Open Index
Build Smarter Agents using Structured Context Discussion | Link
Latest News Archive - Page 425
Danish high schoolers will have to verbally defend written assignments
Article URL: https://mezha.net/eng/bukvy/ca117584_denmark_requires_oral/ Comments URL: https://news.ycombinator.com/item?id=49224294 Points: 573 # Comments: 254
An Amazon data center could have the worst polluting power plant in the country
To power its new West Texas data center, Amazon is investing in the construction of a new power plant that could be one of the largest single producers of greenhouse gases in the US, according to the New York Times. The new gas-burning plant in Pecos County, Texas has received significant investment from Amazon and, […]
Ukraine mourns 'collector of souls' Oleksiy Yukov, killed recovering war dead
Oleksiy Yukov spent two decades recovering the bodies of soldiers — including many Russians. He was killed during a recovery mission after stepping on a landmine.
Amazon Is Creating the Biggest Pollution Source in the Country
Article URL: https://newrepublic.com/post/214111/amazon-data-center-biggest-pollution-source-entire-country Comments URL: https://news.ycombinator.com/item?id=49223845 Points: 215 # Comments: 138
Fastmail offers EU data region
Article URL: https://www.fastmail.com/blog/fastmail-offers-eu-data-region/ Comments URL: https://news.ycombinator.com/item?id=49223082 Points: 291 # Comments: 133
Buc-ee’s dodges John Oliver to sue another small business
Buc-ee's became something of a viral sensation during the World Cup, but it has a troubling history of suing small gas stations and convenience stores. On a recent episode of Last Week Tonight, John Oliver literally begged the company to sue him for selling merch featuring his squirrel mascot, Mr. Nutterbutter, with branding that reads […]
US Senate confirms Trump loyalist Todd Blanche as new AG
Republicans dropped their opposition and confirmed President Donald Trump's former defense lawyer to a Cabinet position. Now, Todd Blanche will act as the nation's top law enforcement officer.
Iran war: IRGC says Hormuz to open if US accepts Iran's terms
Iran's Revolutionary Guard said the reopening of the vital Strait of Hormuz did not depend on the outcome of ongoing talks with Oman. DW has more.
UAE says Iran targeted ADNOC tanker in Strait of Hormuz, no casualties
Gulf and Arab governments condemn Iran over missile strike on Emirati tanker.
AI adoption isn’t the same as AI usage
Every engineering org I’ve talked to this year has some version of the same chart. Seat activations climbing. Token spend The post AI adoption isn’t the same as AI usage appeared first on The New Stack .
Anthropic sets Claude Code to Auto Mode by default to protect developers from bad approvals
Starting August 14, Anthropic will make Auto Mode in Claude Code the default for Pro, Max, and Team plans. The company says it's safer. In tests, the classifier caught 89 percent of dangerous commands, while human reviewers caught only 13.6 percent. For the most widely used AI coding tool, this means developers are shifting further from writing code to monitoring AI output. The article Anthropic sets Claude Code to Auto Mode by default to protect developers from bad approvals appeared first on The Decoder .
Iraq’s Kurdish region wants to stay out of regional conflicts: PM Barzani
In an interview with Al Jazeera, Masrour Barzani denies any 'pressure' to enter the Iran war despite over 1,000 attacks.
Readers rate AI-generated short stories higher than human ones until they learn a machine wrote them
People can't tell ChatGPT-generated short stories from human-written ones, according to a new study. More than 2,500 participants performed no better than chance. The AI-generated texts were actually rated higher, but scores dropped as soon as participants learned a machine wrote them. The article Readers rate AI-generated short stories higher than human ones until they learn a machine wrote them appeared first on The Decoder .
Here’s how we played the massive rebound in AI stocks this week
AI stocks roared back to help lift two of the major averages to fresh records.
Now we have a timeline of the OpenAI accidental attack against Hugging Face
My comment on Now we have a timeline of the OpenAI accidental attack against Hugging Face — Hacker News. I think one of the most interesting details here might be tucked away in that first bulletin point: May 7: OpenAI starts a new training run for an experimental, unreleased model. (Do they mean an evaluation run? They say training run in the video, and later mention a “reward signal to judge how well they’re doing”, so I guess this really was about training a model, not evaluating one that was already trained.) The more I think about this the more I suspect that the fact this happened while training a new model is key to understanding what went wrong. In RLVR - Reinforcement Learning with Verifiable Rewards - you set the model a goal and have it take any steps necessary to achieve that goal. Clearly one aspect of OpenAI's training here is to RLVR their models for cybersecurity tasks. Just like pre-training benefits from dumping in vast sources of knowledge, the more tasks you can feed into RLVR the more of a general purpose capable model you get at the end. This also helps explain why the models had nothing to cause them to hold back. Those safety behaviors are added much later in the process. AND it explains (but does not excuse) why monitoring was so lax. If you're training a new model like this you presumably set it thousands of tasks like this in parallel. I can see how you might miss that a tiny subset of your training agents have started leaving each other messages in filenames on your packaging server. Someone once told me that you can't just leave the racist materials out of your training data if you want a non-racist model: it has to have seen examples of racism in order to later be taught that racism is bad. I can see echoes of that here. If your model doesn't know how to aggressively hack things how do you later teach it not to? (I have little knowledge of how RLVR works in practice so I'm looking forward to hearing from people who can help me understand if I'm on the right track here.) Tags: ai , openai , generative-ai , llms , ai-security-research , openai-hugging-face-incident
Why your KubeVirt VMs can’t move between clusters — and how EVPN fixes it
Your team has moved VMs to Kubernetes with KubeVirt. The workloads are running. The team is gaining confidence. Then someone The post Why your KubeVirt VMs can’t move between clusters — and how EVPN fixes it appeared first on The New Stack .
Is football AI-proof? Why tech investors wanted a slice of the World Cup
What was the thinking of the investors backing the now-canned plan, and are such proposals in the future inevitable?
K-pop's global wave inspires Chile's next pop stars
K-pop conquered the world — now Chile's young creatives are remixing the movement and making it their own.
A domain can now say it is for sale, in DNS
Article URL: https://specification.website/spec/foundations/for-sale-dns/ Comments URL: https://news.ycombinator.com/item?id=49221668 Points: 228 # Comments: 101
Trump's ex-lawyer Todd Blanche narrowly confirmed as US attorney general
The vote cements Trump's former criminal defence lawyer as the nation's top law enforcement official despite a rare pushback from some Republicans.
Lion saved as cub from Ukraine war and brought to US battles rare fungal disease
Four-year-old Taras, who now resides at a Minnesota sanctuary, was diagnosed with an often fatal infection A lion that escaped the war in Ukraine after being found as a four-month-old cub inside a duffel bag at Odesa railway station is facing an even greater battle for survival in his new home in Minnesota from a rare and often fatal fungal disease. Taras; his two sisters, Lesya and Stefania; and an unrelated, slightly older cub from Kyiv named Prada were brought to the US in November 2022 by the Wildcat Sanctuary , a secluded private reserve near Sandstone, Minnesota, that arranged their rescue with the International Fund for Animal Welfare (IFAW). Continue reading...
AI skills start on laptops. Enterprises inherit the mess.
“It’s unrealistic to expect an admin to have context over how every skill in a company should be optimally performing,” The post AI skills start on laptops. Enterprises inherit the mess. appeared first on The New Stack .
Claude Code sessions can now talk to each other and share context across terminals
Claude Code now lets sessions talk to each other. On macOS and Linux, instances running in parallel can send messages, share insights, and check on each other's status. The article Claude Code sessions can now talk to each other and share context across terminals appeared first on The Decoder .