AI Slop Is Ripping Off One of Summer’s Best Games. Copycats Are Proving Hard to Kill
Peak has sold millions of copies and is Aggro Crab’s biggest hit to date. That makes it a prime target for cloning.
A new way to test how well AI systems classify text
Is this movie review a rave or a pan? Is this news story about business or technology? Is this online chatbot conversation veering off into giving financial advice? Is this online medical information site giving out misinformation?
Google adds limited chat personalization to Gemini, trails Anthropic and OpenAI in memory features
Google updated the Gemini app running of Gemini 2.5 Pro to reference all historical chats and offer new temporary chats.Read More
Tiny “talking” robots form shape-shifting swarms that heal themselves
Scientists have designed swarms of microscopic robots that communicate and coordinate using sound waves, much like bees or birds. These self-organizing micromachines can adapt to their surroundings, reform if damaged, and potentially undertake complex tasks such as cleaning polluted areas, delivering targeted medical treatments, or exploring hazardous environments.
OpenAI Designed GPT-5 to Be Safer. It Still Outputs Gay Slurs
The new version of ChatGPT explains why it won’t generate rule-breaking outputs. WIRED’s initial analysis found that some guardrails were easy to circumvent.
Robots learn human-like movement adjustments to prevent object slipping
To effectively tackle a variety of real-world tasks, robots should be able to reliably grasp objects of different shapes, textures and sizes, without dropping them in undesired locations. Conventional approaches to enhancing the ability of robots to grasp objects work by tightening the grip of a robotic hand to prevent objects from slipping.
AOL Will Shut Down Dial-Up Internet Access in September
The move will pinch users in rural or remote areas not yet served by broadband infrastructure or satellite internet. Around 175,000 households still use dial-up internet in the US.
Filtered data stops openly-available AI models from performing dangerous tasks, study finds
Researchers from the University of Oxford, EleutherAI, and the UK AI Security Institute have reported a major advance in safeguarding open-weight language models. By filtering out potentially harmful knowledge during training, the researchers were able to build models that resist subsequent malicious updates—especially valuable in sensitive domains such as biothreat research.
TD Securities taps Layer 6 and OpenAI to deliver real-time equity insights to sales and trading teams
TD Securities rolled out an AI assistant for its equity sales and research teams. The goal is to bring assistants and agents throughout the bank.Read More
OpenAI Scrambles to Update GPT-5 After Users Revolt
GPT-5 was touted as a major upgrade to ChatGPT. Not all users think that’s the case, posting threads like “Kill 4o isn’t innovation, it’s erasure” on Reddit.