Like everyone else, I’ve gotten addicted to Claude Code. (It happens to be the tool I picked up first, but I’ve heard that OpenAI Codex, Cursor, Replit, etc. are all also quite good.) I’m exactly in the most susceptible demographic for it: a former software engineer, product manager, engineering manager, and tech startup co-founder who through circumstance has not had time to code, even for fun, in several years. In my case, it’s because I became a writer and nonprofit leader, and also a dad; but the same thing is happening to tech CEOs and others. All of us are intoxicated by the amazing newfound productivity of AI coding agents, which are now unlocking years-old backlogs of product ideas, bug fixes, and pet projects.

Let’s step back and recap how we got here. The first GPTs were research prototypes, not yet products. GPT-2, launched in February 2019, struggled to produce coherent, logical text (see examples here). Progress came mainly came from scaling up training runs and model size, from hundreds of millions of parameters for GPT-1 to now hundreds of billions or maybe trillions of parameters in the most capable models.

What this created was not a full artificial intelligence but artificial intuition. It could “answer off the top of its head,” it had a superhuman recall for facts, and it could blurt out not just sentences but entire essays. But it was still blurting out all its answers, with no ability to “think” before “speaking,” check its work, or follow an explicit procedure—not even, say, long addition.

By 2022, this had become such a limitation that it was possible to dramatically improve GPT-3’s performance on mathematical reasoning problems simply by concluding the prompt with “Let’s think step-by-step,” which encouraged models to work through the problem explicitly rather than trying to blurt out an answer. Soon this approach was built into the product, in a new class of “reasoning” models, such as OpenAI’s o1, that were given the ability to “think”—that is, to talk to themselves in a scratchpad—before producing a response.

In parallel, the models’ coding ability was growing. At first, LLMs were built into the IDE (“integrated development environment,” kind of a text editor on steroids that software engineers use to write code), and they took the form of autocomplete: the developer would start to type code, and the LLM would complete it. This made it faster to write many routine functions. Later AI-assisted IDEs could write or modify an entire snippet of code from an English description, or answer questions about the code, or discuss it with the engineer. Then the AI got good enough that you could just tell it what to do next, and the output was reliable enough that you didn’t even really need to review it carefully, at least for low-stakes hobby projects—what Andrej Karpathy dubbed “vibe coding.”

The other thing that developed in parallel was the models’ ability to pursue goals, as autonomous agents. At the core, of course, an LLM is simply a statistical model of text that predicts the next token; since any predictor can be made into a generator, this allows it to take a prompt and generate a response. A text generator is not an agent and does not pursue goals—but it was clear from the beginning how an agent might be built from them. Just provide it with a small scratchpad and a few tools it can invoke. Then tell it a goal, and run it in a loop: given the goal, make a plan to achieve it, execute that plan, then check if the goal was achieved; if not, replan and begin again; continue until you succeed.

Early experiments with this (such as AutoGPT, March 2023) were toys: they didn’t have enough intelligence, large enough context windows, or coherence across long enough timescales to accomplish anything of note. But all of this has been improving. The models have been been trained in better tool use, and have been given tools including web search, file access, and code execution; they’ve been given larger context windows, now up to 1 million tokens; and they have steadily been increasing in long-term coherence. Indeed, the length of task (in human-equivalent minutes) that a model can perform has become a key metric of AI progress:

Put all of this together—the “reasoning” mode, better coding, and greater agency—and by late last year we had crossed a tipping point: Some software developers stopped writing code themselves, and started letting agents write 100% of it. The job of the engineer became planning, identification of tasks, directing the AI to specific goals, possibly giving high-level technical direction, testing the output, and (perhaps, depending on how fastidious you are) reviewing the code that is generated. That is, software engineers are becoming more like product managers, engineering managers, and tech leads—as I predicted, humanity stepping up into management. Andrej Karpathy says the term “vibecoding” no longer does justice to what’s possible: it’s now “agentic engineering.”

I felt this shift personally. In early 2025, I was able to use Cursor to write a Python script to analyze CSVs of my workout logs and make some charts of my exercise performance. AI sped me up a lot, especially since I wasn’t familiar with libraries for processing and charting data (like pandas or Matplotlib). The LLM saved me a lot of time reading docs and tinkering; I could focus on the core task I was trying to accomplish. But I was still directing the coding at a detailed level. By early 2026, I was able to act as the product manager, not the developer: to ask Claude Code to simply implement my ideas. Already I’ve built a personal todo app with web and mobile UIs that works offline; a site uptime monitor for the Roots of Progress’s various websites; and an iPad game for my preschool-age daughter to learn two-digit addition.

Part of why this is such a huge unlock is that writing code demands a level of focus I simply don’t have these days. It requires multi-hour blocks of uninterrupted time where you can get your head deep into the problem and the code. Between my day job, my parenting duties, and the need in my 40s to get regular sleep and exercise, that kind of hobby just isn’t possible. But directing coding agents is a different thing altogether: it can be done on “manager schedule” rather than “maker schedule.” Garry Tan describes it using the metaphor that it’s as if he used to be a competitive runner (i.e., engineer) who got a knee injury (went into management). But now he has a knee replacement (coding agents)—and it’s a bionic knee, better than before. Shopify CEO Tobi Lutke has been coding up a storm, and even used Claude to create a viewer for his MRI data.

Out of the box, coding agents lack training and professional maturity. They’re like junior engineers, very smart but fresh out of college and operating like cowboys. When I started my first app, I suggested to Claude Code that I write a product spec, from which it could create a tech design that I would review, before proceeding to implementation. Oh, that sounds like too much process, it told me, why don’t you just tell me your idea and I’ll whip it up? OK fine, we’ll try it your way, I thought. It worked well at first, but then as these things always do, the app started to get buggy. I soon realized Claude wasn’t even writing automated tests (a very basic practice). Over time, I’ve leveled up my Claudes with best practices from the software world: automated regression tests; “test-driven development,” in which you write the tests before the code to make sure the tests actually catch bugs; doing each change on a separate branch which gets reviewed and tested before it is merged into the main line; creating separate testing environments so as not to interfere with real production data; etc.

At first, I thought I would do this by writing one big practices document and having each agent review that at the beginning of each session. But it turns out there’s a better way: “skills”, which are brief documents describing one procedure or technique. These can range from how to use a particular app framework to general best practices the agent should always follow. Agents like Claude Code are able to ingest a large volume of skills, holding only brief, high-level descriptions of each in their context window, and searching for skills when they might be relevant, so the full text can be brought into context only when needed. People are publishing their skills, and there are entire skills marketplaces. I started with a set of basic engineering practices from Jesse Vincent, and then have been writing my own as I notice things Claude could do better. Well, of course, I haven’t been writing them: I’ve been having Claude draft them, and then I’ve been reviewing and commenting on them. The experience is much like having a highly trainable employee who takes feedback and earnestly attempts to improve.

When I started, I would delegate a task and then do something else for a little while until the agent was finished. But once I had 3–4 agents working in parallel across two or more projects, I found that I was fully occupied just reviewing their work and prioritizing next steps; it took all my focus just to keep them busy. What this means is that I’m now able to make progress on software development at the speed of my own review and decision-making—which is amazing.

Stepping back, I think a lot of progress since ~GPT-3 has been in taking the core intuitive faculty provided by statistical language models and adding layers of self-monitoring and self-control, such as reasoning and skills. I find it remarkable how much LLMs are aided by some of the same practices that help humans be more effective: working problems out on a scratchpad, planning before executing, and all of the structure and practices that human engineers, designers, and product managers put in place around software development. Elsewhere, Wilson Lin at Cursor reports on an experiment with getting a large team of agents to implement a web browser from scratch, a large undertaking (although one for which there is already a comprehensive set of formal specifications and acceptance criteria). Just getting a bunch of agents to work off of one big shared task list was too chaotic. What worked was having certain agents dedicated to planning—assessing status and figuring out what was needed next to reach the goal—while other agents acted as implementers, picking tasks off the plan and getting them done without worrying too much about the big picture. Again, systems of self-monitoring and self-control.

The biggest limitation on these systems right now, it seems to me, is memory. They start each session like Leonard Shelby from Memento, with no short-term memories, needing to review all their notes to get context. This is a very limited form of learning. An LLM can’t develop intuition or taste post-training—which, as Dwarkesh pointed out, means it can’t learn on the job the way a human does. Claude’s memory file generated from our chats is about 400 words, ChatGPT’s is not much over 100; a human assistant who had talked to me as much as they have would have a much deeper understanding of me. No doubt this limitation, too, will be removed sooner or later; I agree with Ethan Mollick when he suggests that this will be transformative.

In any case, to produce an acceptable app right now, even a simple one, requires me to act as the product manager and the tech lead. I’m giving direction about what to do; I’m training the agents in best practices and watching to make sure they follow them; I’m even exercising high-level oversight about technical decisions and making technical suggestions. I’m only capable of this because I spent almost twenty years in the tech world, doing these roles professionally across several teams and projects. I can only imagine that users without that experience would have a hard time creating an app of any size and complexity without getting bogged down in confusing product design, annoying bugs, and slow performance. But on the current trajectory, we’re only a year or so away from whole teams of agents that work together like a complete dev shop. A client could come to the process with only a vague, high-level idea of what they need. A product manager agent would interview them to discover requirements. The PM would write a product spec, and a design agent would create UI mockups, both of which the user could review and comment on. Once the spec and design were approved, an engineering agent would produce a tech design; perhaps a second agent with fresh context would review and revise it. A planner agent would turn it into a task list, and a team of implementer agents would execute the coding tasks in parallel, with reviewer agents examining the code for bugs, weaknesses, and best practices. The app would periodically be presented back to the client for user testing and feedback, for as many rounds of iteration as needed to leave the client fully satisfied. On the whole, it would be much like the process performed by humans, but it would take orders of magnitude less money and time.

This is going to change the nature of software. Already I notice a shift in my product thinking: instead of designing an app for a market, I can design it for myself. I don’t have to worry about any other user’s requirements, about competition and gaps, about user onboarding, or about pricing and payment. It simplifies a lot, compared to being a tech founder. I don’t think all software will be bespoke in the future, not even nearly all of it, but there will be a lot more custom software than before.

And it is not hard at all to envision how this will play out in other industries whose work essentially consists of talking to people and producing documents: law, accounting, graphic design, business consulting. Virtual service shops, doing in hours what used to take weeks, for hundreds of dollars instead of tens of thousands.

It is now impossible not to see that this is going to change the world, indeed that the change has already begun and is underway in earnest.

Thank you for having me—it’s great to be here. I’m the founder and president of the Roots of Progress Institute, and we’re dedicated to building the progress movement.

There’s a lot of overlap between the progress movement and the abundance movement—a lot of shared vision and goals, and a lot of the same people are involved. So I was invited here to talk about progress and how it’s relevant to abundance.

I agreed to come, because I love abundance. I love it as a vision and a goal. And I love it as a direction for the Democratic party and for the political left.

The left styles itself the party of science. That’s good, because abundance needs science, in the long term. But it’s not enough: abundance also needs technology and economic growth.

Technology and growth are historically how we have created the abundance we already enjoy. Abundance, after all, is relative, and we have a lot compared to the past. We should always remember how lucky we are to live today instead of 200 years ago—when homes didn’t have electricity, refrigerators, or toilets; when almost no vaccines existed to protect us from disease; when a room like this would have been lit not with clean electric lights but with smelly, polluting oil lamps; when a gathering like this would in fact have been impossible, because to travel across the country was not a six-hour plane flight, but a six-month trek by horse and wagon, Oregon Trail style.

Just as we have abundance compared with the past, we should hope that the future can be just as abundant, compared to the present. Indeed, the recent book Abundance by Ezra Klein and Derek Thompson opens with a imagined scene from a technologically advanced future: energy from solar, nuclear, and geothermal; desalination using microbial membranes; indoor farms where food is grown with light from LEDs; lab-grown meat; drone deliveries; longevity drugs made in space-based pharmaceutical plants; supersonic passenger jets; artificial intelligence raising everyone’s productivity so we can all enjoy more leisure.

The historic pattern of increasing abundance over time, and the hope and promise of an even more abundant future, is what used to be commonly known as progress.

Progressives used to believe in progress. The old left was not just the party of science—it was a party of science, technology, and growth.

Take Teddy Roosevelt—a progressive if there ever was one. One of the signature achievements of his administration was the Panama Canal. This was a massive engineering project, a triumph of hydraulic engineering technology, celebrated at the time as the 13th Labor of Hercules. When FDR launched the New Deal, one of his signature projects was the Tennessee Valley Authority, which created hydroelectric dams to provide electricity for an entire region. And JFK, of course, is the president who called for putting a man on the Moon—one of the greatest technological achievements not just of its era, but of all time. When JFK gave his famous speech about the Apollo program (the one where he said “we choose to go to the Moon”), he put it in the context of the grand story of human progress. He invoked that narrative to inspire the people and justify his aims.

The Moon landing, in 1969, was a peak moment for America: literally the highest we had ever reached. But after that, something changed.

The children of the ‘60s were starting to see technology and growth as responsible for some of the worst problems of the 20th century, such as environmental damage and the horrors of war. Growth had created pollution and acid rain. Technology had created machine guns, chemical weapons, and the atomic bomb.

But instead of just being anti-pollution and anti-war, the new left decided to become anti-technology and anti-growth. And so a party of science, technology, and growth became just a party of science.

That was a mistake, a costly historical error that we should now correct.

What has 50 years of the anti-growth mindset gotten us? Stagnation and sclerosis. We can’t build anything in this country anymore. We can’t build the homes we need to make our cities affordable. We can’t build the transit we need to make those cities livable. We can’t build energy infrastructure, either generation or the power lines to connect it to the grid.

Without economic growth, we don’t have the engine that raises the standard of living for everyone and helps people lift themselves out of poverty. Without growth, people feel they are playing a zero-sum game—and they turn to exclusion. “No, you can’t move to my neighborhood, it’s too crowded.” “No, you can’t immigrate, you’re going to steal my job.” We want abundance thinking instead: “Yes, move to my neighborhood—we’ll build more homes!” “Yes, immigrate here—there’s so much work to be done, we need all the help we can get.”

I think people have grown weary of the anti-growth mindset, weary of stagnation and sclerosis. So I’m glad to see that abundance is now a politically winning issue. And I would love to see it be a new direction for the left.

But the right is also moving to embrace technology and growth—or rather, they’re doing that with one hand, while fighting those things with the other. On the one hand, they’ve embraced technologies like nuclear power, supersonic flight, and AI. On the other hand: They’re fighting vaccines, one of the greatest technologies ever invented. They’re defunding research into mRNA, one of the most promising genetic engineering techniques. They’ve disrupted research funding broadly. They’ve disrupted immigration, including high-skilled immigration, which is one of our best talent pipelines into R&D. And they’ve put tariffs on everything, which almost any economist will tell you is hurting affordability and slowing growth.

So the right has at best a mixed record on abundance. The left can still be the party of abundance, if it wants to be.

But it won’t be easy. It will be uncomfortable. Because to become the party of abundance requires truly embracing technology and growth—and the left has developed an allergic reaction to those things. So there’s some work to be done: some lessons to be unlearned, some old habits to be broken.

But I’m excited to help with that work, and I invite you to talk to me about it. I’m eager to see the party of science become once again a party of science, technology, and growth. And I look forward to the day when progressives once again believe in progress.

PS: I would also like to see the right become, more consistently, the party of abundance. I would like to see both parties competing to be the party of abundance! At some point I may write up an analogous “message to the right.”

Safety, security, and resilience against these hazards is not the default state of humanity. It is an achievement, and in each case it came about deliberately.

A striking theme from the history of such achievements is that there is rarely if ever a silver bullet for risk. Safety is achieved through defense in depth, and through the orchestration of a wide variety of solutions, all working in concert.

Recently, in a private talk, I gave a historical example: the history of fire safety. It resonated so strongly with the audience that I’m writing it up here for wider distribution.

Up until and through the 1800s, city fires were a great hazard. Neighborhoods were full of densely packed wooden structures without flame-retardant chemicals, fire alarms, or sprinkler systems; open flames were used everywhere for lighting, heating, and cooking; there were no best practices in place for storing or handling combustible materials; fire departments lacked training and discipline, and they worked with inadequate equipment and insufficient water supply. All this meant that large swaths of cities regularly burned to the ground: Rome in AD 64; Constantinople in 406; London in 1135, 1212, and 1666; Hangzhou 1137; Amsterdam 1421 and 1452; Stockholm 1625 and 1759; Nagasaki 1663; Boston 1711, 1760, 1787, and 1872; New York 1776, 1835, and 1845; New Orleans 1788 and 1794; Pittsburgh 1845; Chicago 1871; Seattle 1889; Shanghai 1894; Baltimore 1904; Atlanta 1917; and Tokyo 1923 are just a short list of the most well-known.

Fire is not unknown today, but it is far less lethal, and great city fires consuming multiple blocks are largely a thing of the past. Today, if you see a fire truck on the street with its sirens blaring, it is more likely to be responding to an emergency medical call than to a fire. Even if the truck is responding to a fire call, it is more like likely to be a false alarm than an actual fire.

How was this achieved?

Better fire-fighting. Pumps to douse fires with water have existed since antiquity, but for most of history they were man-powered. With the Industrial Revolution, we got steam-powered and later diesel-powered pumps that can deliver much greater throughput of water, and at greater muzzle velocities to reach higher floors of buildings. In the 20th century, horse-drawn fire engines were replaced with fire trucks that could get around the city faster and more reliably.

A high-throughput engine, however, needs a high-volume source of water. In ancient and medieval times, water was provided by the bucket brigade: two lines of people stretching from the fire to the nearest lake or river, passing buckets by hand in both directions. A much better solution was the fire hose, invented in the late 1600s (and improved in strength and reliability over the centuries through better materials, manufacturing, quality control). The fire hose not only allowed a fire engine to be connected to a water source, it also allowed the fire-fighters to get in closer to the base of the fire and dump water directly on it, which is far more effective than just spraying the building from the outside.

A fire hose can be inserted into a natural water source like a pond or cistern, but one of these might not be handy nearby, and they aren’t pressurized, so all the pumping force has to be supplied by the fire engine. They also contain debris that can clog the intake and block the flow. Eventually, cities were outfitted with regularly spaced fire hydrants connected to the municipal water supply. A water system designed to supply city residents with daily needs, however, often proved inadequate in an emergency; these systems had to be upgraded to supply the large bursts that big fires demanded. This is a matter of serious engineering: 19th-century fire-fighting journals are full of technical details and mathematical calculations attempting to precisely nail down questions of optimal hydrant distribution or nozzle size, or the pressure required to force a certain volume of water to a given height at a particular angle.

Finally, fire-fighting teams needed improved organization. Traditionally, fire-fighters were volunteers, often rowdy young men with no training or discipline (there is at least one story of a fist fight breaking out between two rival teams who arrived at a fire at the same time). In the 19th century, fire departments were professionalized and were organized more formally, along almost military lines, as befits responders to a life-threatening emergency.

Faster alarming. Fire, like many of our most dangerous hazards, is a chain reaction. Chain reactions grow exponentially, which means early detection and response time are crucial. Traditionally, fires were spotted by watchmen, either on patrol or from a watch tower, who then had to run, shout, or ring bells or other alarms to alert the fire fighters.

Electronic communications, first via telegraph and later telephone, provided a much faster way to get the alarm to the fire department. The telephone lines could be busy, however, so in the 20th century the 911 emergency response system was created to provide a priority channel.

Far better than having a human sound the alarm, however, is doing it automatically. Smoke detectors and other automatic fire alarms caused the fire to “tell on itself,” saving valuable minutes or even hours. Even more effective was the automatic sprinkler, which combined detection and response into one near-instant system.

Reducing open flames. Better than fighting fires, of course, is preventing them. Before the 20th century, flames from candles and oil or gas lamps provided lighting, and fires in wood- or coal-burning stoves provided heat for building, cooking, and industrial processes. The Great London Fire of 1666 is said to have started in a baker’s shop, Copenhagen 1728 was blamed on an upset candle, Pittsburgh 1845 came from an unattended fire in a shed. Even worse, people often kept these fires going unattended overnight, because even starting a fire was difficult before the invention of matches. Medieval regulations required city- and town-dwellers to cover their fires after a certain hour (the word “curfew” derives from the French couvre-feu, “cover the fire”).

Electric lighting and heating greatly reduced this risk. Electric sparks, however, were also a fire hazard—and initially, electrical installations increased rather than decreased fire risk, owing to shoddy electrical products, fixtures, and wiring. The solution here was improved standards, testing, and certification: the fire insurance companies created an organization, Underwriters Laboratories, specifically for this purpose, and its label became a highly valued marker of quality. (I told the story of UL in The Techno-Humanist Manifesto.) Today, our electronics and appliances are so safe that arson is the cause of more fires than either of them.

Safer construction. Preventing fires by eliminating the sparks or flames that ignite them is like lining up dominoes and then trying hard to make sure the first one never gets tipped over: a fragile proposition. Far more robust is to remove their fuel. Wood construction was widespread through the late 19th century, even in dense city neighborhoods: Daniel Defoe wrote that before the Great London Fire of 1666, “the Buildings looked as if they had been formed to make one general Bonfire.”

Today our cities are built of incombustible brick, stone, and concrete. Building codes enforce safety practices to slow the spread of fire both within a building and between buildings. They specify the quality of materials such as brick, mortar, cement, timber, and iron, including the specific tests it must pass; the materials for walls, and their minimum thickness; and the height of non-fireproof structures; among many other details.

Saving lives. By the early 1900s, in advanced societies, the problem of large city fires that spread over many blocks had mostly been solved; fires were often contained to a single building. That was small comfort, however, for those trapped inside the building. Tragedies such as the Iroquois Theatre Fire of 1903 and the Triangle Shirtwaist Fire of 1911 taught us valuable lessons. Exit paths must be adequate to evacuate entire buildings. Doors must remain unlocked, and they should open outwards in case a stampede presses up against them. Fire-resistant material must be used not only for the construction of the building, but for the interior: sofas, beds, curtains, carpets, wallpaper, paneling. Again, building and safety codes specify and enforce these practices.

So fire safety was achieved through the combination of:

• General-purpose technologies: engines, electronic communications, electric light and heat
• Specific inventions: fire pumps, fire hose, fire alarms
• Infrastructure: municipal water supply, telephone lines
• Standards, testing and certification: of electrical products, fire preventing and fire-fighting equipment, building materials, etc.
• Law: building codes and other fire safety codes
• Education and training: in fire departments, among the public

This is a general pattern. Safety requires:

• both prevention and “cure”
• both technical and social solutions
• among technical solutions, both products and systems
• among social solutions, both education and law

We see the same thing in other domains. Road safety, for instance, was achieved through seat belts, anti-lock brakes, crumple zones, air bags, turn signals, windshield wipers, traffic lights, divided highways, driver’s education, driver’s licensing, and moral campaigns against drunk driving. No silver bullet.

When we think about creating safety and resilience from emerging technologies, such as AI or biotech, we should expect the same pattern. Safety will be created gradually, incrementally, through multiple layers of defense, and by orchestrating a wide combination of products, systems, techniques, and norms.

In particular, there is a line of thinking within the AI safety community that tends to dismiss or reject any proposal that isn’t ultimate—fully robust against the most powerful imaginable AI. There’s a good rationale for this: it’s easy to fall victim to hope and cope, and to lull ourselves into a false sense of security based on half-measures that were “the best we could do”; vulnerabilities are often invisible and are revealed dramatically in disasters; such disasters may be sufficiently catastrophic that we can’t afford to learn from mistakes. But I find the all-or-nothing thinking about AI safety counterproductive. We should embrace every idea that can provide any increment of security. History suggests that the accumulation and combination of such incremental solutions is the path to resilience.

Selected sources and further reading:

• Bruce Hensler, Crucible of Fire (2011)
• Harry Chase Brearley, The History of the National Board of Fire Underwriters (1916) and Symbol of Safety (a history of Underwriters Labs, 1923)
• Dennis Nolan, Fire Fighting Pumping Systems at Industrial Facilities (2011)
• W. Fred Conway, Those Magnificent Old Steam Engines (1996)
• Ramon Klitzke, “Roman Building Ordinances Relating to Fire Protection” (1959)
• Lionel Frost and Eric Jones, “The fire gap and the greater durability of nineteenth century cities” (2007)
• John Rainbird, “The Vigiles of Rome”
• Alex Tabarrok, Firefighters Don’t Fight Fires

Historical and primary sources:

• Charles Frederick T. Young, Fires, Fire Engines, and Fire Brigades (1866)
• James Bugbee, “Fires and Fire Departments” (1873)
• James Braidwood, “On the Means of rendering large supplies of Water available in cases of Fire; and on the application of manual power to the working of fire engines” (1844)
• Report of the Commissioners Appointed to Investigate the Cause and Management of the Great Fire in Boston (1873)
• Clarence Goldsmith, “The Use of Pumpers at Fires” (1930) and “Efficient Utilization of Water for Fire Fighting” (1939)
• Journal of the New England Water Works Association, Vol 7 Issue 1
• National Board of Fire Underwriters, Model Building Code (1905)
• Maurice Webster, “What is ‘Fireproof?’” (The Atlantic, 1946)
• FEMA, “America Burning” (report)
• Charles II, An Act for rebuilding the City of London (1666)

I think this is true but irrelevant.

Consider the weather, a prime example of a complex system. We can predict the weather to some extent, but not far out, and even this ability is historically recent. We still can’t control the weather to any significant degree. And yet we are far less at the mercy of the weather today than we were through most of history.

We achieved this not by controlling the weather, but by insulating ourselves from it—figuratively and literally. In agriculture, we irrigate our crops so that we don’t depend on rainfall, and we breed crops to be robust against a range of temperatures. Our buildings and vehicles are climate-controlled. Our roads, bridges, and ports are built to withstand a wide range of weather conditions and events.

Or consider an extreme weather event such as a hurricane. Our cities and infrastructure are not fully robust against them, and we can’t even really predict them, but we can monitor them to get early warning, which gives us a few days to evacuate a city before landfall, protecting lives.

Or consider infectious disease. This is not only a complex system, it is an evolutionary one. There is much about the spread of germs that we can neither predict nor control. But despite this, we have reduced mortality from infectious disease by orders of magnitude, through sanitation, vaccines, and antibiotics. How? It turns out that this complex system has some simple features—and because we are problem-solving animals endowed with symbolic intelligence, we are able to find and exploit them.

Almost all pathogens are transmitted through a small number of pathways: the food we eat, the water we drink, the air we breathe, insects or other animals that bite us, sexual contact, or directly into the body through cuts or other wounds. And almost all of them are killed by sufficient heat or sufficiently harsh chemicals such as acid or bleach. Also, almost none of them can get through certain kinds of barriers, such as latex. Combining these simple facts allows us to create systems of sanitation to keep our food and water clean, to eliminate dangerous insects, to disinfect surfaces and implements, to equip doctors and nurses with masks and gloves.

For the infections that remain, it turns out that a large number of bacterial species share certain basic mechanisms of metabolism and reproduction, which can be disrupted by a small number of antibiotics. And a small number of pathogens once caused a large portion of deaths—such as smallpox, diphtheria, polio, and measles—and for these, we can develop vaccines.

We haven’t completely defeated infectious disease, and perhaps we never will. New pandemics still arise. Bacteria evolve antibiotic resistance. We can sanitize our food and water, but not our air (although that may be coming). But we are far safer from disease than ever before in history, a trend that has been continuing for ~150 years. Even if we never totally solve this problem, we will continually make progress against it.

So I think the idea that we can’t control complex systems is just wrong, at least in the ways that matter to human existence. Indeed, a key lesson of systems engineering is that a system doesn’t need to be perfectly predictable in order to be controllable, it just has to have known variability.¹ We can’t predict the next flood, but we can learn how high a 100-year flood is, and build our levees higher. We can’t predict the composition of iron ore or crude oil that we will find in the ground, but we can devise smelting and refining processes to produce a consistent output. We can’t predict which germs will land on a surgeon’s scalpel, but we know none of them will survive an autoclave.

So we can tame complex systems, and achieve continually increasing (if never absolute or total) mastery over nature. Our success at this is part of the historical record, since most of progress would be impossible without it. The “complex system” objection to the goal of mastery over nature simply doesn’t grapple with these facts.

So here’s mine. At least I can say I’ve been doing it for as long as Dan: nine years running (proof: 2017, 2018, 2019, 2020, 2021, 2022, 2023, 2024). As usual, this is more of a personal essay/reflection, and not so much of an organizational annual report, although I will start with some comments on…

RPI

Over the last three years, the Roots of Progress Institute has gone from “a guy and his blog” to a full-fledged cultural institute. This year we:

• Held our second annual Progress Conference, featuring speakers including Sam Altman, Blake Scholl, Tyler Cowen, and Michael Kratsios (Director, OSTP). The conference has become the central, must-attend event for the progress community: it is sold out each year, with hundreds on the waitlist, and some attendees report it is literally the best conference they have ever attended.
• Inducted the third cohort of our progress writers fellowship, bringing the total to 74 fellows. Our fellows are having impact: Dean Ball helped draft the Trump administration’s AI Action Plan, Madeline Hart has co-authored a book with the CTO of Palantir on revitalizing the American defense industry, Ryan Puzycki helped legalize single-stair buildings in Austin (a key YIMBY reform), and three other fellows have recently had opinion pieces in the NYT or WSJ.
• Announced our first education initiative: Progress in Medicine, a high school summer career exploration program. I’ve previewed the content for this course and I’m jealous of these kids—I wish I had had something like this when I was a teenager!

And the best part about all of these programs is that I don’t have to run any of them! I have a fantastic staff at RPI who deserves credit for all of these, from design to execution: Emma McAleavy, Ben Thomas, Yel Alonzo, and especially Heike Larson—thanks to them for making our programs a success every year.

We’re a 501(c)(3) nonprofit, supported mostly by donations. There’s still time to get in a last-minute end-of-year contribution! The simplest way to support us is to upgrade to a paid (or founding!) subscription.

Or see here for other ways to support us.

Huge thanks to all those who have already given this year!

My writing

Most of my writing effort this year was devoted to finishing The Techno-Humanist Manifesto, an essay-series-cum-book laying out my philosophy of progress. In 2025 I published the last 14 (out of 21) essays in the series, you can read them all here. Also, as just announced, I’ve signed with MIT Press to publish a revised version of the series in book form. The manuscript is out for comment now, and (given typical publishing schedules) I expect the book to launch in early 2027.

I also wrote eight other essays, and ten links digests. I put the links digest on hold after May in order to focus on finishing the book, but I’m working on bringing it back. All subscribers get the announcements and opportunities at the top, but the rest of the digest is paywalled, so subscribe now to get the full version.

The most-liked posts here on Substack were:

• The future of humanity is in management
• How sci-fi can have drama without dystopia or doomerism
• The Unlimited Horizon, part 2
• The Spirit We Lost, part 1
• In defense of slop

The most-commented posts were:

• How much does it cost to back up solar with batteries?
• We should install a thermostat on the Earth
• Where is the YIMBY movement for healthcare?

My longest post, at over 8,400 words, was:

• The Progress Agenda

I now have well over 55,000 subscribers on Substack, up over 68% YOY.

Social media

Here are some of my most-liked posts and threads of the year:

• One model for Elon is that he is perpetually operating in wartime mode: Move as fast as possible, bring overwhelming force, extract superhuman effort, accept collateral damage. A thread. For some reason many people seemed to think I was praising or defending this: I wasn’t, just observing.
• Grocery stores have price transparency and open competition, which health services don’t (replying to a wrong-headed comment by Rep. Jake Auchincloss)
• Everything has to be invented—even buttons
• The moment the Boom XB-1 test plane went supersonic
• “Daddy, why is a hundred a big number?”, and other difficult questions from my 3yo
• “On paper, titanium should be the world’s #1 structural metal”
• When Tyler Cowen just flat-out says that tariffs are bad, you know they are very straightforwardly bad
• A great moment from a legendary interview with Morris Chang of TSMC, on the Acquired podcast
• Waymo achieves 80–90% reduction across all crash/injury metrics. If there were some intervention that achieved this, safety advocates would be demanding its wholesale adoption immediately, even if it were costly or inconvenient

You can join well over 40,000 people who follow me on Twitter, or find me on your favorite social network; I’m on pretty much all of them.

Speaking and events

Like last year, I tried to mostly say no to events and speaking gigs this year, but there were a few I couldn’t refuse. Some highlights of the year:

• I spoke at “d/acc Day” alongside Vitalik Buterin, Juan Benet, Mary Lou Jepsen, Allison Duettmann, and others. My talk was “d/acc: The first 150 years”: a whirlwind tour of how society has thought about progress, decentralization and defense over the last century and a half
• I gave a short talk at Social Science Foo Camp titled “The Fourth Age of Humanity?”, based on ideas that I later wrote up in The Flywheel
• I did a fun Interintellect salon with Virginia Postrel based on her essay “The World of Tomorrow”
• I hosted a discussion series at Edge Esmeralda with the aim of envisioning the future. Each day there was a ~90-minutes session with a theme like AI, health & bio, or energy
• I went to Mojave to watch the first supersonic flight of the Boom XB-1 test plane. Here’s some video I took of the plane taxiing down the runway, and then the pilot getting out after landing and shaking hands with Boom founder Blake Scholl

In 2026 I hope to do more travel, events and speaking. But maybe I’ll just hole up and write some more.

Reading

I put my monthly “what I’ve been reading” updates on hold at the end of 2023 (!) in order to focus on the book. I’d like to bring these back, too. For now, here are some the highlights from my reading this year (that is, things I thought were interesting and valuable to read, not necessarily things I “liked” or agreed with).

The first few book recommendations are free, paid subscribers get the (much longer) full reading update and (below) some thoughts on what’s next for my writing:

Books and other book-length things I read

Or read at least most of:

Max Bennett, A Brief History of Intelligence. A history of the evolution of the brain, from the first animals through humans. It is organized into five major evolutionary steps—to oversimplify: the worm brain, the fish brain, the mouse brain, the monkey brain, and the human brain. This answered some key questions I had on the topic, very well-written, probably my favorite of the year. Hat-tip to @eshear.

Charles Mann, How the System Works, an essay series in The New Atlantis. It covers four of the major systems that form the foundation of industrial civilization and help deliver our modern standard of living: agriculture, water sanitation, electricity, and public health. Mann thinks of these pieces as the start of a curriculum that should be taught in schools—inspired by a group of “smart, well-educated twenty-somethings” who “wanted the hungry to be fed, the thirsty to have water, the poor to have light, the sick to be well,” but “knew little about the mechanisms of today’s food, water, energy, and public-health systems. They wanted a better world, but they didn’t know how this one worked.” Enjoyed this, recommended.

Brian Potter, The Origins of Efficiency, from Stripe Press, a history of manufacturing efficiency. Light bulbs used to cost ~$50 (adjusted for inflation), now they cost 50 cents; how did that happen? This is a comprehensive and very readable overview of the answer to that question and others like it.

To read the rest, subscribe on Substack.

We define slop as “digital content of low quality that is produced usually in quantity by means of artificial intelligence.” … The flood of slop in 2025 included absurd videos, off-kilter advertising images, cheesy propaganda, fake news that looks pretty real, junky AI-written books, “workslop” reports that waste coworkers’ time… and lots of talking cats. People found it annoying, and people ate it up. … “AI Slop is Everywhere,” warned The Wall Street Journal, while admitting to enjoying some of those cats.

Slop touches a nerve today. When Meta announced a product to create massive amounts of AI-generated short-form video, presumably with no goal other than entertainment to capture clicks and eyeballs, even my generally pro-technology circles exploded in disgust and outrage. Now we have education slop, math slop, drug discovery slop, longevity slop, and “urbanist slop.” Slop exemplifies everything wrong with the modern era; it signifies the gap—some would say the chasm—between what technology enables and what promotes human well-being.

I have no praise for slop itself, but we can be more sanguine about it if we see it as a byproduct of a bigger and more important trend.

People make things when the value of the thing exceeds the cost of creation. When the cost of creation in a medium is high, people are careful only to use it for high-value products. If a movie costs tens or hundreds of millions of dollars to make, you can’t afford to make a bad movie (or at least, not very many of them). You’re going to put a lot of effort into making it, and someone who holds the purse strings is going to have to decide if it’s good enough to fund.

Whenever the cost of creation in a medium falls, the volume of production greatly expands, but the average quality necessarily falls, because many of the new creations are low-quality. They are low-quality because they can be—because the cost of creation no longer prohibits them. And they are low-quality because when people aren’t spending much time or money to create something, they don’t feel the need to invest a lot in it. When you can quickly dash off a tweet, you don’t need to edit it or fact-check it, or even have correct spelling or grammar; when you can quickly create an AI illustration, you don’t need to hold it to high standards of composition, color, or even the right number of fingers. Hence slop.

The Internet lowered the cost of publishing to virtually zero, which enabled many low-quality blogs and other web sites. Social media made it trivial to put thoughts online, and made it much easier to find an audience, which enabled a vast amount more low-effort and low-quality posting. Now AI is arriving, and lowering the costs of creation itself, not just publication and audience-building. And it is enabling new and different forms of slop.

But along with slop, lower costs and barriers get us:

• More experimentation. It can be hard to predict how good or great a piece of writing, art, music or video is going to be. Major Hollywood pictures can be disliked by audiences, critics, or both; books often fail to make money or even pay out their advances. Conversely, sometimes an unknown creator comes out with a work that is initially ignored but goes on to fame and/or fortune. Lowering the bar for creation allows for more experiments, more chances to create something high-quality.
• Removal of the gatekeepers. If it’s hard to predict or evaluate what is good, who decides? Editors, producers, etc., who act as gatekeepers to the means of production and distribution. But gatekeepers are imperfect predictors, and they have blind spots. Harry Potter was rejected by twelve publishers before finding one that would take it: how many potentially great books never found that one editor to champion them, and never saw the light of day? Today there are far fewer gatekeepers for writing or podcasting, but they still exist in music and movies; AI will gradually remove these.
• More chance for people to make a start. E.g., there are many good bloggers who never would have gotten started if they had to first find a job as a journalist.
• More runway for works to find their audience. My writing had about 50 subscribers for the first two and a half years (now well over 50,000). Dwarkesh Patel was “was 2 months away from quitting the podcast for 2 years” before becoming a rocket-ship success. Lowering the cost of production allows these experiments to be incubated for years, kept alive by love and sweat, until they evolve into a more valuable form or catch their break.
• More content for niche audiences. When content is expensive, it has to serve a large audience, and everything converges on bland mainstream taste. When the only significant cost is one creator’s time, it only has to find 1,000 True Fans, and there is much more room for a broad and varied menu to serve many different palates.
• More diversity of content and format. When content is expensive, and gate-kept, it becomes the work of Trained Professionals, who are Serious People, and it should follow Formal Conventions. No serious magazine editor would approve a column that ranges widely across psychiatry, philosophy, politics, science, and epistemology, covering everything from book reviews to academic papers to online controversies. But that’s Scott Alexander, and he’s one of the best and most successful writers of our generation.
• Freedom from the tyranny of finance. When content is expensive, it becomes the domain of large corporations, who have a duty to their shareholders and who frequently succumb to the ruthless logic of financial returns. Hollywood today has found the safest returns in sequels, remakes, and the endless continuation of franchises such as Star Wars or the MCU. Low costs give you more ability to work for the love of the craft and for the sake of the art.

Slop is a byproduct of this overall process, the detritus that accompanies greatly expanded production. Slop is at best annoying and frustrating, and at worst a tool for scams or propaganda. But the overall process will, I believe, usher in a golden age of creativity and experimentation.

We don’t have to like slop, of course. We don’t even have to accept it. We can find ways to minimize it.

First, we need better tools for discovery. Just as the explosion of content on the Internet created a need for directories, search engines, and then social media, the next explosion of content will create a need for new ways to search, filter, etc. AI can help with this, if we apply the right design and product thinking. We can create a future equilibrium that is much better than the pre-AI world, where a thoughtful consumer is able to find more targeted, high-quality writing, video, etc. This is a call to action for the technologists who design and build our information supply chain.

But they key word above is “thoughtful.” The explosion of content raises the bar for everyone to be more conscious in your media consumption. The more stuff is out there, the more of it will be like junk food: enticing, tasty, but not nutritious and ultimately unfulfilling. We all need to be mindful in how we direct and spend our precious, limited attention in a world of increasingly unlimited choice. This is a call to action for every individual, and by extension to parents, teachers, psychologists, and moralists.

Short answer: They overlap 80–90%, and if you’re outside both of them you should probably think of them as variations on the same thing. If you look at the Abundance conference and the Progress Conference, for example, there’s a good amount of overlap in the speakers, attendees, and topics.

What are the differences?

• “Abundance” tends to be more wonkish, oriented towards DC and policy, and focuses on reforming regulations and institutions with a goal of efficiency and being able to build stuff again.
• “Progress” is interested in regulatory reform and efficiency, but it’s also interested in ambitious future technologies, from longevity to nanotech. It’s also more focused on ideas and culture, and on history, philosophy, and economics, in addition to just policy.

Again, just look at the conferences: Abundance is held in DC; Progress Conference in the San Francisco area.

In “The Progress Agenda” I sketched out three broad cause areas for the progress movement: regulatory reform, research institutions, and culture. To my mind, the “abundance agenda” is basically the first of those. So I think of the abundance movement as a part of the progress movement.

After this year’s Progress Conference, Charles Mann suggested: “Abundance wants to make sure everyone has a house. Progress wants to make those houses better.” But IMO, the progress movement is interested in both of those things, so that’s not how I think of the distinction.

Some more background for those who are interested:

“Progress” took off in 2019 when Patrick Collison and Tyler Cowen coined the term “progress studies” in The Atlantic. They proposed it as a field of study, but the article galvanized a movement. (The movement has taken off more than the field, so I tend to talk about the progress movement or progress community rather than “progress studies.”)

There were precursors. Steven Pinker’s Enlightenment Now (2018), David Deutsch’s The Beginning of Infinity (2011), and Virginia Postrel’s The Future and Its Enemies (1998) all had progress as a central concept (two of those have “Progress” in the subtitle). Marian Tupy started HumanProgress.org back around 2012. Peter Thiel, Tyler Cowen, and Robert Gordon were all talking about growth vs. stagnation in the 2010s, although they emphasized those concepts more than “progress.” Max Roser at Our World in Data since 2011, and Hans Rosling before him, worked to communicate the reality of progress, especially global development. Of course, the idea of progress has a history going back centuries at least.

Organizations and projects that put “progress” in the name include the Institute for Progress, Works in Progress magazine, the Human Progress project, and of course us, the Roots of Progress Institute.

“Abundance” was popularized by Derek Thompson, who coined the term “abundance agenda” in a 2022 Atlantic piece. It has since taken off politically, especially with Derek and Ezra Klein’s book titled Abundance that launched this year. Again, there were precursors, including Peter Diamandis’s book also titled Abundance from 2012.

Derek and Ezra credit the progress movement as one of their influences in this podcast.

Things with “abundance” in the name include the Abundance Institute, the Abundance Network, the Metropolitan Abundance Project, the Inclusive Abundance Initiative, and even for a while an Abundance Caucus, which got renamed to the Build America Caucus (because “abundance” has become a partisan term on Capitol Hill, owing to Derek and Ezra being Democrats).

Further reading: Ruy Teixeira captured some of the differences between the progress and abundance movements in his post on the Progress Conference; Steve Teles at Niskanen offers a different take on various factions who have adopted “abundance” ideas.

In “The Progress Agenda,” I mentioned children’s books. One in particular caught people’s attention: You Will Go to the Moon, from 1959 (h/t Virginia Postrel via her blog).

I’ve now heard from a few people who are buying used copies for their kids. Snag yours for Christmas before they’re all gone.

Here are a few more progress-related books I know of—some old and out of print, some quite recent; some I’ve read and can recommend, some I’m just intrigued by.

Subscribe to get posts by email, or upgrade to paid to support my work:

Up Goes the Skyscraper!

This one opens with a clear statement of human needs: “Thousands of people want to work and live on the empty city block. It is a small space for so many people. A skyscraper must be built.” And it concludes by saying “Look up at the skyscraper … it is beautiful.”

It goes through many details—I skip the smaller print for my 4-year-old. She loved learning about I-beams and H-beams, however, and she wondered why there were no G-beams or J-beams. (H/t Luca Gattoni-Celli for this one.)

Evidently there is an updated version of this book (I haven’t seen it yet), and it is only one of dozens of “Explore the World” books by the same author, including How a House is Built, Tool Book. and From Seed to Plant. I think my daughter will get some of these for Christmas.

John Deere, That’s Who!

The story of the invention of the polished steel plow. This is a story of industriousness and resilience: Deere is a hardworking blacksmith who has to find a new home after his workshop burns down. It’s also a story of inventiveness: when the farmers in his new town are having trouble plowing through the thick, sticky mud in their fields, Deere creates a slick, lightweight plow out of polished steel. (H/t Gregory Salmieri.)

Magic Ramen: The Story of Momofuku Ando

The invention of instant ramen might seem an unlikely topic, but it’s actually a very well-told story that draws out the problem-solving process (in the way that Anton Howes talked about here).

Ando has a clear goal of creating a ready meal that people could make easily and quickly. He has to solve one problem after another, and he does it by tinkering and experimenting. First he has to find a good recipe for noodles that don’t stick or fall apart; then he has to figure out how to get the soup flavor into the noodles; and finally he has to figure out how to make it cook in just a few minutes using boiling water from a kettle, instead of cooking for several minutes on a stove. At each stage he doesn’t know what to do and tries many things that don’t work before finally hitting on the solution. A story of vision, drive and persistence, made tangible and real. (I confess to editorializing a bit by inserting at each stage, “and he was very frustrated, but he didn’t give up. He kept trying.”)

Working Boats: An Inside Look at Ten Amazing Watercraft

This one is beautifully illustrated and does not skimp on the detail. I didn’t realize there were so many specialized boats, seemingly one for every species of seafood, as well as tugboats, fire boats, and more. (H/t Katherine Boyle via Twitter.)

Note for parents: most of the books above are probably best for ~6- to 7-year-olds, but I was able to adapt them for my daughter when she was only 3 or 4 by simplifying the language and explaining some of the background concepts.

Le Livre des Progrès

This one is in French and I don’t know if there’s an English translation, but it’s so remarkable I had to mention it. It’s literally titled “The Book of Progress,” and every page shows how we used to do things, and how we do them now.

Here’s a sample page, headlined “From the plow to the tractor.” The page opposite is about construction machines:

Here’s another sample page, “From the broom to the washing machine”:

I’d love an English translation of this.

“We Were There”

“We Were There” is a series of chapter books for older kids, written in the 1950s and ‘60s, basically historical fiction meant to teach history. The books include We Were There at the First Airplane Flight (pictured above), We Were There at the Driving of the Golden Spike (about the completion of the transcontinental railroad), and We Were There at the Opening of the Erie Canal. I haven’t read these yet to evaluate them.

Those are the most remarkable I’ve found so far. What else is out there? Please suggest more in the comments on Substack (ideally not just science explainers, of which there are plenty, but stories of invention or discovery, or explainers about industrial civilization and infrastructure).

This responsibility exists at all levels: from society as a whole, to institutions, to families, down to each individual. Companies should strive to design healthier products—snack foods that aren’t calorie-dense, smartphones with screen time controls built in to the operating system. There is a role for law and regulation as well, but that is a blunt instrument: there is no way to force people to eat a healthy diet, or to ensure that students don’t cheat on their homework, without instituting a draconian regime that prevents many legitimate uses as well. Ultimately part of the responsibility will always rest with individuals and families. The reality, although it makes some people uncomfortable, is that individual choices matter, and some choices are better than others.

I am reminded of a study on whether higher incomes make people happier. You might have heard that more money does not make people happier past an annual income of about $75k. Later research found that that was only true for the unhappiest people: among moderately happy people, the log-linear relationship of income to happiness continued well past $75k, and in the happiest people, it actually accelerated. So there was a divergence in happiness at higher income levels, a sort of inverse Anna Karenina pattern: poor people are all alike in unhappiness, but wealthy people are each happy or unhappy in their own way. This matches my intuitions: if you are deeply unhappy, you likely have a problem that money can’t solve, such as low self-esteem or bad relationships; if you are very happy, then you probably also know how to spend your money wisely and well on things you will truly enjoy. It would be interesting to test those intuitions with further research and to determine what exactly people are doing differently that causes the happiness divergence.

Similarly, instead of simply asking whether social media makes us anxious or depressed, we should also ask how much divergence there is in these outcomes, and what makes for the difference. Some people, I assume, turn off notifications, limit their screen time, put away their phones at dinner, mute annoying people and topics, and seek out voices and channels that teach them something or bring them cheer. Others, I imagine, passively submit to the algorithm, or worse, let media feed their addictions and anxieties. A comparative study could explore the differences and give guidance to media consumers.

In short, we should take an active or agentic perspective on the effects of technology and our relationship to it, rather than a passive or fatalistic one. Instead of viewing technology as an external force that acts on us, we should view it as opening up a new landscape of choices and possibilities, which we must navigate. Nir Eyal’s book Indistractable is an example, as is Brink Lindsey’s call for a media temperance movement.

We should also take a dynamic rather than static perspective on the question. New technology often demands adjustments in behavior and institutions: it changes our environment, and we must adapt. For thousands of years manual labor was routine, and the greatest risk of food was famine—so no one had to be counseled to diet or exercise, and mothers would always encourage their children to eat up. Times have changed.

These changes create problems, as we discover that old habits and patterns no longer serve us well. But they are better thought of as growing pains to be gotten through, rather than as an invasion to be repelled.

When we shift from a static, passive framing to a dynamic, agentic one, we can have a more productive conversation. Instead of debating whether any given technology is inherently good or bad—the answer is almost always neither—we can instead discuss how best to adapt to new environments and navigate new landscapes. And we can recognize the responsibility we all have, at every level, to do so.

The bold, ambitious future is waiting for us. How do we pursue it?

We have the flywheel of progress on our side: progress is a self-reinforcing process. The bigger the world economy, the more resources we have to invest in research and new ventures, the more we will increase total factor productivity, and the more we will grow the economy, in an accelerating cycle.

But the cycle does not operate purely through technological and economic factors. It is mediated by institutions: universities and laboratories, VC and other finance, law and regulation. Institutions develop and allocate both human and financial capital. And institutions, along with all of our activities, are shaped by culture.

Over most of the long sweep of history, institutions and culture improved along with science and technology. In the last four centuries, we developed the scientific method, constitutional democracy, research universities, limited liability corporations, stock exchanges, venture capital. Many of these developments were motivated by progress; all of them helped accelerate it.

But in the 20th century, this loop of the flywheel broke down and even went into reverse. Our institutions today are saddled with the twin legacy of the 20th century: the legacy of technocracy, which left our institutions overly centralized and bureaucratized, and the legacy of the counterculture, which left us with an obstructionist vetocracy. And so, as I described in the introduction, we have deceleration in total factor productivity and per-capita GDP, fields such as manufacturing and transportation that are still using the same basic technologies as in the 1960s, rising costs in housing, education, and healthcare, and a general inability to build or to operate.

Our societal sclerosis is partly the result of deliberate sabotage, motivated by the romantic backlash against progress: the view of material progress as a “destructive engine” and the goal of controlling it, slowing it down, or stopping it altogether. And it is partly a result of natural causes, the institutional senescence that can be staved off only through eternal vigilance. That kind of vigilance requires that our most capable leaders are motivated to maintain institutional competence, and that the public still cares enough about progress to support their efforts. Those values are fostered by a positive and confident philosophy of progress; they disintegrate without one.

The progress agenda is thus one of institutional and cultural reform, in three main cause areas: our laws and regulation, the way we manage research, and the culture of progress itself.

Law & regulation

There are problems we don’t yet know how to solve: curing cancer, or building nanotech. For these, we need advances in science and technology. Then there are problems we know how to solve, but prevent ourselves from solving. These are social problems, and they need legal and regulatory reform.

Building enough housing, for instance, is technologically a solved problem. And yet, we don’t build enough housing to meet demand, and both home prices and rents continue to hit all-time highs. One major index of US home prices has risen over 4x since 1990. The median single-family home price is now 5x the median household income; a price-to-income ratio of 3 is generally considered affordable. In 1990, over 70 of the top 100 US metro areas were under this ratio; now almost all of them are over it.¹ An influential essay in Works in Progress magazine titled “The Housing Theory of Everything” argues that, in addition to the direct cost, more expensive housing hurts productivity, since people can’t move to where better jobs are; slows innovation, which benefits from density of talent; and makes it harder for people to have as many children as they would like.²

Economists such as Ed Glaeser argue persuasively that the problem lies in land-use regulations that artificially restrict housing supply. Since the 1960s, he writes in a Brookings report, we “changed from a country in which landowners had relatively unfettered freedom to add density to a country in which veto rights over new projects are shared by a dizzying array of abutters and stakeholders.”³ In short, housing is a social problem. We have to get out of our own way, by reforming the zoning restrictions, permitting processes, and building codes that restrict supply.⁴

But the building problem is much broader than just housing. It affects any kind of infrastructure: transit, factories, power plants, electrical lines. As of the end of 2024, almost 1.6 TW of electrical generation capacity and over 1 TW of storage capacity was waiting in queues to connect to the grid.⁵ “There is more power in the queue than on the grid today. The average wait time in the interconnection queue is five years and growing, primarily due to permitting timelines. In addition, many projects are cancelled due to the prohibitive cost of interconnection.”⁶

The permitting problem is broad and varied, but one law at the core is NEPA (the National Environmental Protection Act). NEPA requires that all federal agencies, in any “major action” that will significantly affect “the human environment,” must produce a “detailed statement” on those effects, now known as an Environmental Impact Statement (EIS). As Eli Dourado recounts, in the early days, an EIS was less than ten pages, and would go unchallenged. But activists learned to obstruct projects using litigation, by charging that an EIS failed to address some important detail. Every time they won, it raised the standard for how detailed an EIS had to be. Court rulings set precedents that “major action” actually means any action, and “human environment” actually means the entire environment. Over the decades, this process ratcheted up, until today, an average EIS runs several hundred pages (plus over a thousand pages of appendices!) and takes almost 5 years to complete. And if a federal action will not affect the environment, it requires yet another type of statement to confirm this—which now creates more paperwork than the EISes themselves.⁷

NEPA has now become a barrier to projects including housing, transmission lines, semiconductor manufacturing, congestion pricing, and even offshore wind.⁸ None of this was anticipated or intended by Congress or even by environmental activists, who now find even their own “green” projects stifled by the law. Dourado recommends four reforms to bring NEPA back to its original purpose: trust agencies to make findings of no impact without an assessment report; do not require public hearings or comment for every environmental review; raise the bar for judicial injunctions; and establish an exemption for decisions with a strong national interest.⁹

NEPA is a procedural law: it mandates a process to be followed, but doesn’t require any specific outcome. Alec Stapp and Brian Potter at the Institute for Progress contrast this with substantive environmental regulation, which mandates outcomes, such as limiting the amount of carbon monoxide that can be emitted from an automobile. Procedural laws are more flexible, but that very flexibility allows them to be weaponized by activists against growth. “We need a new era of environmentalism that learns from the successes and failures of the past,” they conclude. “Environmentalists rightly tout triumphs over acid rain, ozone depletion, DDT, and lead exposure. But these wins were not the result of preparing ever longer environmental impact statements for specific projects. They were the product of putting a price on pollution, via cap and trade programs, or outright banning a pollutant when necessary.”¹⁰ We could use a shift away from procedural and towards substantive regulation.

More broadly, Michael Catanzaro suggests that we have conflated permitting with compliance. Our processes emphasize obtaining approval, but what actually matters is following the rules. Instead of lengthy and burdensome permitting processes, he suggests, permitting should be made a simple matter of submitting information and certifying that a project will follow the law. Permits would be approved by default unless the information were incomplete or fraudulent, within a tight time window, say 90 days. Then, the enforcement focus could shift to compliance: auditing that a project is in fact following applicable regulations, such as pollution limits.¹¹

Permitting affects all projects, but some industries get special treatment. Nuclear energy, for decades, has faced one of the most burdensome regulatory regimes, which make it slow and expensive to build nuclear plants, and ultimately make nuclear power uneconomical. NuScale Power, after going through its design certification with the NRC (Nuclear Regulatory Commission), complained that they had been “required to evaluate events with frequencies orders of magnitude less than the Commission’s safety goals or the 10–6 per reactor year limit the Commission has stated as a threshold of credibility, and in some cases even where no core damage would result.” The costs of the application exceeded half a billion dollars; NuScale states dryly that this level of effort “may not be repeatable for future reviews.”¹²

It doesn’t have to be this way: one analysis of worldwide nuclear construction since 2000 finds that in South Korea, India, China, and Japan, costs are about 4x lower than in the US or UK.¹³ France built enough nuclear plants in the 1980s and ‘90s that about two-thirds of its electricity is now nuclear (compared to about 20% in the US and only 10% worldwide).¹⁴ We could do the same. There are many credible proposals for reform,¹⁵ but fundamentally what we need is change in regulatory culture: a commitment to balance costs and benefits, and an NRC that sees the development of energy as its job.

Or take the FDA. To bring a new drug to market now takes around 12 years and over $1 billion,¹⁶ and the costs have grown over time, as noted in a well-known paper by Jack Scannell et al that coined the term “Eroom’s Law” for this phenomenon. The paper suggests four factors driving the cost of drugs; some of them are inevitable low-hanging fruit effects, but one of them is over-cautious regulation: “Each real or perceived sin by the industry, or genuine drug misfortune, leads to a tightening of the regulatory ratchet, and the ratchet is rarely loosened, even if it seems as though this could be achieved without causing significant risk to drug safety.”¹⁷ In many cases, the FDA has been too conservative in its approvals, adding needless delay that holds back treatments from patients. Omegaven, a nutritional fluid given to patients with digestive problems (often infants) that helped prevent liver disease, took fourteen years to clear FDA’s hurdles, despite dramatic evidence of efficacy early on, and in that time hundreds to thousands of babies died preventable deaths.¹⁸ A former FDA regulator has stated that in the 1980s, approval of the new drug application for human insulin was delayed despite compelling evidence of safety and effectiveness, because his boss said: “If anything goes wrong, think how bad it will look that we approved the drug so quickly.”¹⁹ One Bayesian analysis that modeled the tradeoff between approving bad drugs and failing to approve good drugs found that the FDA is much more conservative than optimal, especially for terminal illnesses with no existing therapies.²⁰

The problem with regulatory agencies is not that the people working there are evil—they are not. The problem is the incentive structure: Regulators are blamed for anything that goes wrong. They are not blamed for preventing growth and progress. They are not credited when they approve things that lead to growth and progress. So all of the incentives point in a single direction: towards more stringent regulations. And when regulations are put in place, they become very hard to remove. So regulation ratchets upwards.

Nuclear and drugs, at least, were not outright banned. Supersonic flight was. Since 1973, FAA regulations have made it illegal to fly faster than the speed of sound over land in the US; several other countries followed suit.²¹ A 2025 executive order, which I trust will be faithfully implemented, has directed the FAA to repeal the speed limit and replace it with a noise limit instead.²²

This is just an amuse-boucheof the issues involved. Summarizing the theme, Derek Thompson has called for an “abundance agenda” for America, and his book with Ezra Klein on this topic, Abundance,has made a splash in DC.²³ Gavin Newsom has now endorsed an “Abundance Agenda” for California, and a bipartisan “Build America Caucus” has launched, inspired by the same ideas.²⁴

We should take pains to prevent this agenda from succumbing to the tribalism of contemporary politics. Thompson has suggested that it could “take the best from several ideologies”:

It would harness the left’s emphasis on human welfare, but it would encourage the progressive movement to “take innovation as seriously as it takes affordability,” as Ezra Klein wrote. It would tap into libertarians’ obsession with regulation to identify places where bad rules are getting in the way of the common good. It would channel the right’s fixation with national greatness to grow the things that actually make a nation great—such as clean and safe spaces, excellent government services, fantastic living conditions, and broadly shared wealth.²⁵

Growth and progress should be a shared, cross-partisan goal. With progress as the standard to measure our efforts, all sides can then debate, using history, data, and logic, whose policies will actually achieve that goal. A world in which every party was competing to be the party of progress would be far healthier than one absorbed in redistribution and identity politics.

Research institutions

“By many measures, the biological and medical sciences are in a golden age. That fact, which we celebrate, makes it all the more difficult to acknowledge that the current system contains systemic flaws that are threatening its future.”

So begins a perspective piece published in 2014 by four biologists, including the founding chair of the Department of Systems Biology at Harvard, a former president of Princeton, and a Nobel laureate who formerly directed the NIH.²⁶ The paper laments that “biomedical scientists are spending far too much of their time writing and revising grant applications and far too little thinking about science and conducting experiments.” It also warns about “conservative, short-term thinking in applicants, reviewers, and funders”:

The system now favors those who can guarantee results rather than those with potentially path-breaking ideas that, by definition, cannot promise success. Young investigators are discouraged from departing too far from their postdoctoral work, when they should instead be posing new questions and inventing new approaches. Seasoned investigators are inclined to stick to their tried-and-true formulas for success rather than explore new fields.

Others sound a similar warning. In an editorial for the London School of Economics Business Review, Donald Braben and Rod Dowler write: “It is of vital importance right now to avoid suppressing genius in favour of apparent practicality.” They warn that “freedom of research has been severely curtailed”:

Current policies make sense for incremental or near-market research that may well lead to the creation of new technologies based on existing fundamental theories. The casualty of such policies, however, will be hard-to-predict radical discoveries … Uninhibited exploration of these fields would almost certainly reveal unimaginable opportunities for growth and enrichment. However, they are in danger of being strangled by bureaucratic processes that would have denied funding for many of the 20th century’s major discoveries.²⁷

And Sydney Brenner, a medical professor at Cambridge and Nobel laureate, says:

The supporters now, the bureaucrats of science, do not wish to take any risks. So in order to get it supported, they want to know from the start that it will work. This means you have to have preliminary information, which means that you are bound to follow the straight and narrow.²⁸

In an obituary for Fred Sanger, two-time Nobel laureate who pioneered the sequencing of both proteins and DNA, Brenner writes:

A Fred Sanger would not survive today’s world of science. With continuous reporting and appraisals, some committee would note that he published little of import between insulin in 1952 and his first paper on RNA sequencing in 1967 with another long gap until DNA sequencing in 1977. He would be labeled as unproductive, and his modest personal support would be denied. We no longer have a culture that allows individuals to embark on long-term—and what would be considered today extremely risky—projects.²⁹

Similarly, Peter Higgs, whom the Higgs boson is named after, has said:

It’s difficult to imagine how I would ever have enough peace and quiet in the present sort of climate to do what I did in 1964. … Today I wouldn’t get an academic job. It’s as simple as that. I don’t think I would be regarded as productive enough.³⁰

A threat to future scientific breakthroughs is a threat to progress. What has gone wrong?

Accounts differ; what follows is an opinionated and probably oversimplified narrative. In short, I blame two factors: First, the centralization of funding, which has come to be dominated by a small number of large agencies. Second, an increase in competition among researchers, which has created overoptimization for markers of performance that lead to jobs and grants but not necessarily to great science.

The dominance of a few federal agencies in research funding is a post-WW2 phenomenon. During the war, the Office of Scientific Research and Development, under the great Vannevar Bush, developed not only weapons but also peacetime technologies such as radar and penicillin.³¹ Motivated by these successes, both the scientific establishment and the government wanted to keep the collaboration going into peacetime. Bush famously wrote a report to the President, “Science: The Endless Frontier,” in which he advocated for federal funding of basic research as the means to American prosperity and security.³² Bush’s specific proposals were not adopted, but federal funding for science ramped up massively in the 1950s and ‘60s.³³ For decades, US universities have gotten the majority of funding for science and engineering R&D from federal sources such as the NIH ($44B in 2024), the NSF ($7.4B), and DARPA ($4.1B).³⁴

Even as the supply of funding grew, however, demand grew faster. The NIH, for example, received 9,199 grant applications in 1970, and made 3,264 awards; in 2024 it received 6 times as many applications, but only made 3 times as many awards. Thus while the success rate on NIH grants was 30–40% in the 1970s, for the last 20 years it has hovered around 20%.³⁵ It doesn’t help that the NIH’s budget, adjusted for inflation, has not increased since 2003.³⁶

Several consequences flow from all of this.

A key problem with centralization of funding is that any one funder, no matter how well-intentioned and wise, will have blind spots. Katalin Karikó, who shared the Nobel prize in 2023 for mRNA technology, was repeatedly denied grants for her work and was even demoted for failing to raise funds.³⁷ Stanley Prusiner won a Nobel in 1997 for discovering prions, but his work was “treated as heretical,” he struggled to get funding from the NIH, he was dropped by the Howard Hughes Medical Institute, and he initially failed to get tenure at UCSF.³⁸ Research that led to the breast cancer drug Herceptin was denied funding from NIH, and was ultimately supported by private sources including the cosmetics company Revlon.³⁹ Similar stories are told about early work on biodegradable polymers, insulin manufacturing, neural networks, genome sequencing, and expansion microscopy—they are not the exception, but the rule.⁴⁰

The creation of a virtual monopsony for research also allows funders to be insensitive to the needs and desires of researchers. Thus, the grant process has grown slow, cumbersome, and restrictive. One institute within the NIH advises: “Your overall process from planning to award may take as long as two years—even longer if you need to resubmit.”⁴¹ (Resubmission is common.)⁴² The process can be tripped up for trivial reasons, such as minor variations in font size on grant applications.⁴³

Further, since grants come from taxpayer money, they face pressure for oversight and accountability. NIH grants require prior approval for any “change in the direction, aims, objectives, purposes, or type of research training,” including a change in animal model, a shift in emphasis to a new disease area, a change in assay or other new technology, any equipment purchase over $25,000, and any rebudgeting by more than 25% in any category.⁴⁴ Accountability is an understandable goal—but it is directly at odds with the autonomy that scientists need to do great research.

Risk aversion may be exacerbated by the committee-based peer review of grants, a process used by the NIH and NSF. Committees tend to make consensus decisions, but scientific breakthroughs often challenge consensus. And when the committee is formed of experts in the field—essentially, the competitors of the researchers submitting the proposal!—there is a risk that a field converges on a consensus too quickly, prematurely pruning off branches of the search space. Some Alzheimer’s researchers, for instance, believe the field has suffered from an overfocus on amyloid proteins as the cause of the disease; one of the more emphatic has even claimed: “If it weren’t for the near-total dominance of the idea that amyloid is the only appropriate drug target, we would be 10 or 15 years ahead of where we are now.”⁴⁵

To its credit, the NIH, at least, recognizes the problem. Starting in 2004, they created a series of new grant types “out of concerns that the traditional NIH peer review process had become overly conservative, and the belief that NIH required specific means to fund high-risk research.”⁴⁶ The High-Risk, High-Reward Research Program (HRHR) now includes the Pioneer Award, for researchers pursuing groundbreaking new directions, the New Innovator Award, aimed at early-career scientists, the Early Independence Award, which helps junior scientists skip their postdoc and jump-start their lab career, and the Transformative Research Award, for “groundbreaking, unconventional research with the potential to create new scientific paradigms.”⁴⁷

The HRHR program, however, is less than 0.5% of NIH’s budget.⁴⁸ Whatever good these grants have done, they don’t seem to be enough to have reformed the culture of research. Indeed, when the foundation Open Philanthropy reviewed a number of the Transformative Research proposals, they found them to be “a bit on the conventional side.”⁴⁹ Even when researchers are solicited for transformative ideas, they are still influenced by the entrenched culture of risk aversion.

Some of these problems would be alleviated simply by having a larger number of smaller funders. Patrick Collison, a tech CEO and science philanthropist, has suggested that we break up NIH and NSF “into 10+ bodies with fully independent approaches. Every 5-10 years, reassess their budgets.”⁵⁰ There is precedent for a more decentralized funding approach: much of the progress in agricultural science and technology in the US was funded not at the federal level but at the state level, through the state agricultural experiment stations established in the 1880s.⁵¹ This decentralization was crucial in the story of hybrid corn, considered one of the greatest breeding successes of the 20th century: in 1905, work by Edward Murray East on this project was discontinued at the Illinois experiment station, and it survived only because East was invited to continue it at the Connecticut station.⁵²

More broadly than diversification in funding sources, though, we need diversification in organizational structures. The dominance of NIH and NSF has also meant the dominance of a particular model of research: relatively small labs, each led by a principal investigator (PI) who not only runs the lab but also fundraises for it, mostly in the form of relatively small, short-term, project-based grants.

The PI model forces one person to play several roles: research, management, administration, and fundraising; at many universities, teaching is added, not to mention serving as a reviewer for journal papers or grant committees. More importantly, it means that each lab has to sell itself to the outside world, to funders who don’t know the researchers and aren’t working with them closely or talking to them daily. This pushes grants more towards “legible,” incremental work.

Consider, in contrast, the UK Laboratory of Molecular Biology (LMB) in Cambridge, where thirteen scientists won nine Nobel prizes—including Watson & Crick for the structure of DNA, the aforementioned Fred Sanger for protein and DNA sequencing, and Venki Ramakrishnan for the structure of the ribosome. LMB researchers don’t have to rely on external grants: the entire lab is funded as a unit, and research projects are funded out of the core budget. A profile of the LMB in Science reports that researchers are encouraged to tackle big, difficult questions, despite the risk that entails, and that researchers have autonomy and aren’t under pressure to constantly publish. In the words of one group leader: “There’s a tradition of trying to hire smart people and then basically leaving them to it.” Ramakrishnan says that having secure, long-term funding allowed him to focus on discovering the structure of the ribosome. He also says that the lab has a collaborative atmosphere, since the success of any one researcher helps the entire lab do well.⁵³ In a different interview, Sydney Brenner, another one of the lab’s Nobel laureates, commented:

… we never let the committee assess individuals. … We asked them to review the work of the group as a whole. Because if they went down to individuals, they would say, this man is unproductive. He hasn’t published anything for the last five years. So you’ve got to have institutions that can not only allow this, but also protect the people that are engaged on very long term, and to the funders, extremely risky work.⁵⁴

James Phillips, a neuroscientist and former Science and Technology Advisor to the Prime Minister of the UK, points out that many of these features of the LMB also apply to celebrated institutions such as Bell Labs and Xerox PARC. All these labs had “a transformative impact on the world; a widely reported unique culture distinct to the mainstream; minimal obstacles to research; ability to pursue work that defied consensus; and a highly collaborative environment.” All had a single major funding source, providing “hands-off” funding that is open to speculative work, on a multi-decade time horizon. All had shared lab resources, rather than each lab accumulating its own resources in its own little “empire.” All allowed researchers to focus on research instead of management, administration, and teaching. And all evaluated their researchers internally, which is possible only with single-source block funding. In contrast, the typical approach has “each researcher being pulled in a different direction by different external funding criteria, in an atomising and soloist manner preventing the emergence of a vibrant community of people pulling in the same direction.”⁵⁵

A modern biology lab applying some of these lessons is the Arc Institute, based in Palo Alto and partnered with Stanford, Berkeley, and UCSF. Investigators at Arc get their labs fully funded for renewable 8-year terms to pursue curiosity-driven research; institute co-founder Silvana Konermann says that they “want to take away the short-term demand to produce output.”⁵⁶ This is made possible by $650 million in founding pledges, including from institute co-founder Patrick Collison. Although Arc was only founded in 2021, it has already announced the largest AI model trained on genomic data, which they have used to design an improved bacteriophage; a virtual cell model that they say is “to the best of our knowledge … the first model to consistently beat simple linear baselines;” and a new DNA editing technology that can make far more sophisticated edits than CRISPR.⁵⁷

Another alternative to the principal investigator model is exemplified by DARPA. The DARPA model focuses on a program manager with a vision for an ambitious technological goal. The PM contracts out to labs to perform the work, but they control the vision and the budget. This model lends itself well to ambitious technological prototypes, and was the origin of the Internet and GPS.⁵⁸ One privately-funded research organization based on this model is Speculative Technologies, which aims to unlock “big-if-true” technologies through “multi-year programs run by program managers with wide-ranging authority to coordinate several projects towards a single vision.”⁵⁹ The newly-formed UK research agency ARIA also uses ARPA-style PMs.⁶⁰

The PI model also doesn’t lend itself well to big scientific goals that require a large, coordinated team and a lot of funding: sequencing the human genome, or mapping the neural connections of the brain. An increasingly popular alternative is the Focused Research Organization (FRO), which is organized like a startup team with a CEO, but as a nonprofit with a goal to produce a public good such as an open scientific dataset.⁶¹ Convergent Research has now launched FROs for technologies including microorganisms for use in synthetic biology, software for proving mathematical theorems, ocean-based carbon dioxide removal, and brain-computer interfaces.⁶²

Another problem caused by increased competition among researchers is the focus on the peer-reviewed journal paper as the measurable unit of work, with citations as the score. Academic career advancement, including tenure, has become closely tied to citation counts. Hyper-optimization around these metrics leads researchers to focus more on what they can publish than on good science. Seemay Chou, biologist and co-founder of the Astera Institute, writes: “In all my discussions with scientists across every sector, exactly zero think the journal system works well. … Scientists should probably be putting out shorter narratives, datasets, code, and models at a faster rate, with more visibility into their thinking, mistakes, and methods.” She calls journal publishing “fundamentally broken … one of the legacy systems that prevents science from meeting its true potential for society.” In a rare move, Astera refuses to fund journal articles, requiring that research they support “be released and reviewed more openly, comprehensively, and frequently than the status quo”—and not behind a paywall or with a restrictive license. The affiliated lab Arcadia Science prohibits its researchers from publishing in journals. When they made this change, Chou reports, it profoundly reshaped their science: “Our researchers began designing experiments differently from the start. They became more creative and collaborative. The goal shifted from telling polished stories to uncovering useful truths.”⁶³

None of these models is a silver bullet. The point is that our ecosystem of research institutions has become dominated by a narrow set of organizational models. In “A Vision of Metascience,” Michael Nielsen and Kanjun Qiu argue that the ecosystem is in “a state of near stasis, with strong barriers inhibiting the improvement of key social processes.” Someone “seeking to achieve a scalable improvement in the social processes of science”—such as the founders of the organizations mentioned above—they call a “metascience entrepreneur.” They even envision a “metascience accelerator” that could provide support to such efforts, with the ultimate goal of “a flourishing ecosystem of people with wildly imaginative and insightful ideas for new social processes; and for those ideas to be tested and the best ideas scaled out.”⁶⁴ This vibrancy is what the ecosystem of science needs.

Culture

The deepest roots of institutional change are cultural. Cultural change is thus the most important pillar of the progress agenda.

This begins with what our children are taught in school. Today, they are not taught the history or nature of progress. History classes focus on wars and empires; science classes teach concepts and frameworks; the story of technology and economic growth falls between the cracks. Steven Johnson, popular author of dozens of books on the history of technology, reports that in “an otherwise excellent American history textbook” covering the last 150 years, “labor” was mentioned 226 times, and “civil rights” 134 times, but “antibiotics” and “vaccines” were not mentioned once. “Something is fundamentally distorted in the emphasis here.”⁶⁵

As a result, students graduate with no understanding of or appreciation for the system that created and maintains their historically unprecedented standard of living. Author Charles Mann has suggested: “All high-school students should be required to take a course called How the System Works”; he envisioned a course focusing on the systems that provide us with food, water, energy, and sanitation. “The overall theme: these systems, which required decades to create, are triumphs of the human mind and spirit, and it is the task of all of us to ensure that they are passed down to the next generation and improved. It should not be possible for so many terrific young men and women not to know anything about the systems on which their lives depend.”⁶⁶

A basic understanding of industrial civilization—how it works and why we need it—should be considered an essential outcome of an education. Call it “industrial literacy.”

Industrial literacy could start in grade school. Students could learn basic facts relevant to economic life: what crops need in order to grow, what farmers do, and how pests or disease can damage the harvest; what types of materials our world is made of, and how things like metal, glass, ceramic, and textiles are produced; different sources of energy, how to harness them, and how to direct forces using simple machines such as gears and levers. They could learn the stories of specific inventions and inventors: Edison and the light bulb, Stephenson and the locomotive, Bell and the telephone. They could learn basic facts about economic history: that people once lived without heating, air conditioning, plumbing, or electricity; that most people worked manual jobs, and mostly on farms; that severe disease was common, especially in childhood. They could engage in many hands-on activities to get first-hand knowledge of historical and modern processes: gardening, weaving, carpentry, paper-making and printing, navigation with a compass. They could try going a day eating only food they had grown themselves, wearing only clothes they had sewn, or using light only from candles they had dipped.

Children learn through books and other media as well. I’ve found a few relevant books for my young daughter, such as Up Goes the Skyscraper; John Deere, That’s Who!, about Deere’s steel plow; and Magic Ramen, about Momofuku Ando and his instant noodles—but I and other parents I know wish there were more.⁶⁷ There used to be. Steven Pinker tells us: “When I was a boy, a popular literary genre for children was the heroic biography of a medical pioneer such as Edward Jenner, Louis Pasteur, Joseph Lister, Frederick Banting, Charles Best, William Osler, or Alexander Fleming.”⁶⁸ (How many of those names do you recognize?) Another children’s book, from 1959, was titled You Will Go to the Moon; one blogger recounts: “My husband (aged 51), was profoundly changed by this story as a boy.”⁶⁹ Another series from the 1950s–’60s, “We Were There,” includes among its titles not only the Battle of Gettysburg and the Mayflower Pilgrims, but also the first airplane flight with the Wright Brothers, the completion of the transcontinental railroad, and the opening of the Erie Canal.⁷⁰ Topics like these have fallen out of favor, leaving parents mostly with worn and yellowed books from decades ago. We need a new generation of children’s media, to retell these stories for today’s youth and to update them with heroes like Katalin Karikó and Steve Jobs.

By high school, students would be ready to grasp the concept of progress itself: they could see the advancement of science, technology and industry as part of a bigger, integrated historical narrative. They could learn how we solved the grand challenges of human existence: how we learned to feed the world, how we mechanized most labor, how we reduced mortality from infectious disease. They would be ready to approach economic history quantitatively, understanding the significance of crop yields, labor productivity, or energy usage per capita; they would learn key historical charts such as the “hockey stick” of world GDP growth, the dramatic declines in poverty and child mortality, the growth of literacy and democracy.

In university, “progress studies” could be an interdisciplinary field integrating ideas from economics, history, and philosophy.⁷¹ Students would deepen their understanding of these fields, and engage with the centuries-long conversation about the meaning and causes of progress. They would read historical texts like Bacon and Condorcet, contemporary perspectives from authors such as Julian Simon or Joel Mokyr, and critics of progress from Rousseau to Lewis Mumford. Some universities are already starting to offer such classes, such as the University of Toronto, where professor Kevin Bryan is giving a course starting Fall 2025 on “Progress: How to Get Big Things Done.”⁷²

Students who graduated from such a curriculum would be cured of the “industrial amnesia” described in Chapter 1. They would not be “fish in water,” blind to technology and infrastructure all around them; they would see their world with wonder and gratitude. And rather than despair at inevitable doom for the world, they would have ambition to be a part of the story of human progress, to pursue our best opportunities and seek solutions to our biggest problems.

The great educator Maria Montessori saw progress as an essential part of the curriculum, for the purpose of moral development. Writing just after the World Wars, she felt acutely the need for human fraternity and unity—and she felt these were evident in the interconnectedness of society and the economy, and in the great discoveries and inventions made by those who came before us. Thus she wrote:

Children should be made to realise that all great achievements in culture and in the arts, all sciences and industries that have brought benefit to humanity, are due to the work of men who often struggled in obscurity and under conditions of great hardship; men driven by a profound passion, by an inner fire, to create with their research, with their work, new benefits not only for the people who lived in their times, but also for those of the future.⁷³

When this was done in classrooms, she reports:

The children frequently asked to see the portraits of these heroes and delighted in relentlessly pursuing the near-miraculous significance of their work once they had a clear idea of the times in which they had lived, the degree of ignorance of their contemporaries with regard to their research and studies, and the dearth of means at their disposal.

A near-religious respect grew within them for these men who lived in such distant times and places, who belonged to such a diversity of social classes; in this way, they managed to thoroughly grasp, almost concretely, the universal unity for good achieved by the work of men the world over.

In another essay, she wrote:

The child will have the greater pleasure in all subjects, and find them easier to learn, if he were led to realize how these subjects first came to be studied and who studied them. We write and read, and the child can be taught who invented writing and the instruments wherewith we write, how printing came and books became so numerous. Every achievement has come by the sacrifice of someone now dead. Every map speaks eloquently of the work of explorers and pioneers, who underwent hardships and trials to find new places, rivers and lakes, and to make the world greater and richer for our dwelling.⁷⁴

Art and entertainment, too, provide moral education. We should celebrate the work of scientists, inventors and founders in story and song. There should be far more major Hollywood biopics of these figures. Where is the life story of Norman Borlaug, and why isn’t he a household name? Why hasn’t there been a movie about Pasteur since 1936, in black and white? And when these stories are told, they too often focus on the “human drama”—Marie Curie’s love life, or Edison’s business feuds—that is easy to write but tells us nothing about their accomplishments.⁷⁵ These films should highlight instead the drama of discovery and invention. There is conflict, suspense, and adventure there—if only more screenwriters would take an interest in it and learn how to make it accessible and engaging on the screen.

Science fiction, too, can help us visualize a future that we want to live in and are inspired to build. J. Storrs Hall writes:

Science fiction has a long and valuable history of providing us with visions of a better world. Verne, Wells, Burroughs, Gernsback—even Bellamy—much less Campbell, Doc Smith, van Vogt, Heinlein, Asimov, Garrett, Piper, Niven, and Pournelle, provided people with places and lives they could imagine and aspire to create. Science fiction since the Sixties has signally failed in that regard; we have been fed, by and large, a diet of Chicken Little soup in a pot of message, ladled out over leg of Frankenstein.⁷⁶

An objection I often hear is: “You can’t have a story where everyone is happy and everything is perfect! Stories need conflict!” Of course they do—but there are many ways to write a compelling, exciting story without implying that technology makes the world worse or that its main feature is doom. You can tell a man vs. nature story, as in The Martian or Seveneves. You can tell a story of heroic builders vs. the villains who want to stop them, whether they are radical anti-human environmentalists like Ra’s al Ghul from DC Comics, or religious fanatics like the ones in Contact. You can tell a classic human story in a futuristic setting: a detective drama like The Quantum Thief or a political epic like The Moon is a Harsh Mistress. You can have technology as both doom and savior: robots fighting robots, or a genetically engineered cure for a lab-leaked pandemic. You can create conflict over good vs. evil uses of technology: maybe there is a biotechnology that could cure disease and aging, but the villains want to steal it for a bioweapon. You can explore the social implications of technology, from the ethics of embryo selection (Gattaca) to human relationships with AI (Her). The future is not a bland, flawless utopia—it’s a dynamic “protopia,” with plenty of room for conflict and intrigue, heroes and villains.⁷⁷

Music, too, can convey the grandeur of progress—such as the soundtracks for the Civilization game series by composer Christopher Tin. His album To Shiver the Sky glorifies flight and space travel, beginning with the words of da Vinci and ending with JFK’s call to the Moon, all set to choral and operatic music. His song “Live Gloriously,” written for Civilization VII, echoes “the game’s core narrative concept: that history is a global story of human achievement, and that the lessons of the past can still inspire our lives in the present.”⁷⁸

Culture is also mediated by journalism and media. Reporters today can’t even cover medical advances without including concerns from hand-wringing “experts.” A NYT story on Loyal, a startup making longevity drugs for dogs, quotes a bioethicist wondering whether “it is in their best interest to live a little bit longer when there’s some risk to taking these drugs”; the story points out: “The dogs themselves cannot give consent.”⁷⁹ “We have a media defined by cynicism,” says Arena Magazine in a manifesto titled “The New Needs Friends.” Arena defines itself as the opposite: a media company that is “on the side of the future and the people building it.”⁸⁰ We need more journalism like this, or Freethink, or Ashlee Vance’s Core Memory, that covers technology news straightforwardly—not with uncritical boosterism, but without problematizing every new development.

We need more opinion writers who appreciate progress and are focused on how we get more of it—such as Derek Thompson, Ezra Klein, Jerusalem Demsas, Matt Yglesias, or Noah Smith (Klein is at the NYT; the rest are independent, but were formerly at places like The Atlantic, Bloomberg, and Vox). We need more data journalists with this focus too, such as John Burn-Murdoch at the Financial Times. We need entire magazines devoted to progress, like the excellent Works in Progress magazine, or the biology-focused Asimov Press.

We need more popularizations of the history of progress, like the American Innovations podcast from Steven Johnson, or his popular books such as The Ghost Map or How We Got to Now. We need more social media personalities to lend their weight to progress, like Isabelle Boemeke, a fashion model with over 50,000 Instagram followers who uses her platform to promote the value of nuclear energy.⁸¹

The zeitgeist has a lot of inertia. But with slow, steady pressure from multiple directions—from the classroom to the box office to the social feed—it can be turned.

“Society’s course will be changed only by a change in ideas,” Hayek is reported to have said. “First you must reach the intellectuals, the teachers and writers, with reasoned argument. It will be their influence on society which will prevail, and the politicians will follow.” Sounding the same note, Keynes famously wrote that “the ideas of economists and political philosophers, both when they are right and when they are wrong, are more powerful than is commonly understood. Indeed the world is ruled by little else.” Though archrivals in economics and policy, these two men agreed: ideas are upstream of politics. The progress movement, then, is a movement of ideas.

Ideas drive the world from two directions: top-down, by directly influencing a small set of leaders in each field, and bottom-up, by indirectly influencing popular opinion. Only a small minority of the population reads intellectual books or essays, or pays attention to long-form blogs and podcasts—but among the few who do are those who run the world. Ideas in these formats can raise the ambitions of scientists, engineers and founders; can direct the resources of VCs and philanthropists; can inform the policy goals of think tanks and Congressmen. Reaching this minority has outsized influence. But ideas also directly influence those who write textbooks for schools, screenplays for Hollywood, and articles for newspapers and magazines—which in turn influence the public at large, even the vast majority who don’t read books. Public opinion creates headwinds or tailwinds for any project, and not only in politics; if those winds are strong enough, they can make a project impossible, or inevitable. And it is from the public that the next generation of leaders is recruited—among the young are our best chances to inspire ambitious scientists, inventors, founders, funders, and policy reformers. This is the bottom-up influence.

The foundation of any intellectual movement, then, is a body of ideas, expressed primarily in long-form writing—essays, reports, white papers, and especially books. There is already a small progress canon of books, including Pinker’s Enlightenment Now, Klein and Thompson’s Abundance, and Hall’s Where Is My Flying Car? But we need far more. We need books on history, telling the story of progress, as it has never adequately been told and desperately deserves to be. We need books on philosophy, like this one, honing our conception of progress: its nature, definition, and measurement; its benefits, costs, and risks; its root causes and future prospects. We need books about solutions to our biggest challenges: AI safety, media addiction, obesity, climate, technological unemployment. We need books that paint visions of the future, the prospects for longevity or nanotech or space or energy. Think of how many books you would find in any library on environmental studies, religious studies, or gender studies—we need at least as many on progress studies.

Sensing this need back in 2017, I started what was then a tiny blog called The Roots of Progress. In 2019, an article in The Atlantic by Patrick Collison and Tyler Cowen coined the term “progress studies,” which galvanized the formation of a progress community. A few years later, with the community still active and growing, it was clear that some movement-building was needed. Building on the growing audience for my blog, I founded the Roots of Progress Institute, with the mission to build a culture of progress for the 21st century. Our strategy is threefold: lay the intellectual foundation for the movement, in long-form writing; build community around these ideas; and then spread them to the public through education, media and entertainment. You can learn about our programs, including our annual conference and our fellowship for progress writers, at rootsofprogress.org.

In this movement we are joined by many other organizations. There are new think tanks inspired by ideas of progress, including the Institute for Progress, the Abundance Institute, the Good Science Project, and the Inclusive Abundance Initiative. There are also older organizations that have joined the cause, such as the Foresight Institute, the Human Progress project at Cato, the Breakthrough Institute, the Foundation for American Innovation, and the Institute for Humane Studies at George Mason University. Philanthropic foundations have started special funds, such as the $120M Abundance and Growth Fund launched by Open Philanthropy, or the science and technology initiatives of Renaissance Philanthropy.⁸²

There is a role in the progress movement for every individual. If you are a scientist, engineer, or founder: you are already on the front lines of progress! Be ambitious in your work, and take courage and inspiration from the techno-humanist worldview. If you are a VC or philanthropist: make the progress agenda a funding priority. If you are in the humanities, especially history, economics, or philosophy: study progress and incorporate it into your thinking—these fields could benefit from more often adopting the progress lens. If you are an author, journalist, or teacher: communicate ideas of progress, and use this perspective to choose topics and angles on them. If you are a novelist, screenwriter, director, or other storyteller or artist: tell stories of progress, and aim to inspire humanity’s best efforts. If you work in law, regulation, or policy: think about how to remove roadblocks, and about the wisest ways to regulate emerging technologies.

If you are a voter: choose politicians who will support building, economic growth, technology, and science, and weigh in with your representatives when these issues come up. If you are a parent: educate your children about progress, don’t wait for the schools to do it. Tell them stories of discovery and invention, teach them to view industrial civilization with wonder, teach them to have gratitude for the brave souls who went before us and created this amazing world. And no matter who you are: educate yourself, spread the word, and support the efforts that make up this movement any way you can.

To turn the culture around, after the fear and skepticism that marked the end of the 20th century, is a herculean feat. It will require a large-scale cultural movement, on the scale of the environmentalist movement, the neoliberal movement, or the civil rights movement—and like them, it will take a generation to come to fruition.

Such a task may seem daunting. But Francis Bacon inspired generations after him with a vision of useful knowledge leading to practical improvements in human life—and he had only a handful of inventions to point to as examples to prove his case. Today, the case is far stronger: progress is not a theoretical possibility for the future, but the established reality of the past and the living present; its examples are literally all around us. And to communicate our vision, we have not only the printing press, but the internet. If Bacon did it, so can we.

The bold, ambitious future awaits. Let’s go build it.

For more about The Techno-Humanist Manifesto, including the table of contents, see the announcement. For full citations, see the bibliography.