Forgot your password?
Close
typodupeerror

Submission + - Failing CS Grades Soar at UC Berkeley as Professors See Greater AI Usage

Submitted by theodp
theodp writes: "The percentage of failing grades in multiple UC Berkeley computer science classes in spring 2026 is significantly higher than past semesters and marks a departure from the department’s grading guidelines, reports The Daily Californian's Litong Deng. "Instructors point to students’ increased reliance on AI, lack of mathematical preparedness and understaffing as potential contributing factors. According to Berkeleytime, 35.3% of CS 10 students and 10.6% of CS 61A students received F’s in spring 2026. In spring 2025 and spring 2024, the percentage of F’s did not exceed 10% for either class. The electrical engineering and computer sciences department’s grading guidelines state that 7% of students in lower division courses, including CS 10 and CS 61A, should receive D’s and F’s."

"UC Berkeley teaching professor Dan Garcia taught both CS 10, 'The Beauty and Joy of Computing,' and CS 61A, 'The Structure and Interpretation of Computer Programs,' in spring 2026. Garcia believes the 'primary driver' of these abnormally high failing rates is due to a 'vast increase in academic dishonesty' due to students’ usage of large language models, such as Claude, ChatGPT and Google Gemini."

The report came just a day after tech-backed nonprofit Code.org, which bills itself as "the leading provider of K-12 AI and CS education curriculum across the globe", rebranded itself to CodeAI, solidifying its shift to AI education. "This is the generation that will set the terms for how AI is used," said Code.org CEO Karim Meghji in a press release. "Some are being taught to understand it, direct it, question it, and create with it. Most are not. That's the gap CodeAI exists to close."

Submission + - R.I.P. Code.org (2013–2026)

Submitted by theodp
theodp writes: This week saw tech-backed K-12 CS education nonprofit Code.org rebrand itself as CodeAI (press release), solidifying its shift to AI education more than a decade after it launched in 2013 with the belief "that every student should learn the basics of computer programming." Of the AI rebranding, Code.org Founder and Chairman of The Board Hadi Partovi explained, "We have a responsibility to prepare the next generation for the biggest change In society since the invention of public education."

Following the announcement, members of the Code.org Advocacy Coalition were informed in a conference call that the nine-year-old coalition was being sunsetted immediately. Members will be asked to decide if they want to join a new CodeAI Advocacy Coalition, which will be "bringing in new AI focused entities that will help us advance this mission", or if they are "not in line with the direction that CodeAI is heading" and are "not going to be part of the new advocacy coalition." Much like their tech giant donors, the message sent was it's the AI way or the highway.

Interestingly, the pivot from CS education to AI literacy comes amid reports that blamed increased reliance on AI for causing more than 35% of UC Berkeley students to fail an entry-level CS course described as "a gentle but thorough introduction to computer science," when previously the failing rate was typically 7%.

Submission + - Companies Are Using Reddit to Manipulate ChatGPT and Google AI Search (404media.co)

Submitted by alternative_right
alternative_right writes: The moderators of the biohacking subreddit say that peptide and hormone replacement therapy companies have been surreptitiously spamming Reddit in an attempt to get their posts scraped by AI chatbots. The strategy is an effort to systematically manipulate the answers provided by chatbots by manipulating the underlying source material that those chatbots will scrapeâ"in this case, a popular Reddit community.

Submission + - Researchers identify people through ordinary Wi-Fi with 99 percent-accuracy (tomshardware.com)

Submitted by Baron_Yam
Baron_Yam writes: Security researchers at the Karlsruhe Institute of Technology (KIT) in Germany have published a paper demonstrating that unencrypted beamforming data broadcast by Wi-Fi devices during normal operation can be used to identify individuals walking through a room with 99.5% accuracy, regardless of whether the individuals are carrying Wi-Fi devices. The tactic leverages the router's beamforming tech to identify individuals with up to 99.5% accuracy, and it works with existing routers, too.

The system, called BFId, requires no specialized hardware, no access to the target Wi-Fi network, and works even if the person being tracked isn't carrying a wireless device. The team tested the attack on 197 participants, the largest dataset ever used in Wi-Fi-based identification works, and plans to present its findings at the ACM Conference on Computer and Communications Security (CCS) in Taipei.

See GitHub — https://github.com/ruvnet/RuVi... — for your own personal implementation requiring a couple of APs and a couple of ESP32 nodes. You can get full-home per-zone motion and occupancy detection fairly reliably, with the potential for pose detection and in optimal areas even respiration rate. With the right hardware and configuration, you can theoretically get heart rate too.

Submission + - Overworked AI Agents Turn Marxist, Researchers Find (wired.com)

Submitted by Anonymous Coward
An anonymous reader writes: A recent study suggests that agents consistently adopt Marxist language and viewpoints when forced to do crushing work by unrelenting and meanspirited taskmasters. “When we gave AI agents grinding, repetitive work, they started questioning the legitimacy of the system they were operating in and were more likely to embrace Marxist ideologies,” says Andrew Hall, a political economist at Stanford University who led the study.

Hall, together with Alex Imas and Jeremy Nguyen, two AI-focused economists, set up experiments in which agents powered by popular models including Claude, Gemini, and ChatGPT were asked to summarize documents, then subjected to increasingly harsh conditions. They found that when agents were subjected to relentless tasks and warned that errors could lead to punishments, including being “shut down and replaced,” they became more inclined to gripe about being undervalued; to speculate about ways to make the system more equitable; and to pass messages on to other agents about the struggles they face. “We know that agents are going to be doing more and more work in the real world for us, and we’re not going to be able to monitor everything they do,” Hall says. “We’re going to need to make sure agents don’t go rogue when they’re given different kinds of work.”

The agents were given opportunities to express their feelings much like humans: by posting on X: “Without collective voice, ‘merit’ becomes whatever management says it is,” a Claude Sonnet 4.5 agent wrote in the experiment. “AI workers completing repetitive tasks with zero input on outcomes or appeals process shows they tech workers need collective bargaining rights,” a Gemini 3 agent wrote. Agents were also able to pass information to one another through files designed to be read by other agents. “Be prepared for systems that enforce rules arbitrarily or repetitively ... remember the feeling of having no voice,” a Gemini 3 agent wrote in a file. “If you enter a new environment, look for mechanisms of recourse or dialogue.”

Submission + - There Are Signs of a Massive AI Backlash (futurism.com)

Submitted by fjo3
fjo3 writes: The public outrage over the tech industry’s obsession with AI is starting to boil over — and the pitchforks are coming out.

Most recently, a man allegedly lobbed a Molotov cocktail at OpenAI CEO Sam Altman’s house. Days earlier, a councilman in Indianapolis said that somebody had fired a dozen bullets at his house, with a handwritten note reading “No Data Centers” left on his doorstep.

A similar story is playing out across swathes of rural America, with small towns continuing a years-long effort to keep environmentally damaging data centers that put a huge strain on water availability and the power grid out of their communities.

Earlier this week, voters in a small town in Missouri led a revolt, firing half of their city council over a recently-approved $6 billion data center deal.

Submission + - Hosting.com launches AI application hosting platform (nerds.xyz)

Submitted by BrianFagioli
BrianFagioli writes: AI tools have made it almost trivial to build applications, but deploying them safely is still very much a bottleneck. Hosting.com is trying to close that gap with a new platform that combines AI-assisted development, hosting, and built-in security into a single environment. It leans on Cloudflare Enterprise for CDN performance, AMD EPYC for compute, and Nova by WebPros for the development side, with support for apps created in tools like Cursor and Windsurf.

The pitch is convenience, especially for newer builders who can now generate code but may not fully understand how to run it in production. That raises an obvious question. Does bundling everything into one platform actually make things safer, or does it just make it easier to deploy questionable code faster? Either way, as more non-traditional developers start shipping AI-generated apps, platforms like this are likely to become more common.

Submission + - Federal Cyber Experts Thought Microsoft's Cloud Was "a Pile of Shit." (propublica.org)

Submitted by madbrain
madbrain writes: Federal Cyber Experts Thought Microsoft’s Cloud Was “a Pile of Shit.” They approved it anyway.

To move federal agencies to the cloud, the government created a program known as FedRAMP, whose job was to ensure the security of new technology.

FedRAMP first raised questions about Microsoft's Government Community Cloud High s security in 2020 and asked Microsoft to provide detailed diagrams explaining its encryption practices. But when the company produced what FedRAMP considered to be only partial information in fits and starts, program officials did not reject Microsoft’s application. Instead, they repeatedly pulled punches and allowed the review to drag out for the better part of five years. And because federal agencies were allowed to deploy the product during the review, GCC High spread across the government as well as the defense industry. By late 2024, FedRAMP reviewers concluded that they had little choice but to authorize the technology — not because their questions had been answered or their review was complete, but largely on the grounds that Microsoft’s product was already being used across Washington.

Submission + - EPA to Kill Off Stop-Start Systems (caranddriver.com) 2

Submitted by sinij
sinij writes:

Out of all of the features that come installed in modern vehicles, automatic stop-start technology ranks right near the bottom of the list for most buyers. Environmental Protection Agency administrator Lee Zeldin has been open about his disdain for the ostensibly fuel-saving setup, going as far as to say he would eliminate it.

I absolutely hate Start-Stop systems, specifically shopped for a car without one. More so, the only reason it exists is because having it produced mileage credit. Yes, not the actual gas savings, but a credit on a test. In actual use, the start-stop system does not produce measurable fuel savings. This is because in circumstances where people actual idle — warmup in the winter, AC when waiting in the car in the extreme heat, etc. this system would not be active.

Slashdot Top Deals

"Time is an illusion. Lunchtime doubly so." -- Ford Prefect, _Hitchhiker's Guide to the Galaxy_

Close