Five Things AI: Gap Map, Boom, Superforecasters, Jobs, Meta Patents
There is so much AI out there! Find out what it does!
Heya and welcome back to Five Things AI!
.Everyone is telling a story about AI right now, and none of the stories agree. This week you get five of them side by side. Follow the money, follow the jobs, follow the forecasts, follow the open source stack, and follow what Meta wants to strap to your face. Read together, they say more than any single one does on its own.
Five angles on the same boom. Current AI publishes the Gap Map, a live picture of where the open source stack has momentum and where it has holes, which is the roadmap that decides whether affordable AI actually arrives. Groundbrkr runs the numbers on the two labs everyone argues about, and finds Anthropic on solid enterprise footing while OpenAI burns through roughly 57 percent of revenue with no path to positive cash flow this decade. Astral Codex Ten walks through AI superforecasters, frontier models wrapped in scaffolding that turns them into serious prediction machines. The Wall Street Journal catches Big Tech quietly retiring the jobs apocalypse it was selling a year ago, right as the layoffs land. And 404 Media reads a Meta patent for a wearable that tracks your emotions and watches you take your meds, which is exactly the company you would expect to file it.
You do not want to miss this!
Introducing the Gap Map v0.1
The Gap Map comes out of cumulative work to identify the points of highest leverage in the open source AI stack: where to build something new, where to invest in capability, where to open up the tools. By creating an up-to-date visualization of the ecosystem where we can all see both the progress and the gaps in the space, we hope to rally the community around a collective roadmap.
This is so cool and so important. I am sure that open source software will lead the way to affordable AI.
The Second Derivative: Why No One Understands the AI Boom
Anthropic is a speculative-grade, but highly insulated, credit. Its revenue is ~80% enterprise - sticky, recurring, contracted seats - and its unit economics are firmly above water, generating $1.70 of revenue for every dollar of compute. Burn converges to a manageable ~9% of revenue by 2027 as gross margins normalize.[…]
OpenAI is the “naked” borrower, making it the weakest and most volatile credit in the ecosystem. Its revenue mix is fragile - ~60% consumer. With burn hovering at a crushing ~57% of revenue through 2027 and cumulative cash destruction marching toward $115B by 2029, OpenAI has no path to positive cash flow this decade.
It remains an interesting race - while some people assume that the money will dry out soon and the bubble will burst, others do not see a bubble and want everyone to invest even more into AI.
The AI Superforecasters Are Here
An AI superforecaster is an AI - usually a frontier model like ChatGPT or Claude - which has been modified to be good at forecasting. This usually means a “scaffold” - a program that handholds it through a long research process with various prompts, tools, advice about when to create subagents, etc. The overall experience is a lot like using any other AI, but slower and more expensive, because it’s doing more work.
This is a truly impressive topic. Just try to wrap your head around it.
Big Tech Has Suddenly Flipped on the AI Jobs Wipeout Scenario
Collectively, the narrative has shifted from worker-light doomsday scenarios caused by AI to a future in which workers keep their jobs—and get a productivity boost.
The sentiment change isn’t limited to tech leaders: A survey by EY-Parthenon found that the percentage of CEOs who believe AI investments will result in significant reductions in head count fell from around 46% in January 2025 to 20% this May.
Not many people are keen on getting replaced, so it makes a lot of sense to reframe the agentic co-worker idea.
Meta Patents AI Device That Tracks Your Emotions, Watches You Take Your Meds
A wearable that records your every word and divines your emotions would also, necessarily, record your interactions with other people. Meta has pioneered non-consensual public recording with its smartglasses so it’s not shocking to see it file a patent that suggests it’ll move further into that space.
The last time Meta explicitly pursued user’s emotional data, it horrified people. In 2012, the company then called Facebook conducted a study into “emotional contagion” using Facebook’s newsfeed. Meta altered the feeds of 700,000 users to see if it could make them happy or sad just by tweaking what they saw online. Meta found that it could, in fact, alter people’s moods if it wanted. It did this without informing users they’d been part of an experiment.
This sounds like a typical move by Meta - move fast and break things - and ruin your reputation, if you till have some left…
