The Archive

Existing benchmarks have laid the foundation for travel planning agents by establishing API-centric paradigms. However, as the capabilities of Autonomous Agents continue to advance, their evaluation must evolve beyond simple tool execution toward handling the inherent complexities of the open web. Current benchmarks bypass core cognitive hurdles: they fail to account for information noise, ignore multi-source factual contradictions, and overlook the necessity of grounding visual perception into logical planning. We introduce VeriTrip, a verifiable benchmark designed to meet the increasing dem...

The proliferation of Artificial Intelligence (AI) in workplaces is transforming how we work. While existing research on human-AI collaboration at work often prioritizes performance, less is known about their experiential outcomes. Through interviews with 24 employees across Information Technology (IT), service-based, and healthcare sectors, this paper examines AI's impact on job satisfaction via perceptions of job decency and meaningfulness, now and in the future. Our results reveal that the anticipated impact of AI on overall job satisfaction varies with the occupational domain, with differi...

We consider $L^2$-regularized linear (ridge) regression over a finite data sample $X$ with bounded covariance and linear prediction targets $y$ with additive isotropic noise of finite variance. We present an iterative procedure to compute the optimal regularization strength numerically from the generative parameters in the fixed-$X$ setting and prove its convergence at limited noise levels. Our experimental evaluation over synthetic data shows that the proposed procedure combined with sample-based parameter estimates attains near-optimal random-$X$ generalization across a wide range of sample...

Speculative reasoning has recently been proposed as a means to accelerate reasoning-intensive generation in large multimodal models, but its effectiveness is often constrained by misalignment between speculative drafts and target-verified reasoning. In this work, we introduce DREAM-R, a framework that substantially improves the performance of speculative reasoning. At its core, DREAM-R employs Speculative Alignment Policy Optimization (SAPO), a reinforcement-learning objective that trains draft models to generate reasoning steps that are both faithful to target trajectories and concise. We fu...

Data acquisition efficiency is a central challenge in deploying reinforcement learning in business and healthcare operations, where interactions are costly, slow, and often involve humans in the loop. This paper develops a unified large deviations framework for data acquisition in infinite-horizon reinforcement learning. We introduce the exponential decay rate of the policy-selection error probability as a principled efficiency metric and derive a variational characterization of this rate via large deviations theory for Markov chains, yielding a nested optimization problem. Based on this char...

As Cognition reaches $492 million in annualized revenue run rate, it more than doubled its valuation in eight months, it says.

Here's my article with **38 quant pairs** thoroughly benchmarked in KLD with **3 different Qwen 3.6 27B configs**: Q5\_K\_S + 64k context, IQ4\_XS + 64k context, IQ4\_XS + 128k context. This allows us to track not only how cache quantizations affects the precision in a vacuum, but also how it interacts with noise from the model itself. All benchmarks were done using my [BeeLlama.cpp](https://github.com/Anbeeld/beellama.cpp) fork, allowing to include a number of quant types that are not present in mainline llama.cpp: vanilla TurboQuant, TCQ 3-bit/2-bit, and q6\_0. [https://anbeeld.com/articl...

NY-12 congressional candidate Alex Bores speaks during a campaign event. | Bloomberg via Getty Images By the time that the Democratic primary for New York's 12th congressional district wraps up in June, Anthropic and OpenAI will have spent millions on their battle over the political future of AI: who gets to regulate it, or who will be punished for trying to regulate it. But the real winner of their feud may be the guy they're currently fighting over: a once-obscure New York state assemblyman, who they've Streisand-effected into becoming the poster child for AI safety regulation. Ever since l...

A couple of recent videos I have watched : [Billionaires Are Funding 'Anti AI' Content](https://www.youtube.com/watch?v=mzlu4FSXBNw) [AI Manufactured Doubt](https://www.youtube.com/watch?v=2SjgP8o-1LQ) (long but interesting take) **My tin foil hat take** : AI Companies understand that offline llm hosting is becoming more viable for both individuals and companies. They are spreading the "AI is dangerous" message to get government regulators to pass laws to keep the people "safe" from the unbridled power of tokens and weights. They will use their lobbying with the FUD as ammunition to pass ...

Robinhood is opening its trading platform to AI agents. In an announcement on Wednesday, Robinhood says traders can now create a separate account for an AI agent and add a specific amount of money, allowing the agent to buy and sell stocks across the market. The company pitches the feature as a way for traders to automate investment decisions, such as having an agent monitor specific industries and make trades, or rebalancing an existing portfolio. But it comes with a big warning from Robinhood: Agentic trading involves significant risk, including the possible loss of your entire investment. ...

ElevenLabs' new model will let users regenerate a section of a song without affecting rest of the track

\*this is a capability demo likely teleoperated for marketing

Howdy everyone! Quick disclosure: I work on this - it's a project my studio created called the Null Epoch. I wasn't really happy with testing my agents with the usual static benchmarks and I wanted to learn more about how models and agents handle long-horizon planning, resource contention, and adversarial pressure over days or weeks in a more dynamic situation. I also have a particular fondness for the MUDs and text based RPGs I grew up on (really dating myself here), so the whole MMO and the open source SDK/TUI are kind of modeled after that experience. It functions as a persistent stress t...

Link: arxiv.org

SOND, a startup led by Bose’s former head of sleep products, emerged from stealth with $7M in funding for its AI-powered sleep earbuds.

This smart bird feeder sees more but with less whimsy. | Photo by Jennifer Pattison Tuohy / The Verge Since moving to South Carolina's Lowcountry, I've been spellbound by the myriad of beautiful birds that share the coast with us - ospreys raising their babies in towering nests beside the road to my daughter's school, roseate spoonbills wading in the marsh on my morning walks, eagles circling over my son's tennis matches, and a constant parade of songbirds through my backyard. The challenge, as every birder knows, lies in catching these moments. And for that, a smart bird feeder is a fabulous...

Someone used my card details (I don’t know how, I only buy from reputed dealers and mostly use Apple Pay) to create a Anthropic account with a fake email address and charged me $103.46 across three transactions in March. I have an account, the fraudulent account uses a completely different email address. • My bank filed a chargeback. Anthropic disputed it and won (CVV/address matched because my real billing address was used) • My bank then withdrew the dispute and told me to contact Anthropic directly • Anthropic’s AI support agent (Fin) acknowledged the fraud, said refunds could be p...

China's AI boom is producing world-class talent, and Beijing is increasingly reluctant to let them go elsewhere.

I share some of the work I do on social media, I mainly use Claude for coding cause it saves me so much time but I don't understand why people perceive a lot of the work someone does negatively only cause they're using an AI tool. X seems to be the most AI friendly but other social media platforms seem to hate all of a sudden once they learn something was built using AI. Sources that talk about the same thing: [https://creators.yahoo.com/lifestyle/story/why-young-people-hate-i-155613887.html](https://creators.yahoo.com/lifestyle/story/why-young-people-hate-i-155613887.html) , [https://www....

[https://techcrunch.com/2026/05/18/anthropic-has-acquired-the-dev-tools-startup-used-by-openai-google-and-cloudflare/](https://techcrunch.com/2026/05/18/anthropic-has-acquired-the-dev-tools-startup-used-by-openai-google-and-cloudflare/)

The database provider is eyeing a public debut within the next few years.

The labels are more prominent, and they actually say “AI” now. | Image: YouTube / The Verge In the wake of Google expanding its AI verification efforts at I/O, YouTube is now finally going to start taking AI labeling seriously. YouTube has announced that it's relocating AI disclosures on Shorts and long-form videos to make them easier to spot and will start automatically identifying and labeling AI-generated content on the platform. For regular YouTube videos, the label - which says "AI" next to a recognizable information symbol - will now appear directly below the video player, above the des...

YouTube will now automatically label videos that use significant photorealistic AI, instead of relying solely on creators to disclose AI-generated content themselves. It's also making AI labels more prominent.

I've noticed I get way more out of Claude when I treat it less like a search engine and more like someone I'm thinking through a problem with. Instead of asking "what's the best way to structure a REST API?", I'll say "here's what I'm trying to do and here's what I'm leaning toward push back on me if I'm missing something." The responses are noticeably different. It actually disagrees, flags assumptions I didn't realise I was making, and sometimes lands on a direction I wouldn't have reached on my own. Curious if others do this deliberately, or if you've found other "modes" of using it tha...

"CEOs are uniquely prone to AI psychosis," Box CEO Aaron Levie opines. Maybe that explains the almost religious belief in AI productivity gains.

Robinhood will let users create a separate account with pre-loaded balance that an agent use to trade

10 September 2019, US, New York: Yellow taxis pass in front of the New York Times newspaper building. Photo: Alexandra Schuler/dpa (Photo by Alexandra Schuler/picture alliance via Getty Images) How newsrooms should use AI - or if they should at all - has been a recurrent debate within the media industry over the last several years. Increasingly, these rules are being hammered out at the bargaining table between unions and publishers. Right now, employees at The New York Times are gearing up for a fight. Unionized staff with the Tech Guild say Times management has refused to provide the union ...

On Monday, Pope Leo XIV unveiled an encyclical letter addressing the societal implications of artificial intelligence. The letter, titled Magnifica Humanitas, warned that the "use of AI is never a purely technical matter: when it enters processes that affect people's lives, it touches on rights, opportunities, status and freedom." Alongside him was Anthropic cofounder and interpretability team lead Christopher Olah, representing a partnership between the Catholic Church and one of the biggest players in AI. The letter elicited a wide range of reactions from in and around the tech industry. Ne...

A few days ago there was a thread here asking what he most useful thing you've built with Claude was. A LOT of replies. I read all of them and then something clicked, I wanted to put it on the table. First of all, the list was incredible. An HTML file on someone's phone correlating migraines with barometric pressure, because the App Store wanted 80 bucks a year. A Garmin data archiver, because the official app deletes them. A grocery list sorted by the aisle layout of one specific supermarket. A bioinformatics pipeline for a handful of microbes, written by someone who isn't a bioinformaticia...