Ani Rahul

aura farming @proximalhq

[Ani Rahul @Ani_da_dev]

@naoufal_elh Would be cool to see a writeup on this. It seems like one of the harder parts of routing.

The folks @cognition gave this explanation which seems to make sense.

[Ani Rahul @Ani_da_dev]

Do you not get killed on the cache misses with these model routers?

[Cursor @cursor_ai]

Cursor Router keeps improving from millions of in-product user interactions each week.

We intelligently classify and route requests, lowering latency and reducing cost based on the task.

[Ani Rahul @Ani_da_dev]

The researchmaxxed data teams are going to mog the human scalers

[Kyle Wong @ewveggies]

Frontier data is a research problem.

For example, some of the biggest academic CUA papers are solely focused on data synthesis for SFT (OpenCUA, from last August) and RL (CUA-Gym, from this June). Good data requires diligent researchers and engineers.

By next year, I'd expect data talent to look increasingly similar to the current research scientist talent. Already hear small stories of data talent poaching wars and such.

[Ani Rahul @Ani_da_dev]

I like the eval

[Design Arena @DesignArena]

BREAKING: Bland Speech v3 by @usebland debuts as the most human-sounding TTS model on our newest evaluation: Audio Realism Bench.

Traditional speech benchmarks increasingly struggle to separate frontier systems that all sound clean, fluent, and intelligible.

Audio Realism Bench instead asks a harder question: when two voices read the same transcript, which one sounds more like a real person?

Vetted native speakers evaluate recordings blindly across phone agents, conversational assistants, and spoken explainers, judging the subtle qualities that define human-like speech.

We also place real human recordings directly into the evaluation pool, testing whether evaluators can distinguish TTS models from actual human speech.

Bland Speech v3 takes the top spot in our initial rankings, followed by MAI-Voice-2 by @MicrosoftAI and Grok TTS by @SpaceXAI, all demonstrating important steps toward voice systems capable of sustaining genuinely human-feeling interactions.

Congratulations to the @usebland team!

[Ani Rahul @Ani_da_dev]

@MatternJustus QA hunting down reward hacks

[Brooke LeBlanc @brookeleblanc]

Made the top of our Internal "Yapper" Leaderboard at Corgi.

If you've been following awhile, Drop a comment and say hey!

Let me know what you want to see more of.

[Ani Rahul @Ani_da_dev]

@sqx_saqlain Thats actually such bad rng

[Calvin Chen @calvinchen]

Cursor and Proximal in the “nontechnical north” btw

don’t let them psyop you into thinking North Beach / Marina is not a beautiful part of SF

[Ani Rahul @Ani_da_dev]

@thsottiaux Where is the reset tibo

[jpark @ icml @jparkjmc]

yor anime pfp in company slack

certified ball knower

[Logan Kilpatrick @OfficialLoganK]

Automating AI research is going to look a lot more like data cleaning than it is going to look like inventing the transformer

[wh @nrehiew_]

If you are interested in learning more about Opus's over-editing behaviour at high reasoning effort, here's a blog for you:
1) Reasoning models specifically have a default tendency to over-edit
2) It can be alleviated with prompting

The example is ~identical too :)

[Ani Rahul @Ani_da_dev]

audit is about to be so back 👀

[tinah @tinahhong]

Tech has decided that regulation is the enemy of progress, but innovation requires capital

Regulation builds the trust necessary to inject billions into markets

To enable free innovation, Andera raised a $37M series A led by Lightspeed to scale financial oversight with agents

[Ani Rahul @Ani_da_dev]

@tinahhong The goats keeping my stonks fraud free

[Elon Musk @elonmusk]

Grok 4.5 is worth trying

[Anjney Midha @AnjneyMidha]

There’s nothing inherently accelerationist or decelerationist about open weights

Models are just information

Context and timing determine their impact, and sweeping claims otherwise deserve skepticism

[Ani Rahul @Ani_da_dev]

@grx_xce holy bingsu

[Calvin Chen @calvinchen]

we're hosting a dinner at ICML at one of the most sought after reservations in Seoul (chef was on a Netflix show) with General Catalyst!

if you like great food and a small curated group of friends and researchers, would love for you to join!

https://t.co/zTRswg3FzF

[Michael Hla @hla_michael]

I trained an LLM from scratch on pre-1900 text to see if it could come up with quantum mechanics and relativity.

While the model is too small to do meaningful reasoning, it has glimpses of intuition.

When given observations from past landmark experiments, the model can declare that “light is made up of definite quantities of energy” and even suggest that gravity and acceleration are locally equivalent.

I’m releasing the dataset + models and leave this as an open problem to the research community.

I also include what this project has taught me about intelligence in a mini essay linked below.

🧵(1/n)

[Anne Ouyang @anneouyang]

Excited to share @Standard_Kernel's seed round and some reflections on what we’ve learned about kernel generation and what we believe is next. Grateful to our amazing team, supporters, and the broader community pushing this space forward.