Search: “models”

v14 Lite Release Notes: – Distilled the intelligence from HW4 V14 into HW3. This allows HW3 to directly learn how to handle scenarios using HW4 V14 as a guide. This process unlocks the improvements th

Vercel has launched voice agent capabilities on its AI Gateway platform, enabling developers to build real-time applications with speech and transcription features. The announcement indicates new inte

xAI has introduced a Build page feature to the xAI Console, expanding its development tools and capabilities for users working with xAI's platform. This addition likely enables developers to create, t

hard to see how anyone makes much money from all this in the long run. more like the airline industry; very small margins and big expenses pouring trillions in probably wasn’t wise China’s AI playbook

It is worth being very, very careful about how you are approaching routing, especially when the systems are primarily tested on verifiable IT benchmarks, which may lead you to overestimate the ability

My conversation with @ScottWu46, founder and CEO of @Cognition, the company behind Devin, the first AI software engineer. 0:00 Scott Wu's Obsession With Winning 2:06 Competitive Programming, Games And

NEW paper worth reading. Reasoning-data curation is expensive because scoring a trace usually means reading it to the end. This new work from UCLA shows you may not have to. The quality of a reasoning

Ethan Mollick notes that OpenAI has not released a GDP value metric for GPT 5.6, which he identifies as one of the most useful measures for quantifying the economic value generated by AI systems. This

> Better Caching – Cache misses are the easiest way to drive your cost up. All of our requests are cache aware, so we’re reusing a warm cache wherever possible. We do this for you in Deep Agents - see

From playing around with /goal It feels like there's less and less of a need to build any type of workflow manually (whether through code, drag and drop, or a prompt). Instead, specify the goal, let t

It really isn't an exaggeration! LiteParse clocks in at an average of 3ms per page for a reason: it skips the heavy AI processing and cloud overhead entirely. Here is exactly how it pulls off that kin

This article explores the implementation and use of local coding agents—AI systems designed to autonomously write, test, and debug code on local machines without relying on external APIs. The piece li

The talk around openness & AI needs to distinguish between the vibrant and innovative open source movement that is advancing the state-of-the-art on harnesses & other key areas and open weights fronti

This is very interesting. Coinbase seems to have lowered their token spend ($$) to about half, by 1) routing to cheap inference like GLM 5.2 and Kimi 2.7 that are still pretty performant 2) Smart rout

'AI Is Dumber Than Investors Think — ft. Gary Marcus' Ed Elson is joined by Gary Marcus to discuss why he’s concerned about the fact that we’re all-in on AI. They explore why he argues generative AI i

“An AI future where only a few US companies, subject to the whims of the US government, have frontier AI runs a high risk of dystopian outcomes.” An AI future where only a few US companies, subject to

“Gary Marcus … says that investments into scaling AI is the “greatest capital misallocation in history” which leaves everyone “on the hook”. However, he also says that it is still early days for the A

Great experiment testing how good AIs are getting at very ambitious end-to-end coding tasks. Opus 4.7, in 14 hours, was able to build a software package that would take 2-17 weeks of human engineering

“If there is a deflation of the AI bubble, the optimists say that the new infrastructure will remain even if the companies do not — just as railways survived the 19th-century railway bust. However, th

Is it just me, or are today's LLMs less likely to default to building everything in React than they were last year? I used to have to say 'don't use React' in almost all of my frontend web dev prompts

It doesn't take much skill to game this out and see where we're going if we keep this up. Kneecap our own companies. Gate everything. Let regulatory capture and fear mongering win. Have our people and

Ethan Mollick argues that regulatory frameworks for AI will not overlook or exempt open-source AI systems from compliance requirements, implying that future AI regulation will comprehensively address

Hugging Face, a leading open-source machine learning platform, has achieved $100 million in annual recurring revenue (ARR), marking a significant milestone in its growth as a commercial AI company. Th

Crazy how many times people have told me over the last five years that a solution to hallucinations was right around the corner — and yet here we still are. This study tests how often LLMs invent answ

excited to be speaking at @aiDotEngineer World Fair next week on Improving Agents, Continual Learning, and why we think a large part of it is...Data Mining! Trace Mining is how we understand agent beh

This post likely describes how to integrate voice communication capabilities with Hermes agents, enabling two-way spoken interaction rather than text-only exchanges. It probably covers implementation

lots of folks prepping talks next week (congrats!). Some thoughts from RLing on thousands of hours of engineer- and researcher- focused talks: - AI generated svgs > AI generated imgs. MAXIMUM 4 ai slo

Elon Musk argues that the primary class conflict in society exists between productive individuals and those he characterizes as parasitic, suggesting that value creation rather than traditional econom

We exist and work so leftists can rule over us. Our labor and taxes get extracted to build and sustain political clients who deliver power to the left at home and extend its reach abroad. Democrats ex

We just crossed 100M annual run-rate. I know many AI companies are capturing much more $ these days, but still proud of the milestone! Maximizing short-term revenue has never been our priority. In fac

GLM-5.2 achieved token generation speeds of 318 tokens per second, surpassing the previously reported 280 tokens per second benchmark, with indications that further performance improvements are possib

Clem Delangue, CEO of Hugging Face, expressed strong confidence in the future dominance of local and open-source AI solutions over proprietary alternatives. The statement reflects optimism about the c

It's a bird, it's a plane, it's a JEPA! Congrats on that great work that brought SIGReg and JEPAs to the sky--in the real world! Check out the paper! https://arxiv.org/abs/2606.23444 What should a wor

Clem Delangue, CEO of Hugging Face, announced a news update alongside Durga Malladi, likely regarding new developments, products, or initiatives from the company. The post was shared on X (formerly Tw

pretty sick that i get to work with Jake every day on making continual learning + memory accessible at scale for every single agent one common thread here is...the Trace a large part of Continual Lear

Nous Research recommends using the 'hermes skills config' command to enable creative-ideation and other optional skills in their system. This command appears to be a configuration tool for activating

We’ve designed and built our first AI chip: Jalapeño. Designed from the ground up by OpenAI and brought to production with @Broadcom, Jalapeño is purpose-built for the LLM workloads powering ChatGPT,

Harrison Chase suggests that evaluations (evals) are becoming as critical to AI/LLM development as Product Requirements Documents (PRDs) traditionally were, implying a shift in how teams validate and

Ran 10 more tests comparing GLM 5.2 & Opus. On average, GLM 5.2 produced 2x the tokens but was still faster + 3x cheaper with similar quality! I'm open sourcing all these tests tomorrow, including the

Together AI conducted a blind test comparing two landing pages—one created by GLM 5.2 and one by Opus 4.8—to evaluate whether users could distinguish between AI-generated designs. The test highlights

Technological revolutions rarely destroy industries. More often, they destroy the reason those industries used to exist. Photography did not destroy painting. It destroyed painting’s monopoly on reali

The advanced civilizations of sci-fi legend (Banks, Asimov, etc) have some form of simulation to guide society. @joon_s_pk is taking a crack at building that simulator with @simile_ai. As Joon's cofou

A fundamental problem with extending Codex/Cowork/Code to all knowledge work is that they remain very 'software-brained' where the end result (the software) is what is important & that code serves as

Clem Delangue, CEO of Hugging Face, celebrated reaching nearly 1,000 followers on the platform, marking a milestone in the company's social media presence. The post reflects growth in community engage

This work demonstrates that frozen image encoders (neural networks trained on static images without video or physics supervision) contain implicit geometric signals that correspond to physical plausib

1.3 GW secured. 6 GW needed. The gap is the story. Anjney Midha @AnjneyMidha runs AMP, a grid operator that pools AI compute for frontier labs. His own pipeline shows 1.3 gigawatts secured against an