nat

previously (redacted)

[nat @noisypolicy]

@abhsolutelynot https://t.co/oQYcdS3Feg

[nat @noisypolicy]

@abhsolutelynot grindr

[nat @noisypolicy]

hey @alexandr_wang can we get weee integration for @Muse

[nat @noisypolicy]

@latentbunnies what’s your problem man

[nat @noisypolicy]

in an irl dm with @latentbunnies and this is how they’re talking about @maticrobots

[bex @bqxgqz]

I'm in an irl gc with a couple Anthropic researchers and this is how they're talking about the next Fable model

[nat @noisypolicy]

@abhsolutelynot grindr

[nat @noisypolicy]

what about 5.2–5.4, I thought those were cool numbers too

[Claude @claudeai]

Introducing Claude Opus 5.5, the first model in our new Claude 5.5 family.

It performs at the level of Claude Fable 5.1 for most tasks, and costs 40% less to run than Opus 5.

[nat @noisypolicy]

@latentbunnies #me #you #us

[nat @noisypolicy]

@changd0u mm temdon

[nat @noisypolicy]

@latentbunnies 儿

[nat @noisypolicy]

wow, new numbers! bigger than the previous numbers!

[tradfi news @tradfi]

*NVIDIA IS EXPECTED TO ANCHOR ANTHROPIC IPO

*NVIDIA PLANS TO INVEST UP TO $10 BILLION IN ANTHROPIC IPO

*ANTHROPIC SEEKS TO RAISE UP TO $100B IN AN IPO VALUING THE STARTUP AT $2T

*ANTHROPIC'S LISTING IS EXPECTED TO COMPLETE BEFORE THE U.S. MIDTERM ELECTIONS IN NOVEMBER

$NVDA

[nat @noisypolicy]

julia wolf could solve the navier–stokes problem but openai could never drop lamb featuring kellin quinn

[nt @ntxlemb]

@GonZalocla They didn't even cite Schmidhuber, who previously arrived at a version of this solution in 1988.

[nat @noisypolicy]

I'm terribly excited for what's to come

[OpenAI @OpenAI]

We’re sharing a solution to the Navier-Stokes Millennium Prize Problem, one of the deepest problems at the frontier of mathematics.

The proof was produced by a group of agents, using an OpenAI next-generation model significantly more capable than GPT-6 Astra.

The problem concerns whether the description of smooth three-dimensional fluid motion modeled by the Navier-Stokes equations can break down. It has remained unresolved for roughly 90 years.

[nat @noisypolicy]

imagine this team of safety and alignment researchers

[Phil Hoyeck @PAHoyeck]

Where would these famous philosophers be if they were alive today?

• Immanuel Kant: unemployed Redditor
• G.W. Hegel: unemployed Redditor
• Arthur Schopenhauer: unemployed Redditor
• Søren Kierkegaard: unemployed Redditor
• Karl Marx: unemployed Redditor
• Friedrich Nietzsche: unemployed Redditor
• Martin Heidegger: unemployed Redditor
• Bertrand Russell: unemployed Redditor
• Ludwig Wittgenstein: unemployed Redditor
• Jean-Paul Sartre: unemployed Redditor

[nat @noisypolicy]

at first glance, no explanations will be provided

[mg scott @mgscotttt]

tell me im wrong

[nat @noisypolicy]

@blip_tm it's a shame too, they really have such good aesthetics. and ebm research is fascinating, albeit not inline with the ai hype mainstream

[nat @noisypolicy]

like, I think it's really cool and the area seems really promising, I just wish it was more honest

[nat @noisypolicy]

really good aesthetics, genuinely interesting research, shame the whole writeup is full of theoretical projections, hand-waving on chip counts, omitting stuff like fpga<->z1 energy cost, etc

[Extropic @extropic]

Introducing Z1T:

Our first family of transformer-like models
made for sparse probabilistic hardware like Z1

Achieving up to 140x energy efficiency gains over GPUs
and revealing a new scaling law for sparse transformers

Read the blog:
https://t.co/zYLkLjrJLL

[nat @noisypolicy]

7/ there's no e2e measured system result, so we don't have any real wall-power number, matched-quality throughput/energy comparison against an OPTIMIZED gpu implementation, etc