Sam Altman, partway into his understated keynote at the first ever OpenAI DevDay:

Even though this is a developer conference, we can’t resist making some improvements to ChatGPT.

Before this moment, I only fuzzily understood that OpenAI operates in two separate but related markets. They have:

1. A consumer product called ChatGPT. This includes mobile apps, websites, secondary models like DALL·E 3, and integrated tools like the web browser and code interpreter. ChatGPT, used by 100M weekly actives, costs $20/month. Despite Monday being “DevDay”, OpenAI launched major new consumer-facing features. Most notably, it now allows anyone to build custom “GPTs”.

2. A growing suite of cloud services. Before OpenAI DevDay, I might have referred to this as the “OpenAI API” — a small piece of code that interfaces with OpenAI’s foundation models. Now, this seems too simplistic as the API quickly evolves into multiple value-add (and margin-add!) services like the Assistant API. This is similar to how AWS adds value (and margin) on top of its few core services. OpenAI’s cloud, with 2M registered developers, undoubtedly generates significant revenue.

Finding oneself in two very different product and market segments — suddenly and at scale — is no small feat! Yes, these markets utilize the same underlying technologies. However, their pricing models, sales strategies, and value propositions are quite different and are likely to diverge over time. This isn’t exactly unheard of in the tech industry, but it’s a significant undertaking for a company that’s been around for less than a decade.

Sam Altman, with charming understatement:

About a year ago […] we shipped ChatGPT as a “low-key research preview”… and that went pretty well.

Very few consumer services reach ChatGPT’s scale. In under a year, OpenAI accidentally defined a new consumer product category and became its heavyweight. I don’t think we need to look any further to find the future of consumer chatbots. Especially with the introduction of custom GPTs, it’s hard to see much room for creating differentiated consumer-facing chat services. Instead, developers will need to bring their unique data assets and external compute capabilites to the ChatGPT interface. The Canva demo is a compelling example: Canva is a decacorn but the expectation is that users will still start in ChatGPT.

As for the growth of a new kind of cloud, focused not on virtual compute and object storage but instead on foundation models, we should expect to see many more value-add services in the future. The logic behind this year’s launches seems straightforward: OpenAI simply observed the emerging architectural patterns behind LLM applications and implemented versions that are better and easier to use, readily available from the same vendor that provides the LLM itself. Like AWS, OpenAI is the heavyweight in this emerging cloud segment. While there will always be room for smaller players along competitive axes like model customization and privacy, and I hope we see plenty of innovation there, the default place to start building services that need LLMs will probably be OpenAI for the foreseeable future.

Today we launched our newest PSL spinout, Otis Health.

In its first iteration, Otis offers a free discount pharmacy card to the underserved 1099 worker market. A shocking percentage of contract workers have no insurance whatsoever; saving even a handful of dollars on medications can make a meaningful difference. Otis can often save much more than that.

I’m excited about Otis on three fronts. First, it brings a modern design and user experience to a market that until today has sorely lacked it. Second, Otis’ discount card is free to use; all the economic magic happens behind the scenes, in the form of complex contracts between retail chains, benefits managers, distributors, and manufacturers. Finally, Otis has its eyes firmly fixed on several adjacent services that we think will also meaningfully improve the lives of 1099 workers in the future.

Oh, and one more (the most important!) reason to be excited: the team. It’s been a blast working with Aaron, Luke, Sanford, Sharon, and Steve to get this thing out the door.

GPT-3 shipped two years ago. Its capabilities, and that of descendant language models like OpenAI Codex, astonished me on day one; two years later, I’m still just as astonished.

Earlier this week OpenAI announced ChatGPT, a new variant intended to be used in interactive dialogue. With ChatGPT I find myself astonished all over again.

One delightful discovery is that back-and-forth conversation is a good way to build code. It’s a form of literate programming that Knuth probably never imagined.

I paid for GitHub Copilot the moment I could. Copilot, which builds on top of Codex, is more than just astonishing: it’s useful. Yet, as impressive as Copilot is, playing with ChatGPT makes it clear that Copilot has the potential to do vastly more even today.

The Internet has done its work and there are too many fun ChatGPT code samples to choose from. The most compelling examples refine code just by talking it over. There are also countless short examples that might have been accomplished with the previous generation of Codex or GPT-3 but that seem to have renewed potency today, like explaining a bug or exploiting a buffer overflow. But if I were to highlight just one example that seems to capture the moment, it might be this complete absurdity from Riley Goodside:

Computers pretending to be gangsters with a knack for complexity theory. That is where we’re at today.

We’re at one of those handful of moments when our industry undergoes a deep and lasting phase transition. It’s easy to draw parallels with previous transitions like the advent of the Web in 1991. Then, as now, a new technology was introduced that collected recent advances into a package that felt wholly new. Even in its first version, that new technology was instantly useful in spite of its obvious flaws. It was easy to imagine a long road of improvements ahead. Most of all: people couldn’t look away. The Web struck like lightning. Large language models? Much the same.

Having said this, there’s also something that feels entirely different to me about this moment. We’re playing with language. Language is primal. It is quintessentially human. It is fire. It’s no accident that GPT-3 was built by an organization whose stated ambition is to develop the world’s first artificial general intelligence.

What lies on the other side of this transition? I don’t know; until we get there, I’ll have to content myself with those fast-talkin’ wise guys. Hopefully they can teach me another thing or two about computer programming along the way.

Crypto is a wild mix of culture, code, and capitalism. Its proponents speak of “going down the rabbit hole” in a positive way, as if a fundamental truth awaits on the other side. Crypto’s rabbit hole is deep, yes, but in my view it is full not of pleasant truths, but of distressing surprises that seem to animate the surrounding chaos.

On its surface, crypto’s “culture” feels both youthful and vibrant. I’m reminded of the early days of the web, when sites were weird and creators built without constraint. Then as now, nobody quite knew which way the industry’s winds blew; they built anyway. Crypto is the center of a new design trend that hearkens back to Space Jam, bringing its aesthetics into a technotronic future. It’s home to new writing and criticism that feels like the work of whip-smart semiotics majors. Like the early web, there’s lofty rhetoric about crypto’s utopian potential. New social clubs seem as enthusiastic about throwing underground raves in Paris as they are with the market value of their tokens.

This bizarre emerging culture arrives with what seems to me to be a giant asterisk. I think @pinboard said it best when he described crypto as “an unregulated casino with a hip bar scene”. He continued: “there’s nothing morally neutral about the criminality at the heart of the endeavor”. Every big-time speculator, punter, grifter, money launderer, and ransomware bandit in the world is hard at work in the casino. No matter how literate the bar scene, it’s the fruit of a poison tree.

Just inside the crypto rabbit hole, we find a genuinely new and expressive culture emerging from the grimy bits of unregulated capitalism run amok.

Code animates the crypto ecosystem; its more starry-eyed believers claim that code is all that matters. And, to be sure, there is very interesting code behind crypto. A dozen years ago, Bitcoin introduced the world to a new form of distributed consensus. Half a decade later, Ethereum proved the viability of smart contract programming. Since then, algorithmic advancements have led to low-cost high-throughput chains. Emerging financial protocols on top of blockchains — MakerDAO, which creates a dollar-stable asset through collateralization; Uniswap, an exchange without an order book; and Compound, a decentralized interest rate market — all have innovative technical and economic underpinnings. (Are they long-term valuable underpinnings? I admit my skepticism.)

The technical history of the Internet is, in part, the history of the birth, adoption, and stewardship of distributed protocols by its broader community. Blockchains follow in this tradition but also depart radically from it: they are the first protocols to arrive with an asset class attached. Protocols like SMTP and HTTP created immense value for the world but captured little for their inventors; blockchains and the protocols built on top of them upset this balance, allowing inventors to capture considerable value for themselves. Historically successful protocols typically see slow early adoption; crypto’s new protocols break this mold by nearly requiring substantial up-front speculation (or wise investment!) in order to achieve escape velocity. As a result, it’s alarmingly easy to launch systems that look indistinguishable from Ponzi schemes or multi-level marketing.

Midway down the crypto rabbit hole, we find a technically intriguing class of Internet protocols that upend the historical balance of value creation and capture, leading to surprising and problematic new dynamics.

Ultimately, healthy economies need productive ends. While speculation almost certainly drives crypto’s outsize market cap today, it’s hard to ignore the community’s rapid experimentation. Perhaps crypto will never uncover true sources of value creation; perhaps there are none to be found. (That’s certainly my instinct at the moment!) Or perhaps the essential ingredients are already in the kitchen. At the heart of every crypto experiment lies the simple abstraction of tokens: marks on ledgers maintained by the network. Anyone can wave a “magic wand” and declare the existence of a new token that the world will honor merely by collective agreement. Of course, waving wands alone isn’t enough to create value. Yesterday, we experimented by attaching goofy images of punks and apes to our tokens. Today, we’re starting to staple metadata with complex structure. Tomorrow, perhaps, we’ll wrap our tokens with increasingly sophisticated code: code that grants permissions and access rights, code that defines and enforces behavior, and code that describes how to interact with crypto’s new payment rails. There’s an immense design space to explore; it may be too bearish to entirely dismiss its potential.

The collective agreement that makes crypto’s “magic wand” possible at all is, to me, a puzzle. By nature, digital content can be reproduced at zero marginal cost. Why is it that we’ve decided to recreate scarcity in the digital world? Yes, we’ve grappled with digital scarcity in limited domains before. Video games sell digital items; the appeal is easy to understand. Software phones home to check its license. Intellectual property itself is a form of manufactured digital scarcity. Yet crypto seems to throw these doors wide open. Perhaps this was inevitable. Perhaps the centuries we’ve spent building societies and economic systems around the unavoidable problem of physical scarcity makes digital scarcity feel natural to us even though it’s unnatural to information itself. Whatever the case, the deluge of capital that floods the crypto markets today practically guarantees that we’ll experiment with manufactured digital scarcity for quite some time to come.

Deep down the crypto rabbit hole, we see that while crypto might portend a revolution, at its heart is a simple reassertion of scarcity in the digital realm — an assertion that is by no means a foregone conclusion.

Where does crypto go from here? With so many cards — cultural, technical, economic, and regulatory — in the air, it’s impossible to predict its future trajectory. But the technology industry is immensely path dependent. Particularly when buckets of money appear, feedback loops can form whose outcomes seem all but inevitable. Speculators and venture capitalists alike have inundated crypto with cash. Blockchains may become the future only because their story was told, speculated on, invested in, and told some more. Or they may disappear quickly, in the bursting of speculative and fraudulent excess. Whatever drives us, if society decides to wander further down this road, we should expect to see the same unequal outcomes we see today merely reproduced in the digital realm. Despite the utopian rhetoric, economic power always concentrates: crypto may mint new winners, but the tune will stay the same.

It turns out that having a second monitor can substantially improve a teacher’s remote teaching experience. That’s why I helped my friends Matt, Mike, Jesse, and Kenshi build and launch Two Screens For Teachers this morning. A video:

We’re pairing teachers with individual donors and getting monitors shipped around the USA. If you’re able, please donate today.