hustwindmaple

Re:Google's Open Agentic Orchestrator on Sunday September 20, 2026 @07:51PM (#49781395)

more like create a big launch for promo, then maintain it briefly then slowly let it die

Re:World Model Architectures: A Survey (2018–2026) on Wednesday August 12, 2026 @08:27PM (#49280354)

Nice work!

Re:SpaceX IPO demand is approaching four times oversubscribed on Tuesday June 09, 2026 @07:37PM (#48469292)

There is an IPO tab. Click that and register your interest. Schwab will notifiy you with allocation.

Re:Alphabet announces $80B equity capital raise to expand AI infra and... on Monday June 01, 2026 @10:07PM (#48365086)

They've already issued $80B in 6 markets/currencies. How much more do you think they can raise at a decent rate? I think they might issue more next year.

Issuing new equity might be a financial engineering experiment. No other mag7 has tried it. Plus they got BH name on the plate.

Re:Designing AI Chip Software and Hardware on Monday March 23, 2026 @09:20PM (#47497500)

Oh wow, amazing doc

Re:Autoresearch: Agents researching on single-GPU nanochat training au... on Saturday March 07, 2026 @08:19PM (#47293336)

I suspect Ant is already doing this for Claude. Takes a sh*t ton of compute though.

Re:Fei-Fei Li's World Labs raised $1B from A16Z, Nvidia to advance its... on Wednesday February 18, 2026 @07:27PM (#47068361)

entertainment: gaming, videos, VR

learning

AGI: DeepMind is trying to build general agents within Genie (+SIMA)

Re:Google Titans architecture, helping AI have long-term memory on Sunday December 07, 2025 @09:52PM (#46187786)

The amazing thing about this is the first author has published multiple high-impact papers with Google Research VPs! And he is just a 2nd-year PhD student. Very few L7/L8 RS/SWEs can even do this.

Re:Amazon launches Trainium3 on Tuesday December 02, 2025 @08:58PM (#46129416)

I met a AWS engineer a couple of weeks ago and he said Trainium is actually being used for Anthropic model inference, not for training. Inferentia is basically defected Trainiums chips that nobody wants to use.

Re:The fall of Labubus and the mush of modern internet trends on Tuesday November 25, 2025 @09:27PM (#46053527)

Chinese trash

A Jax-Native LLM Post-Training Library

Posted by hustwindmaple on Friday October 03, 2025 @02:28AM
from the github.com dept.

https://github.com/google/tunix

( Read More... | no comments | 2 points | on HN )

Re:Integer Programming (2002) [pdf] on Monday September 08, 2025 @08:39PM (#45176025)

There are some new research that applies transformers to MIPs. I'm looking forward to LLMs cracking MIPs at some point and beat those commercial solvers.

Re:Google has eliminated 35% of managers overseeing small teams in pas... on Wednesday August 27, 2025 @07:34PM (#45046591)

It's not just for engineers. There are some non-engineering managers who have been demoted into ICs because they don't have enough ppl to manage.

Re:Google execs say employees have to 'be more AI-savvy' on Tuesday July 29, 2025 @09:19PM (#44730106)

Any decent company would have comprehensive tests and peer review in place to mitigate AI bs

Re:Building a ML Transformer in a Spreadsheet [video] on Monday March 14, 2022 @09:23PM (#30680733)

impressive