Instagram Story Save

Humans shouldn't be making ER diagnoses alone.

A Harvard study put AI against real ER doctors on 76 cases. AI hit 67% accuracy. Doctors hit 55%. On clinical reasoning, the model scored perfectly on 98% of cases.

Jason's take: every doctor needs an AI copilot.

The ER is overworked. We need AI to take the front door.

The fastest growing segment in healthcare right now is digital-first primary care. AI handles the triage. Humans step in when it gets serious. Trey says the front door of healthcare has already changed.

They blocked the JetBlue/Spirit merger to give consumers more options.

Spirit went bankrupt anyway. The big airlines got the gates. If you actually want more competition, stop blocking mergers and start fixing how airport gates get allocated. That's the real problem.

Fear made Anthropic the most trusted AI company.

They never said AI would destroy humanity. The media did that for them. Now Claude is the default dev tool. Not because it's the best. Because people trust who built it and why.

Let’s go Knicks!

They built Samantha from Her and it flopped.

They tweeted. They got 100 likes. They moved on.

Five months later OpenAI found the demo and built all of ChatGPT Voice on top of it. That's how ChatGPT found its voice.

ChatGPT thinks palm reading is real. And it's your fault.

It got trained on the wild west of the internet. It doesn't know what's quackery and what's not. Now ChatGPT is an expert at palm reading, tarot and astrology.

“Is this a fake startup?”

A beanie with 100,000 sensors that reads your thoughts. No surgery needed. Just look like a hipster and start thinking. Sabi does the hard work.

They shot an entire $70M movie on a gray screen.

No sets. Just an AI-generated background. Only the actors are real. Gal Gadot said she's fine with it. She can just focus on performing. This could lead to more movies being made than ever before.

All-In Summit returns to Los Angeles, September 13–15, 2026.

The best speakers.
The best parties.
The Besties
An elite attendee list.

See you there.

Applications are now open: allin.com/events

Jason Calacanis: Anthropic and OpenAI are trying to KILL OpenClaw

Why? Because an open source agent platform is an existential threat to frontier model companies.

Jason:

“ People are going to say I'm a conspiracy theorist, but the number one goal, I believe, in the large language model, frontier model space, is to kill (OpenClaw).

This is a giant movement to stop it, because this is the equivalent of having an open source Android-like player in the market, and that could be incredibly disruptive.

Because, I believe, open source is going to win the day on the large language models and take 90% of the token usage, and I think the entire frontier model space could be undercut by open source.

And I think they realize that SLMs, the smaller language models that are verticalized now, that will run on desktops and laptops and are even starting to run on the top ones, that is their biggest competitive threat, and I hope it happens.”

Just tell your agent "a model dropped,” and it installs itself.

We used to wait in line at the library to get an Ethernet card installed. That was the internet in 2001.

#ai #openclaw