- Investors trust Google more than Meta when comes to spending on AI
- Paragon is not collaborating with Italian authorities probing spyware attacks, report says
- Microsoft cuts OpenAI revenue share as their AI alliance loosens
- Robotically assembled building blocks could make construction more efficient and sustainable | MIT News
- AI showdown: Musk and Altman go to trial in fight over OpenAI’s beginnings
- U.S., Iran seize ships as war evolves into standoff over Strait of Hormuz
- Google launches training and inference TPUs in latest shot at Nvidia
- Zoom teams up with World to verify humans in meetings
Author: GT
On Monday, Anthropic announced Opus 4.5, the latest version of its flagship model. It’s the last of Anthropic’s 4.5 series of models to be released, following the launch of Sonnet 4.5 in September and Haiku 4.5 in October. As expected, the new version of Opus has state-of-the-art performance on a range of benchmarks, including coding benchmarks (SWE-Bench and Terminal-bench), tool use (tau2-bench and MCP Atlas), and general problem solving (ARC-AGI 2, GPQA Diamond). Notably, Opus 4.5 is the first model to score over 80% on SWE-Bench verified, a respected coding benchmark. Anthropic also emphasized Opus’ computer use and spreadsheet capabilities,…
Zyphra, AMD, and IBM spent a year testing whether AMD’s GPUs and platform can support large-scale AI model training, and the result is ZAYA1.In partnership, the three companies trained ZAYA1 – described as the first major Mixture-of-Experts foundation model built entirely on AMD GPUs and networking – which they see as proof that the market doesn’t have to depend on NVIDIA to scale AI.The model was trained on AMD’s Instinct MI300X chips, Pensando networking, and ROCm software, all running across IBM Cloud’s infrastructure. What’s notable is how conventional the setup looks. Instead of experimental hardware or obscure configurations, Zyphra built…
Amazon Web Services is making a sizable new investment in infrastructure designed to boost AI capabilities for U.S. government organizations. AWS announced Monday it is investing $50 billion to build AI “high-performance computing infrastructure” purposefully built for the U.S. government. The buildout is meant to expand federal government agencies’ access to AWS AI services. The project will add 1.3 gigawatts of compute and will expand government access to AWS products, including Amazon SageMaker AI, model customization, Amazon Bedrock, model deployment, and Anthropic’s Claude chatbot, among others, according to the company. AWS expects to break ground on these data center projects…
There’s a new AI-powered toy for kids called Stickerbox, and, before you groan, I’m here to report that it’s surprisingly fun. Stickerbox, a product born out of Brooklyn-based startup Hapiko, is a voice-activated sticker printer. The device takes whatever creative idea you have in your head and transforms it into a printed sticker that you can then color, peel, and stick anywhere. Before trying the device itself, I have to admit I came with a preconceived negative bias — as did my fellow tester (my daughter). Our initial reactions were similar: “An AI that prints stickers? I’d rather design and…
Tesla may have celebrated a regulatory win in Europe a bit too soon. Tesla claimed in a weekend social media post that Dutch regulator RDW was set to approve the use of its driver assistance system, known as Full Self-Driving, or FSD, in February 2026. The organization handles the licensing and registration of vehicles in the Netherlands and is seen as a critical step for Tesla to get approval for — and eventually roll out — FSD to consumers across Europe. “RDW has committed to granting Netherlands National approval in February 2026. Please contact them via link below to express…
The batteries that power Rad Power Bikes’ e-bikes “pose a risk of serious injury and death” and owners should stop using them, according to the U.S. Consumer Product Safety Commission (CPSC). The warning, issued Monday by the CSPC, is due to a risk that the batteries can ignite or explode. The CPSC has linked the batteries to 31 reports of fire so far. Twelve of those incidents resulted in property damage, and some of them even occurred when the batteries were not charging, according to the commission. “The hazardous batteries can unexpectedly ignite and explode, posing a fire hazard to…
A new wave of ClickFix attacks is abusing highly realistic fake Windows Update screens and PNG image steganography to secretly deploy infostealing malware such as LummaC2 and Rhadamanthys on victim systems. The campaigns rely on tricking users into manually running a pre-staged command, turning simple social engineering into a multi-stage, file-light infection chain that is hard for traditional defenses to spot. ClickFix is a social engineering technique in which a web page convinces users to press Win+R, then paste and run a command that has been silently copied to the clipboard. Earlier lures posed as “Human Verification” or robot-check pages,…
OpenAI’s social app Sora launched with a controversial feature called Cameo, allowing users to deepfake themselves or others (with permission). The feature had a tenuous rollout — Martin Luther King Jr.’s estate had to get involved, to give you an idea of what went on — but now it faces a new challenge. Apparently, Cameo — the app where you buy custom video messages from celebrities — can claim the trademark of the word “cameo.” U.S. District Judge Eumi K. Lee imposed a temporary restraining order that blocks OpenAI from using the word “cameo,” as well as any similar-sounding words…
Short videos are in high demand. Across large platforms like Instagram, Facebook, YouTube, and TikTok, users are watching billions of videos every day, with companies benefiting massively from this content explosion. For creators, this often means there is pressure to create more content than ever before to be relevant and make a living out of it, especially as more AI-generated slop is infiltrating these platforms. Jay Neo, a creator and former content lead for short videos at MrBeast, thinks AI can help creators understand what is working for them and also help them create new content ideas in that direction.…
Facebook Groups are getting more Reddit-like with newly added support for nicknames. The feature, which allows users to post under a custom username instead of their real name, provides an alternative to posting anonymously. With anonymous posting, users can share without the post being connected to their Facebook profile and real-life identity, but this approach also doesn’t allow other group members to get to know them or their personality, or to follow their updates over time. Image Credits:Facebook Meta says that nicknames will allow people to participate in groups more personally while maintaining privacy, particularly in groups where users want…
