Archive: December 2022
Anker Eufy Window Camera E10 Review: A Weirdly Good 'Outdoor' Camera
OpenClaw and Moltbook: A DIY AI agent and social media for bots

Credit: https://openclaw.ai/ If you're following AI on social media, even lightly, you will likely have come across OpenClaw. If not, you will have heard one of its previous names, Clawdbot or Moltbot. Despite its technical limitations, this tool has seen adoption at remarkable speeds, drawn its share of notoriety, and spawned a fascinating "social media for AI" platform called Moltbook, among other unexpected developments. But what on Earth is it? What is
Thinking of AI-written vows? A study explains why it can backfire

Credit: Olga Lazurenko from Pexels Psychologists at the University of Kent are suggesting people think wisely about their use of ChatGPT this Valentine's Day after new research has revealed that we judge people most when they use AI to write love letters, apologies, and wedding vows. Their findings, published in Computers in Human Behavior, suggest that people tend to judge those who outsource tasks to artificial intelligence more negatively than those who
YouTube promises custom feeds and a lot more AI later this year - Ars Technica

The annual Made on YouTube event has just wrapped up, offering an early look at the features you can expect to come to Google’s dominant online streaming platform in the coming months. Yes, there’s plenty of AI for both viewers and creators, but there’s also a handful of non-AI updates mixed in there, too. The design of YouTube’s homepage feed comes under fire often, but the platform is preparing to give you more options. With Custom Feeds, you’ll be able
Working group established to consider implementing “blocking”; Ministry of Internal Affairs and Communications to accelerate legal framework development - ITmedia NEWS

Ministry of Internal Affairs and Communications = Tokyo, Kasumigaseki The working group will be established under an expert panel convened by the Ministry of Internal Affairs and Communications at the beginning of last month to strengthen cybersecurity measures by telecommunications operators, and will hold its first meeting in late November. It is expected to compile an interim report as early as 2027. Regarding blocking, the government’s emergency measur
Uber puts all 30,000 employees on ChatGPT Work
ChatGPT Work was introduced in July as an enterprise-focused AI agent that corporate teams could deploy for various workflows. Uber originally piloted ChatGPT Work with 1,000 employees earlier this year but has now scaled up its usage to the entire company. That earlier pilot was deployed to workers from across Uber’s various departments, including finance, operations, and product. Indeed, a majority of the early adopters — some 80% — were non-engineers, S
OpenAI makes its Jev competitor widely available as decision model competition heats up
TypeSafe’s Jev model created a category for deep learning models that act as super classifiers: Given a set of options, they can quickly make choices and offer a confidence rating in their decision. Software developers quickly found that these decision or System One models could replace LLMs in a variety of workflows, operating more quickly and a at dramatically lower cost. Nikunj Handa, the product lead for OpenAI’s APIs, told TechCrunch that his colleagu
Data centers are turning to old jet engines as the wait for power grows
The engines, pulled from retired aircraft and rebuilt to burn natural gas, aren’t anyone’s idea of a permanent power plant. But they’re another sign of how far power supply has fallen behind the AI buildout. For companies stuck in those lines, the main draw of converted jet engines is how quickly the engines can be up and running. “It’s all about getting power as fast as possible,” Rystad Energy analyst Audun Martinsen told the Times. With jet-powered unit
Astra’s 0.016% minimally invasive surgery cured Qwen’s 30,000-token infinite loop

When “understanding weights” and “rewriting weights” are connected by the same model, “AI improving AI” gains a shortcut that bypasses gradient descent, and the boundary between model maintenance and the amplification of model capabilities begins to blur. Author | Zheng Jiamei Editor | Cen Feng A model tackling a math problem generated more than 30,000 tokens while continuing to repeat similar reasoning, and ultimately failed to produce even an answer. In