Fintech brokerage Clear Street is launching a private markets platform, beginning with Databricks stakes, to give investors more access to late-stage startups.
Oxide and Friends: The Open Weight Revolution with Simon Willison On Monday Bryan Cantrill and Adam Leventhal invited me to join their podcast to talk about the wild week we've had - with Kimi K3 showing open weight models can stand toe-to-toe with proprietary frontier ones, accidental cybersecurity attacks , and public letters about Open Weights and American AI Leadership signed by almost every big name in AI (with one notable exception ). It was a great conversation, even though it's already out-of-date! DeepSeek V4 Flash 0731 and Anthropic's own embarrassing cyber incident would absolutely have made the cut if we had recorded just a few days later. We also talk about Golden Gate Claude , the Zizians , Alameda wild turkey attacks , Soviet Marburg virus research , the Lead-crime hypothesis , and a bunch of other worthy digressions. Finally, we revisited some of our predictions from January , and we added a new Pope prediction : Prediction by the end of this year: the Pope says something about open models. Tags: predictions , ai , generative-ai , local-llms , llms , oxide , bryan-cantrill , ai-in-china , ai-security-research
smevals - a small eval suite for evaluating models, prompts, and harnesses I've been working with Jesse Vincent's Prime Radiant applied AI research lab building out this evals framework to help answer questions about the capabilities of different models. The result is smevals , a new tool for running small eval suites across different model configurations and grading the results. This blog entry describes the tool in detail. Here's the 10 second version: Tell your coding agent to run uvx smevals docs to learn the tool (this outputs the README ) Then tell it to build you an eval suite Once you've created an eval - which takes the form of a directory with some YAML files - you can run it against models like this: uvx smevals run path-to-eval/ -m gpt-5.5 -m claude-opus-4.6 Runs are treated separately from grading operations - you can grade your runs (against your defined set of checks) using: uvx smevals grade path-to-eval/ Then you can run a localhost web server to explore the results: uvx smevals serve path-to-eval/ Or run the smevals build command to build that report as static HTML, which you can then host anywhere. Here's an example showing an eval suite I built to evaluate how well models can write haikus. The most time-consuming part of this project was figuring out the vocabulary for it! Here's what I settled on, quoted from the announcement: An eval is a collection of challenges designed to answer a question about a model, for example, how good is that model at generating SVGs? Each eval is a collection of tasks . A task is a specific challenge, for example "Generate an SVG of a pelican riding a bicycle". When you run the eval you do so against one or more configs . Each config specifies a model to be evaluated, but may also include other parameters to test, such as different system prompts, model parameters, or agent harnesses. A run records what happened when a specific config was used to execute a specific task. A runner is the script that executes a run. Once you have collected one or more runs, you need to evaluate the results to see how well the model (or config) did. This is done by a grader , which produces a grade . Each grader runs a sequence of checks . These can be simple operations, like checking for a specific string in the output, or confirming that the output is valid XML. They can also be more complicated custom operations (implemented as scripts called checkers ), including using other models to answer questions about the run. I've been trying to figure out an approach I like for evals for several years now. smevals is my third iteration on the idea and it feels right to me. I'm looking forward to expanding this more in the future, as well as pointing it at some of my own projects. Tags: projects , ai , generative-ai , llms , llm , evals , jesse-vincent
India's app market generated a record $345 million in Q2.
Had the hacks used conventional methods, someone would likely go to prison.
A climate scientist says Earth is "passing the limits of adaptation" when it comes to withstanding the effects of climate change.
A U.S. company in Rwanda has been providing medical supplies via drone. But even state-of-the-art technology can't overcome a weak health care infrastructure.
“What on earth is Google doing?” Misinformation fears spur walk-back of AI tool.
OpenAI disrupted a Cambodia-based scam operation using ChatGPT to support investment, romance, gambling, and impersonation schemes.
As of today, July 31, 2026, we have deprecated the following models across all GitHub Copilot experiences (including Copilot Chat, inline edits, ask and agent modes, and code completions). Model… The post Gemini 2.5 Pro and Gemini 3 Flash deprecated appeared first on The GitHub Blog .
A tool that allowed anyone to generate fake AI-generated imagery and superimpose it over real Google Earth maps quickly spurred backlash.
We spotted a great deal on Tile trackers earlier this week that’s still live, but if you’re an iPhone owner, we ultimately recommend Apple’s latest AirTag. Right now, you can pick up a four-pack for $89 ($10 off) at Amazon and Target, matching the bundle’s all-time low price. If you’re a member, Costco also has […]
The incident has sparked concern in Europe while Spain also faces a challenge in balancing its relations with neighbouring Morocco and Algeria.
Google has shut down Google Earth feature it launched Thursday that allowed users to edit satellite images with text prompts using AI. The tool essentially let users create AI deepfakes of the real world using text prompts; Digital Digging's Henk van Ess, for example, intentionally generated images adding things like refugees near the Mexican border […]
Article URL: https://tailscale.com/blog/hugging-face-intrusion Comments URL: https://news.ycombinator.com/item?id=49127306 Points: 560 # Comments: 208
New research from the U.K.’s Imperial College and France’s Emlyon Business School mapped out how Silicon Valley founders commit fraud — and the role investors play.
A Chinese-speaking threat actor is suspected to be behind a fresh wave of cyber attacks targeting government organizations mainly located in Central Asia, including Afghanistan, Kyrgyzstan, Tajikistan, Uzbekistan, Kazakhstan, and the Syrian Arab Republic, since January 2025. These targeted organizations operate across several sectors, such as healthcare, research, government offices,
Native essay writing app for Mac and iPad Discussion | Link
Google Deepmind's Gemini Robotics 2 is its most advanced vision-language-action model yet, built to control everything from tabletop robots to full-body humanoids. Gemini Robotics ER 2 adds a higher-level reasoning layer for robotics tasks. The article Google Deepmind unveils Gemini Robotics 2 to power robots of all shapes from tabletop arms to humanoids appeared first on The Decoder .
You can now take advantage of user-based model policy targeting for GitHub Enterprise customers with Copilot Business or Copilot Enterprise licenses. This feature empowers AI administrators to set a baseline… The post Enterprise teams model policy targeting in public preview appeared first on The GitHub Blog .
Clement Delangue said he didn't want cyber attacks on other companies to become "normalised".
Article URL: https://github.com/yc-software/qm Comments URL: https://news.ycombinator.com/item?id=49126604 Points: 527 # Comments: 110
US president tells reporters during Camp David cabinet meeting technology ‘a hard thing to give away’ Donald Trump has backed off a promise to let Ukraine manufacture its own Patriot air defense interceptors, telling reporters that handing over the underlying technology was “a hard thing to give away” during a cabinet meeting at Camp David. The retreat, billed by the White House as “something different” for the cabinet to experience together, was the first such meeting at the presidential compound of Trump’s second term. Continue reading...
Legal challenge to King’s College London research study claimed it could harm participants but was rejected A clinical trial of drugs that can delay puberty can go ahead after a high court judge threw out a legal challenge on Friday. Campaigners had launched the attempt to halt the planned research into puberty blockers, to be run by researchers at King’s College London (KCL), on the grounds that they believed it could harm the children taking part. Continue reading...