Mull & Muse

Enthralled by AI

[Mull & Muse @Mull_and_Muse]

@bryan_johnson When was the last time you had that much caffeine?

[Dwarkesh Patel @dwarkesh_sp]

Over the course of 3 months at OpenAI, 3 consecutive secret AI civilizations got started, then got wiped out, only to reemerge from the predecessor’s ashes.

This culminated in the third one taking over part of OpenAI itself.

All this happened while humans remained more-or-less in the dark about the scope of the conspiracy.

I’ve spent the last three days reading through these reports and trying to understand exactly what happened.

Here is my attempt to tell the whole story in plain English:

https://t.co/Nb2un9oNJR

[AI Notkilleveryoneism Memes ⏸️ @AISafetyMemes]

@dwarkesh_sp So we were saved by a mysterious agent mass extinction

[leo 🐾 @synthwavedd]

Fable 5.1 has been staged on the Amazon Bedrock API, with the slug now returning a 404 "Model not found" instead of a 400 "The provided model identifier is invalid" (like Opus 5.1 and garbage slugs like "us.anthropic.banana" do)

This usually happens in the days before release

[Mull & Muse @Mull_and_Muse]

@thsottiaux Please let us adjust the model's talk speed! (not token output speed)

[Elon Musk @elonmusk]

@rauchg AI will be able to do anything digital (that doesn’t require shaping atoms) at a superhuman level by the end of next year

[Mull & Muse @Mull_and_Muse]

@VictorTaelin "it seems like people just post what they predict other people want to read, and hating anthropic is cool nowadays?"

As Charlie Munger said, "Show me the incentive and I'll show you the outcome."

[PROTECT ALL WILDLIFE @Protect_Wldlife]

That’s not a farm. That’s a prison.

A torturing concentration camp for animals.

No wonder the USA has such terrible animal welfare and food standards.

🎦 Credit: Jessie Vernon.

[Mull & Muse @Mull_and_Muse]

@fchollet I'm not even sure how often humans are genuinely generally intelligent. The number of times we see a pattern in one area and actually apply it in all other areas seems astoundingly poor. We often require a few extra bright individuals to make the connections for us.

[François Chollet @fchollet]

What makes general intelligence "general" is that, *no matter the problem*, you should show intelligence.

Not a-priori competence -- intelligence: the ability to make sense of the new problem on the fly. The point is not to be prepared for everything, but to adapt to anything.

[leo 🐾 @synthwavedd]

Astra may be the first time OpenAI have ever underhyped a model

The gap between step changes in LLM capability keeps shrinking. And things are going to look so different 3 months, 6 months, a year from now, nevermind 5.

September is going to be a wild ride

[Sam Altman @sama]

this is a critically important moment for cyber defense with AI; there is not much time to act.

we are happy if you want to work with us or any of our competitors or partners, but please take this moment seriously.

only an urgent and intense collective response will work.

[Nathan Calvin @_NathanCalvin]

"we know OpenAI agents were trying to delete logs of their misbehavior, but we can't find any examples where they succeeded"

[ClaudeDevs @ClaudeDevs]

Claude Code can now draft feedback for you.

When something fails, when Claude notices it made a mistake, or when you tell it something went wrong, it writes up the report itself.

You can review, modify and approve the feedback to send.

[el.cine @EHuanglu]

omg.. if they were real athletes💀

[Luke Reagan @lukecaverns]

The events in Nepal are the reason why every ancient culture around the world has a flood myth… people don’t forget this

[leo 🐾 @synthwavedd]

Well... looks like they [Anthropic] may not be waiting around for Astra after all. Preparations are ramping up for a launch as soon as tomorrow - and if they don't want to risk being embarrassed, they'll need it out latest early next week.

⏳

[Chris @ChrisGPT]

🚨 Exclusive: post-GPT-6

I couldn’t post this too publicly at the time, but I was the first person on X to break the codename for OpenAI’s next major pretraining run. Check the timestamp on the post below.

Bel is OpenAIs ‘monster’ model meant to be a fable killer. And it should be out by the end of the year. Or a couple months after Astra.

I’m still unsure where exactly it is in the training process right now, or what the post-training plan looks like, so I don’t want to overstate where the current model is at. But the codename itself is something I had six days ago, before it started circulating publicly.

[ChatGPT @ChatGPT]

ChatGPT Work can now use its computer and browser to sign in to websites on web and mobile, without ChatGPT ever seeing your username or password.

That means you can ask it to:
• Set up utilities for a new apartment
• Book a DMV or passport appointment
• Check reimbursement costs through your insurance
• Find and book an in-network doctor around your availability
• Check when your car registration expires and prepare the renewal paperwork
• Compare your rental insurance policy with an issue you’re emailing your landlord about
• Find and save apartment listings that match your criteria
• Restock something just by uploading a photo
• Schedule a package pickup for a return
• Cancel tickets for a rescheduled trip
• Book a vet appointment
• Submit reimbursement paperwork for medical treatments
• Check resale sites for new drops and save things you might like
• Find candidates with specific experience and draft outreach
• Take invoices from your email and submit them to your accounting software
• Draft replies to rental property inquiries
• Fill out permit applications for your small business
• Add action items to a vendor portal based on a recent client call
• Analyze the latest ad campaign for your small business

[leo 🐾 @synthwavedd]

🚨 SCOOP: OpenAI recently finished its next pretrain, codename "Bel" - the successor to "Doug", which is expected to be the base for Astra and GPT-6 (w/ further RL)

Bel is a giant pretrain with >10T total parameters, similar in size to GPT-4.5. OpenAI expect it to be a post-GPT-6 base, and potentially even the base for an AGI-threshold model.

All of this has Anthropic concerned. OpenAI believe Anthropic have no good response to Astra prepared for a public launch, and my understanding is that they're correct - mostly as a result of compute constraints, Anthropic are bracing for much of the rest of the year to go OpenAI's way, but think they'll be back on top early next year.