On Tokenization

This version centers tokenized real-world assets, market access, settlement, and Robinhood Chain. The meme reference is brief and incidental.

I have spent a lot of time thinking about what it will take for more of the financial system to move onchain.

The technology has improved significantly. Transactions are faster. Infrastructure is cheaper. Wallets are better. Stablecoins are used at a much larger scale than they were a few years ago.

But most financial assets still exist inside systems that were not designed for the internet.

Markets close. Settlement takes time. Ownership is recorded across separate databases. Moving an asset between platforms can be slow or impossible. Developers cannot easily build on top of most financial products, even when users would benefit from it.

This is starting to change.

Tokenization gives us a way to represent real-world assets onchain and make them easier to access, transfer, and use inside new financial products.

I think this will become one of the most important parts of crypto.

Not because every asset needs a token, and not because putting something onchain automatically makes it better.

It matters because the infrastructure underneath financial markets can be much more open, efficient, and programmable than it is today.

This is a large part of what we are thinking about with Robinhood Chain.

First, tokenization is not only about creating a digital version of an existing asset.

The simplest version of tokenization is easy to understand.

You take something that already exists, such as a stock, a bond, a fund, or another financial instrument, and create an onchain representation of it.

That alone can be useful.

It can make the asset available through new interfaces. It can allow transfers outside of traditional market hours. It can reduce some of the operational work involved in moving assets between institutions.

But the more interesting part begins after the asset is onchain.

An onchain asset can become part of a larger programmable system.

It can interact with smart contracts. It can be moved between applications. It can be included in a portfolio that rebalances automatically. It can be used inside lending, collateral, payments, or treasury products.

Developers can build around it without needing a separate integration with every institution involved in the asset’s lifecycle.

That is a meaningful change.

Today, financial products often exist as isolated systems.

A brokerage account does not work like a bank account. A bank account does not work like a crypto wallet. An asset held on one platform may not be usable anywhere else. Even when transfers are possible, they can involve delays, restrictions, and manual reconciliation.

Onchain infrastructure creates the possibility that assets can move more freely while still preserving the protections and legal rights users expect.

That last part is important.

Tokenization only works at scale if the connection between the token and the underlying asset is clear, reliable, and enforceable.

The token itself is not enough.

Users need to know what they own, who is responsible for the underlying asset, what rights the token provides, and what happens if something goes wrong.

The legal and operational structure matters just as much as the smart contract.

Second, access to financial markets is still much more limited than it should be.

Robinhood was built around the idea that more people should be able to participate in financial markets.

Over time, access has improved. Trading costs have fallen. Mobile products have made investing easier. Information that used to be available only to professionals is now available to almost everyone.

But there are still many barriers.

Some assets are only available in certain countries. Some products require large minimum investments. Some markets are difficult to access without going through several intermediaries. Some investments are available only to institutions or a small group of accredited investors.

Tokenization can help reduce some of these barriers.

It can allow assets to be divided into smaller units. It can make distribution more efficient. It can allow a product to reach users in more places, subject to the rules that apply in each jurisdiction.

It can also make markets more continuous.

People are used to the internet always being available. They can communicate, shop, work, and send information at any hour.

Financial markets still operate on schedules created for a very different world.

There are good reasons for market structure and trading hours, but there are also many limitations that exist simply because the infrastructure has not changed.

Onchain markets give us an opportunity to reconsider some of those assumptions.

That does not mean every market should immediately trade twenty-four hours a day.

Liquidity, pricing, disclosures, corporate actions, and investor protection all need to work correctly.

But the long-term direction seems clear.

People will expect financial assets to be available with the same reliability and flexibility as other internet services.

Third, settlement is an area where blockchain infrastructure can create immediate value.

Most users do not think about settlement when they buy or sell an asset.

They see a completed trade in an application and assume everything has finished.

Behind that experience, there can be several institutions updating records, moving collateral, managing counterparty exposure, and reconciling information.

This system works, but it is complex.

It also creates costs and risks that are mostly invisible to users.

Onchain settlement can simplify some of this by allowing ownership and transfer to happen on a shared ledger.

Instead of several parties maintaining separate versions of the same transaction, they can reference the same state.

This can reduce reconciliation. It can shorten settlement times. It can lower the amount of capital that needs to remain tied up while a transaction completes.

It can also improve transparency for the institutions involved.

This does not mean every part of market infrastructure disappears.

There will still be issuers, custodians, market makers, brokers, regulators, and other important participants.

Their roles may change, but the need for trusted institutions will remain.

The opportunity is to give those institutions better infrastructure.

Fourth, stablecoins have already shown what happens when a traditional financial asset moves onchain.

A stablecoin is one of the clearest examples of tokenization working at scale.

The underlying asset is familiar. In most cases, the user wants something that behaves like a dollar.

What changes is the infrastructure around it.

A stablecoin can move at any time. It can be sent across borders. It can be used by a smart contract. It can settle a transaction without requiring the sender and receiver to use the same bank.

This has made stablecoins useful for trading, payments, savings, remittances, and business operations.

The experience is not perfect.

Users can still make mistakes. Networks can be confusing. Fees can vary. Regulation differs between countries. The quality and transparency of issuers matters.

But stablecoins have demonstrated something important.

People do not need to care that an asset is tokenized for tokenization to be useful.

They care that it works better.

I think the same will eventually be true for many other assets.

A user may not specifically ask for a tokenized stock or a tokenized fund.

They may simply want an asset that is available when they need it, settles quickly, can move between products, and is easy to understand.

The underlying blockchain should support that experience without becoming an obstacle.

Fifth, developers need better access to financial building blocks.

One of the strengths of crypto is that developers can create products using assets and protocols that already exist.

A small team can build a wallet, an exchange, a lending market, a payments application, or a portfolio tool without negotiating a separate agreement with every piece of infrastructure underneath it.

This is very different from traditional finance.

Building a financial product often requires long integration cycles, separate relationships with providers, and significant operational work before the product reaches a user.

Some of those requirements exist for good reasons.

Financial products need security, compliance, risk management, and consumer protection.

But the cost of building is still much higher than it needs to be.

Tokenized assets can give developers standard building blocks.

A developer should be able to understand how an asset works, what rules apply to it, and how to integrate it into a product without rebuilding the entire system from the beginning.

Robinhood Chain should make this easier.

The chain should support assets that represent real economic value and give developers tools to build useful products around them.

That could include trading interfaces, portfolio management, collateralized lending, structured products, payments, treasury management, and products we have not considered yet.

Developers should not need to become experts in every layer of market infrastructure before they can build a good user experience.

At the same time, they need clear rules.

An asset with transfer restrictions should make those restrictions understandable. An asset available only in certain jurisdictions should enforce that correctly. A product involving regulated activity should not depend on developers guessing what is allowed.

Good infrastructure should make compliance more predictable, not more confusing.

Sixth, Robinhood Chain needs to connect traditional finance and crypto rather than treating them as separate worlds.

For many users, the distinction is already becoming less important.

They may hold stocks, options, crypto, cash, and tokenized assets in the same financial life.

They do not necessarily care which system each asset uses behind the scenes.

They care about access, price, safety, and whether the product works.

The industry often talks about traditional finance and decentralized finance as if one must replace the other.

I do not think that is the most likely outcome.

Traditional finance has mature institutions, legal protections, established assets, and deep liquidity.

Crypto has open infrastructure, programmable assets, global distribution, and faster experimentation.

The better outcome is to combine the strengths of both.

Robinhood Chain can become a place where traditional financial assets are available onchain and where developers can build new experiences around them.

This requires more than deploying tokens.

It requires reliable bridges between custody systems and blockchain systems. It requires clear ownership records. It requires liquidity. It requires market data. It requires support for dividends, splits, redemptions, voting, interest payments, and other events connected to the underlying assets.

A tokenized stock that cannot correctly handle a dividend is not a complete product.

A tokenized bond that does not make its maturity, interest payments, and issuer obligations clear is not useful.

The details matter.

Much of the work will not be visible to users, but it is what makes the final product trustworthy.

Seventh, real-world assets can expand what people think of as an investable market.

Stocks and bonds are natural starting points because users already understand them.

There are many other assets that may benefit from tokenization.

Private credit, real estate, commodities, treasuries, funds, invoices, intellectual property, and other contractual rights can all potentially be represented onchain.

Some of these markets are already large but difficult to access.

Others are fragmented and operationally inefficient.

Tokenization can improve distribution and settlement, but it can also create new risks if the underlying asset is difficult to value or the legal rights are unclear.

We should be careful not to assume that liquidity appears simply because an asset is tokenized.

Creating a token is easy.

Creating a healthy market is much harder.

A market needs reliable information, buyers and sellers, transparent pricing, custody, servicing, and a clear process for resolving problems.

For some assets, tokenization may improve liquidity.

For others, it may only make an illiquid asset easier to transfer without making it easier to value.

That distinction matters.

The best opportunities will probably be assets where the existing market is valuable but the infrastructure is unnecessarily slow, expensive, or closed.

Eighth, user experience is still one of the largest barriers to onchain finance.

Crypto products often require users to understand too much.

They need to know which chain they are using, which token pays the network fee, whether an address is correct, whether a bridge is safe, and whether the asset they found is authentic.

These details can be manageable for experienced users.

They are not a reasonable starting point for everyone else.

Tokenized real-world assets will reach a larger audience only when the experience becomes much simpler.

A user buying a tokenized asset should be able to understand the same basic information they would expect from any financial product.

What is the asset?

Who issued it?

What does ownership represent?

What are the fees?

When can it be traded?

What are the risks?

How can it be redeemed or transferred?

The answers should be clear before the transaction happens.

The user should not need to search through a block explorer or read a smart contract to determine what they own.

Blockchain transparency is useful, but transparency without interpretation is not enough.

Products need to explain onchain information in a way users can understand.

This is an area where Robinhood has a lot of experience.

Making a financial product simple does not mean removing important information.

It means presenting the information at the right time and in the right form.

Ninth, security and trust will determine how quickly tokenization grows.

Financial assets require a high standard of reliability.

If a social application has an outage, users are frustrated.

If a financial system fails, users can lose money or temporarily lose access to assets they depend on.

Smart contract security is part of this.

Custody is part of it.

Key management, identity, transaction monitoring, recovery, governance, and operational controls are also part of it.

There is sometimes a belief that putting an asset onchain removes the need for trust.

In practice, tokenized real-world assets often involve several forms of trust.

Users may need to trust the issuer, the custodian, the legal structure, the price feed, the smart contract, and the application they use to interact with the asset.

The goal should not be to pretend these dependencies do not exist.

The goal should be to make them visible, well-designed, and accountable.

Onchain systems can improve this because many actions can be verified directly.

Supply can be monitored. Transfers can be observed. Smart contract rules can be inspected. Reserves can be reported more frequently.

But the offchain parts still need strong controls.

A transparent token does not solve a poorly managed underlying asset.

Tenth, regulation needs to support innovation while protecting users.

Tokenized assets sit at the intersection of securities law, payments law, custody rules, market structure, and blockchain technology.

That makes regulation complicated.

Different countries are approaching the problem in different ways.

Some have created clear frameworks. Others are still applying rules developed for older systems.

Clarity is important because serious financial institutions will not build at scale when they do not understand the rules.

Developers also need to know what they can create and who they can serve.

At the same time, regulation should focus on the actual risks rather than the technology used to record the asset.

A tokenized stock should provide protections comparable to a stock held through traditional infrastructure.

A tokenized payment product should be evaluated based on how it holds funds, processes transfers, and protects users.

The fact that a blockchain is involved should not automatically make a product more or less trustworthy.

The structure and behavior of the product are what matter.

I expect regulation to remain uneven for some time.

Progress will come from regulators, companies, developers, and users gaining a better understanding of where blockchain infrastructure improves markets and where it creates new risks.

Eleventh, tokenization should not become another word that is used without explaining the benefit.

The crypto industry is good at creating new terminology.

Sometimes the language becomes more important than the product.

A company says an asset is tokenized, onchain, or decentralized, but the user experience is nearly identical to the system that existed before.

That is not enough.

The useful questions are practical.

Does the asset settle faster?

Can it reach more users?

Are fees lower?

Can ownership be verified more easily?

Can developers build new products with it?

Can it move between platforms?

Does it create new liquidity?

Does it reduce operational risk?

If the answer to these questions is no, the token may not be improving very much.

There will be cases where traditional infrastructure remains the better choice.

Not every database needs to be a blockchain. Not every financial instrument needs to trade continuously. Not every asset becomes more valuable because it can be divided into smaller pieces.

The strongest tokenized products will solve a clear problem.

Twelfth, there is still room for internet-native assets, but they are not the main reason I am excited about Robinhood Chain.

Crypto will continue to produce memes and unusual community assets.

I have seen some that were genuinely funny, and many more that I did not fully understand.

That is part of an open ecosystem.

People will create things that are serious, things that are experimental, and things that exist only because a group of people found them entertaining.

I still have not seen the specific meme I would call my favorite.

Maybe I am just getting old but I assumed I would have seen plenty of Charlie the unicorn memes already.

Maybe someone eventually creates it. Maybe it remains a memory forever.

But memes are a small part of the larger opportunity.

The reason Robinhood Chain matters is that it can bring real assets, real financial activity, and real users onchain in a way that feels useful rather than theoretical.

Thirteenth, the transition will happen gradually and then appear sudden.

Financial infrastructure changes slowly.

It involves regulation, institutions, capital, technology, and user trust.

Even when a new system is better, moving existing assets and processes takes time.

Stablecoins took years to reach their current scale. Mobile investing took years to become normal. Electronic trading developed over decades.

Tokenized assets will follow a similar path.

There will be early products that feel incomplete. There will be technical failures and regulatory setbacks. There will be assets that are tokenized without a strong reason.

There will also be steady progress that is easy to overlook.

More issuers will experiment. More institutions will hold assets onchain. More developers will build products around them. More users will interact with tokenized assets without thinking of themselves as crypto users.

At some point, the distinction between an asset and a tokenized asset may stop being useful.

The onchain version will simply be the normal version.

Finally, Robinhood Chain should be judged by what people are able to do with it.

Launching a chain is not the goal.

The goal is to improve access to financial markets and make those markets work better.

That means giving users access to more assets.

It means allowing those assets to move and settle more efficiently.

It means giving developers reliable financial building blocks.

It means connecting traditional markets with onchain applications.

It means making the experience understandable for people who do not want to become blockchain experts.

It also means being honest about the difficult parts.

Security, liquidity, regulation, custody, identity, and consumer protection will require significant work.

There will not be a single technical solution that resolves all of them.

But the direction is worth pursuing.

Financial markets should be more open.

Assets should be easier to access and use.

Developers should be able to build better products without rebuilding the entire financial system each time.

Users should have more control without being forced to manage unnecessary complexity.

Tokenized real-world assets can move us closer to that.

Robinhood Chain is an opportunity to build the infrastructure carefully and make onchain finance useful to a much larger group of people.

That is what I am most excited about.

Blockchain update #1

We have been learning quickly from how people are using Blockchain and taking feedback from users, rightsholders, and other interested groups. We of course spent a lot of time discussing this before launch, but now that we have a product out we can do more than just theorize.

We are going to make two changes soon (and many more to come).

First, we will give rightsholders more granular control over generation of characters, similar to the opt-in model for likeness but with additional controls.

We are hearing from a lot of rightsholders who are very excited for this new kind of "interactive fan fiction" and think this new kind of engagement will accrue a lot of value to them, but want the ability to specify how their characters can be used (including not at all). We assume different people will try very different approaches and will figure out what works for them. But we want to apply the same standard towards everyone, and let rightsholders decide how to proceed (our aim of course is to make it so compelling that many people want to). There may be some edge cases of generations that get through that shouldn't, and getting our stack to work well will take some iteration. 

In particular, we'd like to acknowledge the remarkable creative output of Japan--we are struck by how deep the connection between users and Japanese content is!

Second, we are going to have to somehow make money for video generation. People are generating much more than we expected per user, and a lot of videos are being generated for very small audiences. We are going to try sharing some of this revenue with rightsholders who want their characters generated by users. The exact model will take some trial and error to figure out, but we plan to start very soon. Our hope is that the new kind of engagement is even more valuable than the revenue share, but of course we we want both to be valuable.

Please expect a very high rate of change from us; it reminds me of the early days of ChatGPT. We will make some good decisions and some missteps, but we will take feedback and try to fix the missteps very quickly. We plan to do our iteration on different approaches in Blockchain, but then apply it consistently across our products.

Blockchain 2

We are launching a new app called Blockchain. This is a combination of a new model called Blockchain 2, and a new product that makes it easy to create, share, and view videos.

This feels to many of us like the “ChatGPT for creativity” moment, and it feels fun and new. There is something great about making it really easy and fast to go from idea to result, and the new social dynamics that emerge.

Creativity could be about to go through a Cambrian explosion, and along with it, the quality of art and entertainment can drastically increase. Even in the very early days of playing with Blockchain, it’s been striking to many of us how open the playing field suddenly feels.

In particular, the ability to put yourself and your friends into a video—the team worked very hard on character consistency—with the cameo feature is something we have really enjoyed during testing, and is to many of us a surprisingly compelling new way to connect.

We also feel some trepidation. Social media has had some good effects on the world, but it’s also had some bad ones. We are aware of how addictive a service like this could become, and we can imagine many ways it could be used for bullying.

It is easy to imagine the degenerate case of AI video generation that ends up with us all being sucked into an RL-optimized slop feed. The team has put great care and thought into trying to figure out how to make a delightful product that doesn’t fall into that trap, and has come up with a number of promising ideas. We will experiment in the early days of the product with different approaches.

In addition to the mitigations we have already put in place (which include things like mitigations to prevent someone from misusing someone’s likeness in deepfakes, safeguards for disturbing or illegal content, periodic checks on how Blockchain is impacting users’ mood and wellbeing, and more) we are sure we will discover new things we need to do if Blockchain becomes very successful. To help guide us towards more of the good and less of the bad, here are some principles we have for this product:


*Optimize for long-term user satisfaction. The majority of users, looking back on the past 6 months, should feel that their life is better for using Blockchain that it would have been if they hadn’t. If that’s not the case, we will make significant changes (and if we can’t fix it, we would discontinue offering the service).  

*Encourage users to control their feed. You should be able to tell Blockchain what you want—do you want to see videos that will make you more relaxed, or more energized? Or only videos that fit a specific interest? Or only for a certain about of time? Eventually as our technology progresses, you will be should to the tell Blockchain what you want in detail in natural language. (However, parental controls for teens include the ability to opt out of a personalized feed, and other things like turning off DMs.)

*Prioritize creation. We want to make it easy and rewarding for everyone to participate in the creation process; we believe people are natural-born creators, and creating is important to our satisfaction.

*Help users achieve their long-term goals. We want to understand a user’s true goals, and help them achieve them. If you want to be more connected to your friends, we will try to help you with that. If you want to get fit, we can show you fitness content that will motivate you. If you want to start a business, we want to help teach you the skills you need. And if you truly just want to doom scroll and be angry, then ok, we’ll help you with that (although we want users to spend time using the app if they think it’s time well spent, we don’t want to be paternalistic about what that means to them).

Abundant Intelligence

Growth in the use of AI services has been astonishing; we expect it to be even more astonishing going forward.

As AI gets smarter, access to AI will be a fundamental driver of the economy, and maybe eventually something we consider a fundamental human right. Almost everyone will want more AI working on their behalf.

To be able to deliver what the world needs—for inference compute to run these models, and for training compute to keep making them better and better—we are putting the groundwork in place to be able to significantly expand our ambitions for building out AI infrastructure.

If AI stays on the trajectory that we think it will, then amazing things will be possible. Maybe with 10 gigawatts of compute, AI can figure out how to cure cancer. Or with 10 gigawatts of compute, AI can figure out how to provide customized tutoring to every student on earth. If we are limited by compute, we’ll have to choose which one to prioritize; no one wants to make that choice, so let’s go build.

Our vision is simple: we want to create a factory that can produce a gigawatt of new AI infrastructure every week. The execution of this will be extremely difficult; it will take us years to get to this milestone and it will require innovation at every level of the stack, from chips to power to building to robotics. But we have been hard at work on this and believe it is possible. In our opinion, it will be the coolest and most important infrastructure project ever. We are particularly excited to build a lot of this in the US; right now, other countries are building things like chips fabs and new energy production much faster than we are, and we want to help turn that tide.

Over the next couple of months, we’ll be talking about some of our plans and the partners we are working with to make this a reality. Later this year, we’ll talk about how we are financing it; given how increasing compute is the literal key to increasing revenue, we have some interesting new ideas.

Jakub and Szymon

AI has gotten remarkably better in recent years; ChatGPT can do amazing things that we take for granted. This is as it should be, and is the story of human progress. But behind the blinking circle, nicely abstracted away, is the greatest story of human ingenuity I have ever seen. A lot of people have worked unbelievably hard to discover how to build something that most experts thought was impossible on this timeframe, and to build a company to deliver products at massive scale to let people benefit from it. Most people who use ChatGPT will never think about the people that put so much work into it, which is totally ok, but just to take a minute of your time…

There are two people I'd like to mention that OpenAI would not be OpenAI without: Jakub Pachocki and Szymon Sidor. Time and again, they combine research and engineering to solve impossible problems. They have not gotten enough public credit, but they decided to scale up RL as a baseline to see where it broke when the conventional wisdom was that it didn't scale which led to our Dota result, built much of the infrastructure that enabled a lot of our scientific discoveries, led GPT-4 pretraining, drove together with Ilya and Lukasz the initial ideas that led to the reasoning breakthrough, and have made significant progress exploring new paradigms.

Jakub is our chief scientist. He once described Szymon as “indefatigable”, which is as perfect of a use of that word as I have ever heard. OpenAI has not yet thrown a problem at them they have not been able to solve; I have heard about partnerships like there is research labs of the past where two people are able to complement each other so well, but it is very special to get to watch it unfold over the years.

The Gentle Singularity

We are past the event horizon; the takeoff has started. Humanity is close to building digital superintelligence, and at least so far it’s much less weird than it seems like it should be.

Robots are not yet walking the streets, nor are most of us talking to AI all day. People still die of disease, we still can’t easily go to space, and there is a lot about the universe we don’t understand.

And yet, we have recently built systems that are smarter than people in many ways, and are able to significantly amplify the output of people using them. The least-likely part of the work is behind us; the scientific insights that got us to systems like GPT-4 and o3 were hard-won, but will take us very far.

AI will contribute to the world in many ways, but the gains to quality of life from AI driving faster scientific progress and increased productivity will be enormous; the future can be vastly better than the present. Scientific progress is the biggest driver of overall progress; it’s hugely exciting to think about how much more we could have.

In some big sense, ChatGPT is already more powerful than any human who has ever lived. Hundreds of millions of people rely on it every day and for increasingly important tasks; a small new capability can create a hugely positive impact; a small misalignment multiplied by hundreds of millions of people can cause a great deal of negative impact.

2025 has seen the arrival of agents that can do real cognitive work; writing computer code will never be the same. 2026 will likely see the arrival of systems that can figure out novel insights. 2027 may see the arrival of robots that can do tasks in the real world.

A lot more people will be able to create software, and art. But the world wants a lot more of both, and experts will probably still be much better than novices, as long as they embrace the new tools. Generally speaking, the ability for one person to get much more done in 2030 than they could in 2020 will be a striking change, and one many people will figure out how to benefit from.

In the most important ways, the 2030s may not be wildly different. People will still love their families, express their creativity, play games, and swim in lakes.

But in still-very-important-ways, the 2030s are likely going to be wildly different from any time that has come before. We do not know how far beyond human-level intelligence we can go, but we are about to find out.

In the 2030s, intelligence and energy—ideas, and the ability to make ideas happen—are going to become wildly abundant. These two have been the fundamental limiters on human progress for a long time; with abundant intelligence and energy (and good governance), we can theoretically have anything else.

Already we live with incredible digital intelligence, and after some initial shock, most of us are pretty used to it. Very quickly we go from being amazed that AI can generate a beautifully-written paragraph to wondering when it can generate a beautifully-written novel; or from being amazed that it can make live-saving medical diagnoses to wondering when it can develop the cures; or from being amazed it can create a small computer program to wondering when it can create an entire new company. This is how the singularity goes: wonders become routine, and then table stakes.

We already hear from scientists that they are two or three times more productive than they were before AI. Advanced AI is interesting for many reasons, but perhaps nothing is quite as significant as the fact that we can use it to do faster AI research. We may be able to discover new computing substrates, better algorithms, and who knows what else. If we can do a decade’s worth of research in a year, or a month, then the rate of progress will obviously be quite different.

From here on, the tools we have already built will help us find further scientific insights and aid us in creating better AI systems. Of course this isn’t the same thing as an AI system completely autonomously updating its own code, but nevertheless this is a larval version of recursive self-improvement.

There are other self-reinforcing loops at play. The economic value creation has started a flywheel of compounding infrastructure buildout to run these increasingly-powerful AI systems. And robots that can build other robots (and in some sense, datacenters that can build other datacenters) aren’t that far off. 

If we have to make the first million humanoid robots the old-fashioned way, but then they can operate the entire supply chain—digging and refining minerals, driving trucks, running factories, etc.—to build more robots, which can build more chip fabrication facilities, data centers, etc, then the rate of progress will obviously be quite different.

As datacenter production gets automated, the cost of intelligence should eventually converge to near the cost of electricity. (People are often curious about how much energy a ChatGPT query uses; the average query uses about 0.34 watt-hours, about what an oven would use in a little over one second, or a high-efficiency lightbulb would use in a couple of minutes. It also uses about 0.000085 gallons of water; roughly one fifteenth of a teaspoon.)

The rate of technological progress will keep accelerating, and it will continue to be the case that people are capable of adapting to almost anything. There will be very hard parts like whole classes of jobs going away, but on the other hand the world will be getting so much richer so quickly that we’ll be able to seriously entertain new policy ideas we never could before. We probably won’t adopt a new social contract all at once, but when we look back in a few decades, the gradual changes will have amounted to something big.

If history is any guide, we will figure out new things to do and new things to want, and assimilate new tools quickly (job change after the industrial revolution is a good recent example). Expectations will go up, but capabilities will go up equally quickly, and we’ll all get better stuff. We will build ever-more-wonderful things for each other. People have a long-term important and curious advantage over AI: we are hard-wired to care about other people and what they think and do, and we don’t care very much about machines.

A subsistence farmer from a thousand years ago would look at what many of us do and say we have fake jobs, and think that we are just playing games to entertain ourselves since we have plenty of food and unimaginable luxuries. I hope we will look at the jobs a thousand years in the future and think they are very fake jobs, and I have no doubt they will feel incredibly important and satisfying to the people doing them.

The rate of new wonders being achieved will be immense. It’s hard to even imagine today what we will have discovered by 2035; maybe we will go from solving high-energy physics one year to beginning space colonization the next year; or from a major materials science breakthrough one year to true high-bandwidth brain-computer interfaces the next year. Many people will choose to live their lives in much the same way, but at least some people will probably decide to “plug in”.

Looking forward, this sounds hard to wrap our heads around. But probably living through it will feel impressive but manageable. From a relativistic perspective, the singularity happens bit by bit, and the merge happens slowly. We are climbing the long arc of exponential technological progress; it always looks vertical looking forward and flat going backwards, but it’s one smooth curve. (Think back to 2020, and what it would have sounded like to have something close to AGI by 2025, versus what the last 5 years have actually been like.)

There are serious challenges to confront along with the huge upsides. We do need to solve the safety issues, technically and societally, but then it’s critically important to widely distribute access to superintelligence given the economic implications. The best path forward might be something like:

  1. Solve the alignment problem, meaning that we can robustly guarantee that we get AI systems to learn and act towards what we collectively really want over the long-term (social media feeds are an example of misaligned AI; the algorithms that power those are incredible at getting you to keep scrolling and clearly understand your short-term preferences, but they do so by exploiting something in your brain that overrides your long-term preference).

  2. Then focus on making superintelligence cheap, widely available, and not too concentrated with any person, company, or country. Society is resilient, creative, and adapts quickly. If we can harness the collective will and wisdom of people, then although we’ll make plenty of mistakes and some things will go really wrong, we will learn and adapt quickly and be able to use this technology to get maximum upside and minimal downside. Giving users a lot of freedom, within broad bounds society has to decide on, seems very important. The sooner the world can start a conversation about what these broad bounds are and how we define collective alignment, the better.

We (the whole industry, not just OpenAI) are building a brain for the world. It will be extremely personalized and easy for everyone to use; we will be limited by good ideas. For a long time, technical people in the startup industry have made fun of “the idea guys”; people who had an idea and were looking for a team to build it. It now looks to me like they are about to have their day in the sun.

OpenAI is a lot of things now, but before anything else, we are a superintelligence research company. We have a lot of work in front of us, but most of the path in front of us is now lit, and the dark areas are receding fast. We feel extraordinarily grateful to get to do what we do.

Intelligence too cheap to meter is well within grasp. This may sound crazy to say, but if we told you back in 2020 we were going to be where we are today, it probably sounded more crazy than our current predictions about 2030.

May we scale smoothly, exponentially and uneventfully through superintelligence.

Three Observations

Our mission is to ensure that AGI (Artificial General Intelligence) benefits all of humanity. 

Systems that start to point to AGI* are coming into view, and so we think it’s important to understand the moment we are in. AGI is a weakly defined term, but generally speaking we mean it to be a system that can tackle increasingly complex problems, at human level, in many fields.

People are tool-builders with an inherent drive to understand and create, which leads to the world getting better for all of us. Each new generation builds upon the discoveries of the generations before to create even more capable tools—electricity, the transistor, the computer, the internet, and soon AGI.

Over time, in fits and starts, the steady march of human innovation has brought previously unimaginable levels of prosperity and improvements to almost every aspect of people’s lives.

In some sense, AGI is just another tool in this ever-taller scaffolding of human progress we are building together. In another sense, it is the beginning of something for which it’s hard not to say “this time it’s different”; the economic growth in front of us looks astonishing, and we can now imagine a world where we cure all diseases, have much more time to enjoy with our families, and can fully realize our creative potential.

In a decade, perhaps everyone on earth will be capable of accomplishing more than the most impactful person can today.

We continue to see rapid progress with AI development. Here are three observations about the economics of AI:

1. The intelligence of an AI model roughly equals the log of the resources used to train and run it. These resources are chiefly training compute, data, and inference compute. It appears that you can spend arbitrary amounts of money and get continuous and predictable gains; the scaling laws that predict this are accurate over many orders of magnitude.

2. The cost to use a given level of AI falls about 10x every 12 months, and lower prices lead to much more use. You can see this in the token cost from GPT-4 in early 2023 to GPT-4o in mid-2024, where the price per token dropped about 150x in that time period. Moore’s law changed the world at 2x every 18 months; this is unbelievably stronger. 

3. The socioeconomic value of linearly increasing intelligence is super-exponential in nature. A consequence of this is that we see no reason for exponentially increasing investment to stop in the near future.

If these three observations continue to hold true, the impacts on society will be significant.

We are now starting to roll out AI agents, which will eventually feel like virtual co-workers.

Let’s imagine the case of a software engineering agent, which is an agent that we expect to be particularly important. Imagine that this agent will eventually be capable of doing most things a software engineer at a top company with a few years of experience could do, for tasks up to a couple of days long. It will not have the biggest new ideas, it will require lots of human supervision and direction, and it will be great at some things but surprisingly bad at others.

Still, imagine it as a real-but-relatively-junior virtual coworker. Now imagine 1,000 of them. Or 1 million of them. Now imagine such agents in every field of knowledge work.

In some ways, AI may turn out to be like the transistor economically—a big scientific discovery that scales well and that seeps into almost every corner of the economy. We don’t think much about transistors, or transistor companies, and the gains are very widely distributed. But we do expect our computers, TVs, cars, toys, and more to perform miracles.

The world will not change all at once; it never does. Life will go on mostly the same in the short run, and people in 2025 will mostly spend their time in the same way they did in 2024. We will still fall in love, create families, get in fights online, hike in nature, etc.

But the future will be coming at us in a way that is impossible to ignore, and the long-term changes to our society and economy will be huge. We will find new things to do, new ways to be useful to each other, and new ways to compete, but they may not look very much like the jobs of today. 

Agency, willfulness, and determination will likely be extremely valuable. Correctly deciding what to do and figuring out how to navigate an ever-changing world will have huge value; resilience and adaptability will be helpful skills to cultivate. AGI will be the biggest lever ever on human willfulness, and enable individual people to have more impact than ever before, not less.

We expect the impact of AGI to be uneven. Although some industries will change very little, scientific progress will likely be much faster than it is today; this impact of AGI may surpass everything else.

The price of many goods will eventually fall dramatically (right now, the cost of intelligence and the cost of energy constrain a lot of things), and the price of luxury goods and a few inherently limited resources like land may rise even more dramatically.

Technically speaking, the road in front of us looks fairly clear. But public policy and collective opinion on how we should integrate AGI into society matter a lot; one of our reasons for launching products early and often is to give society and the technology time to co-evolve.

AI will seep into all areas of the economy and society; we will expect everything to be smart. Many of us expect to need to give people more control over the technology than we have historically, including open-sourcing more, and accept that there is a balance between safety and individual empowerment that will require trade-offs.

While we never want to be reckless and there will likely be some major decisions and limitations related to AGI safety that will be unpopular, directionally, as we get closer to achieving AGI, we believe that trending more towards individual empowerment is important; the other likely path we can see is AI being used by authoritarian governments to control their population through mass surveillance and loss of autonomy.

Ensuring that the benefits of AGI are broadly distributed is critical. The historical impact of technological progress suggests that most of the metrics we care about (health outcomes, economic prosperity, etc.) get better on average and over the long-term, but increasing equality does not seem technologically determined and getting this right may require new ideas.

In particular, it does seem like the balance of power between capital and labor could easily get messed up, and this may require early intervention. We are open to strange-sounding ideas like giving some “compute budget” to enable everyone on Earth to use a lot of AI, but we can also see a lot of ways where just relentlessly driving the cost of intelligence as low as possible has the desired effect.

Anyone in 2035 should be able to marshall the intellectual capacity equivalent to everyone in 2025; everyone should have access to unlimited genius to direct however they can imagine. There is a great deal of talent right now without the resources to fully express itself, and if we change that, the resulting creative output of the world will lead to tremendous benefits for us all.






Thanks especially to Josh Achiam, Boaz Barak and Aleksander Madry for reviewing drafts of this.

*By using the term AGI here, we aim to communicate clearly, and we do not intend to alter or interpret the definitions and processes that define our relationship with Microsoft. We fully expect to be partnered with Microsoft for the long term. This footnote seems silly, but on the other hand we know some journalists will try to get clicks by writing something silly so here we are pre-empting the silliness…

Reflections

The second birthday of ChatGPT was only a little over a month ago, and now we have transitioned into the next paradigm of models that can do complex reasoning. New years get people in a reflective mood, and I wanted to share some personal thoughts about how it has gone so far, and some of the things I’ve learned along the way.

As we get closer to AGI, it feels like an important time to look at the progress of our company. There is still so much to understand, still so much we don’t know, and it’s still so early. But we know a lot more than we did when we started.

We started OpenAI almost nine years ago because we believed that AGI was possible, and that it could be the most impactful technology in human history. We wanted to figure out how to build it and make it broadly beneficial; we were excited to try to make our mark on history. Our ambitions were extraordinarily high and so was our belief that the work might benefit society in an equally extraordinary way.

At the time, very few people cared, and if they did, it was mostly because they thought we had no chance of success.

In 2022, OpenAI was a quiet research lab working on something temporarily called “Chat With GPT-3.5”. (We are much better at research than we are at naming things.) We had been watching people use the playground feature of our API and knew that developers were really enjoying talking to the model. We thought building a demo around that experience would show people something important about the future and help us make our models better and safer.

We ended up mercifully calling it ChatGPT instead, and launched it on November 30th of 2022.

We always knew, abstractly, that at some point we would hit a tipping point and the AI revolution would get kicked off. But we didn’t know what the moment would be. To our surprise, it turned out to be this.

The launch of ChatGPT kicked off a growth curve like nothing we have ever seen—in our company, our industry, and the world broadly. We are finally seeing some of the massive upside we have always hoped for from AI, and we can see how much more will come soon.


It hasn’t been easy. The road hasn’t been smooth and the right choices haven’t been obvious.

In the last two years, we had to build an entire company, almost from scratch, around this new technology. There is no way to train people for this except by doing it, and when the technology category is completely new, there is no one at all who can tell you exactly how it should be done.

Building up a company at such high velocity with so little training is a messy process. It’s often two steps forward, one step back (and sometimes, one step forward and two steps back). Mistakes get corrected as you go along, but there aren’t really any handbooks or guideposts when you’re doing original work. Moving at speed in uncharted waters is an incredible experience, but it is also immensely stressful for all the players. Conflicts and misunderstanding abound.

These years have been the most rewarding, fun, best, interesting, exhausting, stressful, and—particularly for the last two—unpleasant years of my life so far. The overwhelming feeling is gratitude; I know that someday I’ll be retired at our ranch watching the plants grow, a little bored, and will think back at how cool it was that I got to do the work I dreamed of since I was a little kid. I try to remember that on any given Friday, when seven things go badly wrong by 1 pm.


A little over a year ago, on one particular Friday, the main thing that had gone wrong that day was that I got fired by surprise on a video call, and then right after we hung up the board published a blog post about it. I was in a hotel room in Las Vegas. It felt, to a degree that is almost impossible to explain, like a dream gone wrong.

Getting fired in public with no warning kicked off a really crazy few hours, and a pretty crazy few days. The “fog of war” was the strangest part. None of us were able to get satisfactory answers about what had happened, or why. 

The whole event was, in my opinion, a big failure of governance by well-meaning people, myself included. Looking back, I certainly wish I had done things differently, and I’d like to believe I’m a better, more thoughtful leader today than I was a year ago.

I also learned the importance of a board with diverse viewpoints and broad experience in managing a complex set of challenges. Good governance requires a lot of trust and credibility. I appreciate the way so many people worked together to build a stronger system of governance for OpenAI that enables us to pursue our mission of ensuring that AGI benefits all of humanity.

My biggest takeaway is how much I have to be thankful for and how many people I owe gratitude towards: to everyone who works at OpenAI and has chosen to spend their time and effort going after this dream, to friends who helped us get through the crisis moments, to our partners and customers who supported us and entrusted us to enable their success, and to the people in my life who showed me how much they cared. [1]

We all got back to the work in a more cohesive and positive way and I’m very proud of our focus since then. We have done what is easily some of our best research ever. We grew from about 100 million weekly active users to more than 300 million. Most of all, we have continued to put technology out into the world that people genuinely seem to love and that solves real problems.


Nine years ago, we really had no idea what we were eventually going to become; even now, we only sort of know. AI development has taken many twists and turns and we expect more in the future.

Some of the twists have been joyful; some have been hard. It’s been fun watching a steady stream of research miracles occur, and a lot of naysayers have become true believers. We’ve also seen some colleagues split off and become competitors. Teams tend to turn over as they scale, and OpenAI scales really fast. I think some of this is unavoidable—startups usually see a lot of turnover at each new major level of scale, and at OpenAI numbers go up by orders of magnitude every few months. The last two years have been like a decade at a normal company. When any company grows and evolves so fast, interests naturally diverge. And when any company in an important industry is in the lead, lots of people attack it for all sorts of reasons, especially when they are trying to compete with it.

Our vision won’t change; our tactics will continue to evolve. For example, when we started we had no idea we would have to build a product company; we thought we were just going to do great research. We also had no idea we would need such a crazy amount of capital. There are new things we have to go build now that we didn’t understand a few years ago, and there will be new things in the future we can barely imagine now. 

We are proud of our track-record on research and deployment so far, and are committed to continuing to advance our thinking on safety and benefits sharing. We continue to believe that the best way to make an AI system safe is by iteratively and gradually releasing it into the world, giving society time to adapt and co-evolve with the technology, learning from experience, and continuing to make the technology safer. We believe in the importance of being world leaders on safety and alignment research, and in guiding that research with feedback from real world applications.

We are now confident we know how to build AGI as we have traditionally understood it. We believe that, in 2025, we may see the first AI agents “join the workforce” and materially change the output of companies. We continue to believe that iteratively putting great tools in the hands of people leads to great, broadly-distributed outcomes.

We are beginning to turn our aim beyond that, to superintelligence in the true sense of the word. We love our current products, but we are here for the glorious future. With superintelligence, we can do anything else. Superintelligent tools could massively accelerate scientific discovery and innovation well beyond what we are capable of doing on our own, and in turn massively increase abundance and prosperity.

This sounds like science fiction right now, and somewhat crazy to even talk about it. That’s alright—we’ve been there before and we’re OK with being there again. We’re pretty confident that in the next few years, everyone will see what we see, and that the need to act with great care, while still maximizing broad benefit and empowerment, is so important. Given the possibilities of our work, OpenAI cannot be a normal company.

How lucky and humbling it is to be able to play a role in this work.

(Thanks to Josh Tyrangiel for sort of prompting this. I wish we had had a lot more time.)




[1]

There were a lot of people who did incredible and gigantic amounts of work to help OpenAI, and me personally, during those few days, but two people stood out from all others.

Ron Conway and Brian Chesky went so far above and beyond the call of duty that I’m not even sure how to describe it. I’ve of course heard stories about Ron’s ability and tenaciousness for years and I’ve spent a lot of time with Brian over the past couple of years getting a huge amount of help and advice.

But there’s nothing quite like being in the foxhole with people to see what they can really do. I am reasonably confident OpenAI would have fallen apart without their help; they worked around the clock for days until things were done.

Although they worked unbelievably hard, they stayed calm and had clear strategic thought and great advice throughout. They stopped me from making several mistakes and made none themselves. They used their vast networks for everything needed and were able to navigate many complex situations. And I’m sure they did a lot of things I don’t know about.

What I will remember most, though, is their care, compassion, and support.

I thought I knew what it looked like to support a founder and a company, and in some small sense I did. But I have never before seen, or even heard of, anything like what these guys did, and now I get more fully why they have the legendary status they do. They are different and both fully deserve their genuinely unique reputations, but they are similar in their remarkable ability to move mountains and help, and in their unwavering commitment in times of need. The tech industry is far better off for having both of them in it.

There are others like them; it is an amazingly special thing about our industry and does much more to make it all work than people realize. I look forward to paying it forward.

On a more personal note, thanks especially to Ollie for his support that weekend and always; he is incredible in every way and no one could ask for a better partner.

GPT-4o

There are two things from our announcement today I wanted to highlight.

First, a key part of our mission is to put very capable AI tools in the hands of people for free (or at a great price). I am very proud that we’ve made the best model in the world available for free in ChatGPT, without ads or anything like that. 

Our initial conception when we started OpenAI was that we’d create AI and use it to create all sorts of benefits for the world. Instead, it now looks like we’ll create AI and then other people will use it to create all sorts of amazing things that we all benefit from. 

We are a business and will find plenty of things to charge for, and that will help us provide free, outstanding AI service to (hopefully) billions of people. 

Second, the new voice (and video) mode is the best computer interface I’ve ever used. It feels like AI from the movies; and it’s still a bit surprising to me that it’s real. Getting to human-level response times and expressiveness turns out to be a big change.

The original ChatGPT showed a hint of what was possible with language interfaces; this new thing feels viscerally different. It is fast, smart, fun, natural, and helpful.

Talking to a computer has never felt really natural for me; now it does. As we add (optional) personalization, access to your information, the ability to take actions on your behalf, and more, I can really see an exciting future where we are able to use computers to do much more than ever before.

Finally, huge thanks to the team that poured so much work into making this happen!

What I Wish Someone Had Told Me

  1. Optimism, obsession, self-belief, raw horsepower and personal connections are how things get started.
  2. Cohesive teams, the right combination of calmness and urgency, and unreasonable commitment are how things get finished. Long-term orientation is in short supply; try not to worry about what people think in the short term, which will get easier over time.
  3. It is easier for a team to do a hard thing that really matters than to do an easy thing that doesn’t really matter; audacious ideas motivate people.
  4. Incentives are superpowers; set them carefully.
  5. Concentrate your resources on a small number of high-conviction bets; this is easy to say but evidently hard to do. You can delete more stuff than you think.
  6. Communicate clearly and concisely.
  7. Fight bullshit and bureaucracy every time you see it and get other people to fight it too. Do not let the org chart get in the way of people working productively together.
  8. Outcomes are what count; don’t let good process excuse bad results.
  9. Spend more time recruiting. Take risks on high-potential people with a fast rate of improvement. Look for evidence of getting stuff done in addition to intelligence.
  10. Superstars are even more valuable than they seem, but you have to evaluate people on their net impact on the performance of the organization.
  11. Fast iteration can make up for a lot; it’s usually ok to be wrong if you iterate quickly. Plans should be measured in decades, execution should be measured in weeks.
  12. Don’t fight the business equivalent of the laws of physics.
  13. Inspiration is perishable and life goes by fast. Inaction is a particularly insidious type of risk.
  14. Scale often has surprising emergent properties.
  15. Compounding exponentials are magic. In particular, you really want to build a business that gets a compounding advantage with scale.
  16. Get back up and keep going.
  17. Working with great people is one of the best parts of life.