While you might know Kagi best as the paid competitor to Google's ever-worsening search product, the company launched its Kagi Translate tool back in 2024, saying at the time that it was a 'simply better' competitor to tools like Google Translate and DeepL. At launch, the company said Kagi Translate 'uses a combination of LLMs, selecting and optimizing the best output for each task,' a fact that 'can occasionally lead to quirks that we're actively working to resolve.'
Our measure, 'observed exposure,' compares the tasks LLMs are theoretically capable of to the tasks people actually use Claude for at work. We find that actual usage is far from reaching theoretical capability.
Nvidia CEO Jensen Huang forecast that capital expenditure (CapEx) on datacentres would increase from the $300-400bn mark today to $3-4tn by 2030, effectively claiming datacentre spending would increase tenfold during this period.
Meta was recently granted a patent in Dec, 2025 that would essentially allow the social media platform to post on a dormant user's behalf-whether they took a break from social media or long after they've passed away. The patent, first filed in 2023, describes a large language model that "simulates" a user's social media activity, using a user's comments, likes, or content to respond to other users and also references technology that would simulate video or audio calls with users.
Yaghi describes AI not as a silver bullet, but as an advanced form of statistical pattern recognition-tools that can identify trends in data that may be difficult or time-consuming for people to uncover on their own. The real opportunity, he says, depends heavily on what farms are already doing. Operations that are consistently collecting and digitizing high-quality data are better positioned to benefit, whether the goal is lowering per-cow costs in a dairy, improving financial analysis, or identifying operational efficiencies.
For the first time, speech has been decoupled from consequence. We now live alongside AI systems that converse knowledgeably and persuasively-deploying claims about the world, explanations, advice, encouragement, apologies, and promises-while bearing no vulnerability for what they say. Millions of people already rely on chatbots powered by large language models, and have integrated these synthetic interlocutors into their personal and professional lives. An LLM's words shape our beliefs, decisions, and actions, yet no speaker stands behind them.
Scientists are increasingly turning to artificial-intelligence systems for help drafting the grant proposals that fund their careers, but preliminary data indicate that these tools might be pulling the focus of research towards safe, less-innovative ideas. These data provide evidence that AI-assisted proposals submitted to the US National Institutes of Health (NIH) are consistently less distinct from previous research than ones written without the use of AI - and are also slightly more likely to be funded.
Vibe coding is a relatively new programming paradigm that emerged with the rise of AI-powered development tools. The term was coined by Andrej Karpathy, a prominent AI researcher and former Director of AI at Tesla, to describe an intuitive way of coding where developers interact with AI models using natural language commands rather than traditional coding syntax. Instead of meticulously writing every line of code, developers simply "vibe" with the AI, describing what they want, and letting the AI generate the necessary code.
