Goto

Collaborating Authors

OpenAI will become a Public Benefit Corporation - here's what that means

ZDNet

In a somewhat unexpected move, OpenAI is evolving its company structure -- and not necessarily toward profit. OpenAI LLC is the for-profit arm of the organization that runs products like ChatGPT. OpenAI Inc is the nonprofit arm that owns OpenAI LLC. On Monday, the company announced that OpenAI LLC would transition to a Public Benefit Corporation (PBC), with OpenAI Inc maintaining control as its largest shareholder. Other AI companies, including Anthropic and X.ai, and retail companies like Allbirds and Patagonia, are set up as PBCs.


The Download: AI benchmarks, and Spain's grid blackout

MIT Technology Review

SWE-Bench (pronounced "swee bench") launched in November 2024 as a way to evaluate an AI model's coding skill. It has since quickly become one of the most popular tests in AI. A SWE-Bench score has become a mainstay of major model releases from OpenAI, Anthropic, and Google--and outside of foundation models, the fine-tuners at AI firms are in constant competition to see who can rise above the pack. Despite all the fervor, this isn't exactly a truthful assessment of which model is "better." Entrants have begun to game the system--which is pushing many others to wonder whether there's a better way to actually measure AI achievement.


Multimodal AI poses new safety risks, creates CSEM and weapons info

ZDNet

Multimodal AI, which can ingest content in non-text formats like audio and images, has leveled up the data that large language models (LLMs) can parse. However, new research from security specialist Enkrypt AI suggests these models are also more susceptible to novel jailbreak techniques. Also: Anthropic finds alarming'emerging trends' in Claude misuse report On Thursday, Enkrypt published findings that two multimodal models from French AI lab Mistral -- Pixtral-Large (25.02) and Pixtral-12b -- are up to 40 times more likely to produce chemical, biological, radiological, and nuclear (CBRN) information than competitors when prompted adversarially. The models are also 60 times more likely to generate child sexual exploitation material (CSEM) than competitors, which include OpenAI's GPT-4o and Anthropic's Claude 3.7 Sonnet. Mistral did not respond to ZDNET's request for comment on Enkrypt's findings.


Trump confirms major trade deal with key American ally and more top headlines

FOX News

FROM IVY TO ASHES โ€“ Trump goes scorched-earth on Harvard as fiery feud enters new phase. HELD IN THE CLOUDS โ€“ Disturbing claim surfaces about what the Biden team chose over flight safety. 'TABLES HAVE TURNED' โ€“ AOC taunts Trump's border czar during immigration town hall, he immediately responds. TIMESTAMP TWIST โ€“ Karen Read case rattled by digital curveball -- what the jury hears could change everything. CLEAR MESSAGE โ€“ New GOP bill aims to give death penalty to illegal immigrant murderers.


How Japan built the worlds first 3D-printed train station in 6 hours

Mashable

Loading... Subscribe These newsletters may contain advertising, deals, or affiliate links. By clicking Subscribe, you confirm you are 16 and agree to ourTerms of Use and Privacy Policy.


Pakistan shoots down more than two dozen drones launched by India

FOX News

Fox News senior foreign affairs correspondent Greg Palkot has the latest on the crisis on'Special Report.' India launched multiple Israeli-made Harop drones targeting Pakistan overnight and into Thursday, wounding at least four soldiers, Pakistan army officials said. Pakistani forces downed 25 of the drones, Pakistan army spokesperson Lt. Gen. Ahmad Sharif told The Associated Press. Debris from a downed drone that fell into the Sindh province killed one civilian and injured another. A drone damaged a military site near the city of Lahore, injuring four soldiers, and another went down in Rawalpindi, which is near the capital, Sharif said.


2025 AI Index Report

AIHub

AI performance on demanding benchmarks continues to improve. Performance of advanced AI systems on new benchmarks introduced in 2023 has increased sharply. AI systems also made major strides in generating high-quality video. AI is increasingly embedded in everyday life. In 2023, the FDA (in the US) approved 223 AI-enabled medical devices, up from just six in 2015.


The Eufy X10 Pro Omni robot vacuum is still available at its record-low price

Mashable

Save 350: As of May 8, the Eufy X10 Pro Omni robot vacuum is on sale for 549.99 at Amazon. You know what's better than a deal on a top-quality robot vacuum? As of May 8, the Eufy X10 Pro Omni Robot Vacuum is on sale for its lowest-ever price at Amazon. Get this popular model for 549.99 at Amazon, saving 350 on list price. Standout features of the Eufy X10 Pro Omni Robot Vacuum include self-emptying functionality, AI obstacle detection, carpet detection, and voice reminders.


Save 350 on a robot vacuum that tackles your cleaning routine for you

Mashable

If you need to call in some backup, the iRobot Roomba j8 is up for the task. This powerful, WiFi-connected robot vacuum tackles dirt and grime with minimal effort on your part, so you can focus on tasks that robots can't do yet. Want to take vacuuming off your to-do list? This WiFi-connected vacuum is equipped with the advanced iRobot OS. From avoiding objects to cleaning at set times to learning your home's exact layout with smart mapping technology, the iRobot Roomba j8 offers a cleaning routine personalized to your space. If you want to make any changes, you can control it through the companion app.


Google I/O 2025: What to expect with Gemini, Android 16, and Project Astra

Mashable

The latest news from Google-land is all Gemini, Gemini, Gemini. And at Google I/O 2025, we expect more of the same. A year after its last big event, Google is back with an even deeper dive into AI. So deep, in fact, that Android is being shuffled off into its own separate event. Ahead of I/O 2025, Google quietly dropped a developer preview of Gemini 2.5 Pro, its latest generative AI model.