The One

Am what eyes cannot see, am behind the scene.

[The One @anneacan]

@CrediCorpNG Locally assembled laptop?
Is 2024 MacBook Pro M4 in the list of those locally assembled laptops?

[Slava S. @slvDev]

28.9M LLM running on an $8 esp32, writing a story on a screen! fully on the chip, nothing goes to a server.
that's ~100x bigger than the model people ran on this chip in 2023.
it grew out of @karpathy llama2.c and google's per-layer embeddings, which is what lets a big model live in flash instead of ram.

[The One @anneacan]

@krishnanrohit 😂😂😂😂

[clem 🤗 @ClementDelangue]

So proud of our security team! They caught, contained & publicly disclosed an attack unlike anything we've seen before, and did it at record speed.

Also massively grateful to @Zai_org: they shared GLM5.2 as open weights (for free!) with the world and it became a key part of our defense.

This is day one for cybersecurity in the age of agents & we're all learning that secrecy is not the answer & that all defenders (not just a few selected ones) everywhere need more powerful models without restrictions, especially open ones!

[The One @anneacan]

@OmoledaRocha @boniface_jrn 😂😂😂😂😂😂😂
This is crazy unbelievable! Never crashed into any pew. This must be Ai. The congregation don't seem surprised

[The One @anneacan]

@natolambert Another way of telling the public that their model is too dangerous them use. Fuck off OAI... Let it escape. Another model will be created to chase it down. That's how the march of progress is maintained.

[Garry P. Nolan @GarryPNolan]

Been using Fable for some perfectly reasonable, non-biology, physics work. It has been doing amazing work so far for me.

I'm a professor of Pathology at Stanford... working in cancer biology. Not developing megawatt laser weapons, not hacking quantum computer encryption, not doing anything nefarious except working on techniques for measuring things at the atomic level. My eventual bio-goal is the 3D dynamic code of DNA and how it relates to immune-cancer outcomes.

Suddenly, Fable decides to hit a safety guardrail when it's almost finished (after 6 hours of work), after spending $100s of my credits. It terminates and forces me to hand over to Opus 4.8 or Sol. It doesn't even tell me "why" it hit a safety guard, so you could potentially protest or alter queries in the future.

So now I am left wondering how "dumbed down" the result will be?

And Anthropic wonders why their reputation continues to be sullied.

Anthropic-- I want my credits back! @anthropicAI

[jack @jack]

we're launching BUZZ!

a new groupchat platform for teams of people and agents of all sizes, built to reduce our dependency on slack and github. model-agnostic, decentralized, self-sovereign, and open source. 🐝

https://t.co/8IaMVeTQNo

[Pi @pidotdev]

Hi kids, gramps here, taking over for the intern. Just released pi 0.81.0 which features first class integration with @ggerganov wonderful llama.cpp server.

Professional video demo recorded in an echo-y Austrian office below.

Shout out to our friends at @huggingface.

[Kimi Developers @KimiDevs]

A Kimi staff member used K3 on Kimi Code to build a VR companion.

She can listen, reply, show different expressions, and move between scenes like a café and a shop.

[The One @anneacan]

@Abu_Jawaadd E get one particular person wey him want sack. He's not doing that out of patriotism

[krishna @OccupyingM]

for those like me who need a @3blue1brown video to understand what this means say thanks to fable

[Bojan Tunguz @tunguz]

This is incredible. I’ve already posted a few of my thoughts about this result, so I won’t repeat myself. I just wanted to point out a few things:

1. As of this reposting, the original tweet has garnered 10M views. For a nerdy post in pure math, that’s insane.

2. This result announcement is yet another example of how AI-driven research results are being reported - on Twitter, with no carefully written long article in highly gatekept publications, such as journals, arxiv, and other such BS.

This is how scientific communications were originally conducted. I am so excited that we are going back to that model.

[levent @__alpoge__]

hello there the jacobian conjecture is false thanx to my close friend akhil for asking about it and my other close friend fable for working during the world cup final

((1+xy)^3 z + y^2 (1+xy) (4+3xy), y + 3 x (1+xy)^2 z + 3 x y^2 (4+3xy), 2 x - 3 x^2 y - x^3 z): \C^3\to \C^3, has jacobian determinant -2, and sends (0, 0, -1/4), (1, -3/2, 13/2), and (-1, 3/2, 13/2) to (-1/4, 0, 0)

[The One @anneacan]

@OoTheNigerian @Kimi_Moonshot China is currently winning the AI arms race. @DavidSacks confirmed it today. American frontier labs are currently in panic. With what Anthropic did recently, China winning is everyone winning. I regret not buying the 1 year plan when it was still available. Long live Open Source!

[The One @anneacan]

@WarMonitorINTL US is Israel and vice versa 😂

[The One @anneacan]

😂😂😂

[Richard Artoul @richardartoul]

OpenAI and Anthropic right now

[Hamel Husain @HamelHusain]

Bookmarked 100 times 🤣🤣

[Shubham Saboo @Saboo_Shubham_]

wtf is a graph.

Loops made Agent behavior programmable. Graphs make agent orgs programmable.

Dynamic Agent Org is where the graph rewrites itself while the work is happening.

[The One @anneacan]

@HummyLong @Pinovibes Lol! His naiveity is very annoying. Because of you i have forgiven him. 😂