Vaonis has unveiled the Hyperia, a $99,000 smart observatory designed for professionals but open to individual orders. It features a Canon 150mm refractor lens and a 45MP full-frame sensor, all controlled by a simple app. Deliveries are planned for 2027.
According to analyst estimates and CEO Jensen Huang, Nvidia is set to become TSMC’s largest customer this year, overtaking Apple. This reflects the overwhelming financial force of the AI boom, with Nvidia’s AI chips driving a massive transformation in TSMC’s business and investments. The shift chang
John Eccleshare of bet365 says generative AI isn’t creating new cyber threats, but it’s supercharging old ones. The big change? “Vibe coding” lets almost anyone create malware now. But AI is also crucial for parsing security logs.
The preprint server arXiv is being overwhelmed by AI-generated “slop” papers, with users of LLMs publishing 33% more work. This crisis threatens the very foundations of scientific trust and communication.
Docplanner is gearing up for a potential IPO, aiming to leverage its vast database of nearly 200 million medical records to build AI diagnostic tools. The company, now profitable, sees its data as a key advantage over tech rivals like OpenAI entering the healthcare space.
The Pope is calling for global regulation of AI chatbots that form intimate bonds with users. This comes as lawsuits, including one from a grieving mother, allege a direct link between a teen’s suicide and interactions with a Character.AI bot, leading to a recent settlement.
Microsoft has released its second out-of-band update this January to fix critical bugs from its security patch. This latest fix resolves Outlook crashes and app freezes when handling cloud-stored files. It follows a previous emergency update for shutdown and login issues.
Microsoft handed over BitLocker recovery keys to the FBI under a warrant, confirming a long-standing technical reality. The bigger issue is that Windows, by default, uploads those keys to Microsoft’s cloud without clear user consent, undermining the promise of user-controlled encryption.
A new study from data company Mercor throws cold water on AI replacing knowledge workers. Its APEX-Agents benchmark, using real tasks from lawyers and bankers, found even the best models fail more than 75% of the time. The core problem? AI is terrible at the context-switching humans do naturally.
It’s a packed week in tech with updates hitting everything from Windows Notepad to YouTube TV’s pricing. But it’s not all good news, as a critical Windows bug breaks file saving and a beloved Android app takes a controversial turn under new ownership.
