AI
Why language models hallucinate
Hallucinations persist partly because current evaluation methods set the wrong incentives. While evaluations themselves do not directly cause hallucinations, most evaluations measure model performance in a way that encourages guessing rather than honesty about uncertainty.
Pay-per-output? AI firms blindsided by beefed up robots.txt instructions.
The "Really Simple Licensing" (RSL) standard evolves robots.txt instructions by adding an automated licensing layer that's designed to block bots that don't fairly compensate creators for content.
Free for any publisher to use starting today, the RSL standard is an open, decentralized protocol that makes clear to AI crawlers and agents the terms for licensing, usage, and compensation of any content used to train AI, a press release noted.
All IT work to involve AI by 2030, says Gartner, but jobs are safe
Judge: Anthropic’s $1.5B settlement is being shoved “down the throat of authors”
Critics fear Anthropic will get off cheaply, striking a deal with authors suing that covers less than 500,000 works and paying a small fraction of its total valuation (currently $183 billion) to get away with the massive theft.
GPT-5 Is Making Huge Factual Errors, Users Say
AI pricing is currently in a state of ‘pandemonium’ says Gartner
If you can find the T&Cs, which are often hidden, you may spot hidden costs and nasties galore
Politics
FBI releases images of 'person of interest' and offers $100,000 reward for information in Charlie Kirk shooting
Charlie Kirk Was Practicing Politics the Right Way
Poland says it shot down Russian drones after airspace violation
"This situation brings us the closest we have been to open conflict since World War Two." This is the first time Russian drones have been downed over Nato territory since Moscow's full-scale invasion of Ukraine in 2022.
Poland Invokes NATO's Article 4 Over Russia's Drones As Tensions Rise
Following an emergency meeting of the North Atlantic Council, NATO Secretary General Mark Rutte issued a strong statement condemning Russia's violation of Polish airspace. Rutte confirmed that allied nations expressed full solidarity with Poland and denounced what he called Russia's "reckless behaviour."
Stop Acting Like This Is Normal
Ezra Klein defends Schumer's refusal to shut down the government in March, but not a single argument Schumer made then is valid now. Donald Trump is corrupting the government — he is using it to hound his enemies, to line his pockets and to entrench his own power. This is what Trump is doing to the government. This is what Democrats cannot fund. This is what they have to try to stop.
Trump Media Launching ETFs: Here’s What They Target—And How They Could Benefit From Administration’s Policies
US high school students lose ground in math and reading, continuing yearslong decline
A decade-long slide in high schoolers’ reading and math performance persisted during the COVID-19 pandemic, with 12th graders’ scores dropping to their lowest level in more than 20 years, according to results released Tuesday from an exam known as the nation’s report card.
ICE’s Raid on Hyundai Was as Stupid as It Was Cruel
Open letter against the EU's proposed Chat Control regulation
It is simply not feasible to perform detection of known and new CSAM for hundreds of millions of users with
an acceptable level of accuracy.
Geoengineering will not save humankind from climate change
New research debunks some speculative climate fixes.
Infosec
Tesla’s Dangerous Doors
When Teslas lose power, crashes can turn into deadly races against time.
Bluesky will comply with age-verification laws in South Dakota and Wyoming after exiting Mississippi
Instead of requiring Bluesky to restrict access to all unverified users, users in South Dakota and Wyoming can verify their ages through the Kids Web Services’ (KWS) solution.
The service allows users to choose from multiple methods to verify their ages, which may include payment cards, an identity document, an anonymous face scan for facial age estimation, or other alternatives. Bluesky is using the same service to comply with the U.K.’s Online Safety Act, which has similar requirements.
Claude’s new AI file-creation feature ships with security risks built in
IT enables users to generate Excel spreadsheets, PowerPoint presentations, and other documents directly within conversations on the web interface and in the Claude desktop app. While the feature may be handy for Claude users, the company's support documentation also warns that it "may put your data at risk" and details how the AI assistant can be manipulated to transmit user data to external servers.
Anthropic's Claude Code runs code to test if it is safe – which might be a big mistake
Checkmarx does not dismiss the value of AI security review, but asks developers to take note of the warnings the product itself gives, including that "Claude can make mistakes" and that "due to prompt injection risks, only use it with code you trust." There is a near contradiction here, in that if the code is completely trusted, a security review would not be necessary.
Foothill College wins top ranking in annual survey
This is the second time since 2023 that the Los Altos Hills-based college has received the top national ranking calculated through student loan amounts, alumni earnings, college acceptance data, graduation rates and diversity grades among many other factors.
|