.henry

[.henry @harryfiedwrld]

@Shaun_Fosmark

[.henry @harryfiedwrld]

Quick, someone check on which mathematicians recently said that they had a breakthrough!

[Chubby♨️ @kimmonismus]

This essentially confirms that OpenAI will soon have officially solved the next Millennium Problem. The rumors were true.

Via New York Times

“In addition, since the completion of Navier-Stokes, we have made substantial progress on another Millennium Prize problem. We are working through how to share these results thoughtfully.”

[.henry @harryfiedwrld]

How funny would it be if openAI was just tricking anthropic into wasting millions of dollars in compute? Being so vague about which exact problem was partially solved also makes them waste 5x as much compute.

Either way this could be bad news also because it's probably a boring counterexample. It could be some self-referential bs that makes P≠NP.

[Andrew Curran @AndrewCurran_]

OpenAI confirmed to the New York Times that they have made "substantial progress" on another Millennium Prize problem in the last five days, and are preparing to announce.

The rumors for the last 48 hours have been OpenAI solved the Hodge Conjecture, and that Anthropic has solved the Birch and Swinnerton-Dyer Conjecture. Since Navier-Stokes rumors abound, so I was reluctant to post about either. However, OpenAI's statement to the NYT now gives the Hodge rumors some very serious support.

In general people have not updated yet that the new unnamed OpenAI model, the one that finished training about two weeks ago, which I believe will be named Aeon, is massively better at math than Astra, which two weeks ago was the best in the world. Aeon solved Navier-Stokes in 88 hours, start to finish. Follow the trend line. That means everything is on the table. Literally everything. And this does not end with math. Please update. We are taking off.

[BSCN @BSCNews]

Only XRP and HBAR ETFs see inflows this week...

Only spot @Hedera and @Ripple ETFs saw net inflows on September 8, with +$431K and +$1.55M respectively.

Both product suites have been among the best-performing ETFs this year, with Canary's spot Hedera ETF holding 1.7% of HBAR's supply on its own.

[Hedera Community @HederaCommunity]

Tune in live 🔔
What if the hardest part of building on Hedera stopped being the code?

Hedera Hispana Hub #13 is a hands-on workshop: @nicolasgarc_ia shows how the Claude skill turns a real problem into a working build on @hedera

https://t.co/niFZrgzm8O

[.henry @harryfiedwrld]

@hilbertspaess im going to help build agi!

agi comes

wtf why are we doing this guys, I wasn't involved please I quit and wrote a blog I'm different

[.henry @harryfiedwrld]

@sama 👀👀👀

[.henry @harryfiedwrld]

@infowars men aren't applying for jobs because gta6 is dropping soon

[.henry @harryfiedwrld]

@BLUECOW009 @thsottiaux guess this will help https://t.co/nAoxXqbAYl

[.henry @harryfiedwrld]

@sysxplore Azure is most of their revenue, in a way they now make more money from linux then they do from windows

[.henry @harryfiedwrld]

multi tasking final boss

[.henry @harryfiedwrld]

@aienthusiast09 Astra finished under the quota . Fable burned the window

[.henry @harryfiedwrld]

GPT-6 Astra vs Claude Fable 5.1
Decided to take the old "pelican on a bike" prompt a step further into actual multi-body physics, so I tried a perpetual motion test: getting a rainbow slinky to walk down an upward escalator indefinitely.

Put GPT-6 Astra Max on the left and Claude Fable 5.1 Max on the right.

few things to note :

Resource hit: Astra was super fast and barely touched my quota. Fable burned through my entire 5-hour window and actually crashed on the max output limit before finishing cleanly.

Astra: Pretty underwhelming. It runs fast, but it completely ignores coil collisions the spring literally clips straight through itself. Definitely faking the motion rather than simulating it.

Fable: Visually way better. It actually passes the eye test at first glance, though if you look closely at the flip cycle, the escalator speed dips and the tension looks a bit weird across the steps.

Fable definitely takes the win on visual plausibility, while Astra wins on raw speed and efficiency.

Anyone spot other physical tells in the video, or have ideas for edge-case physics prompts to test next?

Original prompt:

Make me a single HTML file of a rainbow slinky going down an up-escalator forever. No libraries, just canvas and code. The slinky should be a chain of springs, each coil a different color of the rainbow. It starts folded in half like a horseshoe draped over a step. When dropped it flips end-over-end down the steps and because the escalator keeps moving up it tumbles in place and never reaches the bottom. Include a drop button and a reset button.

[.henry @harryfiedwrld]

This is common knowledge, no? 😅 This is why I have my agent write software fresh, in small isolated tools, that interface together through API. No code is duplicated. And code that I mark okay for re-use is first carefully and architecturally vetted by me.

[Mario Zechner @badlogicgames]

exactly right

[.henry @harryfiedwrld]

What's the point of motorcycle, i don't understand You get no comfort or stability the of car, and you get no exercise or freedom of a bicycle

[.henry @harryfiedwrld]

Code comments nowdays be like

[.henry @harryfiedwrld]

i genuinely dont know how anthropic will come back from this lol it projectile shits all over fable and opus, it even talks like OG OG opus 4.6 its actually amazing.
we've genuinely opened pandoras box LOL

[.henry @harryfiedwrld]

@The_Alex I'm letting astra X high loose on my project to see what it can do & how fast it munches my usage 😄

[.henry @harryfiedwrld]

its quite important to remain at the Ballmer Peak.

[Tyler @rezoundous]

AI lab is another name for drug dealer

[.henry @harryfiedwrld]

@Taniyatweets_ Linkedin is so ahh...