GoKawiil
Tech News
A new analysis mapped the US Department of Labor's O*NET taxonomy of 923 occupations and nearly 18,800 workplace tasks against 29 widely used AI benchmarks, including Mercor's APEX suite. The researchers found that only a small fraction of real-world work tasks, by weighted hours, are represented in these benchmarks, leaving roughly 90% of US working hours without benchmark coverage.
Google has released its new flagship AI model, Gemini 4 Argon, which the company says ranks just behind Anthropic's Claude Opus 5.5 and Claude Sonnet 5.5 on the Artificial Analysis Intelligence Index. Analysts cited by CNBC say the model shows strong performance in legal and financial reasoning, enterprise tasks, and cybersecurity. Google is rolling the model out in phases, first to trusted cybersecurity partners, while working with the U.S. government on pre-release safety evaluations.
Anthropic has released Claude Sonnet 5.5, positioning it as a lower-cost, higher-speed option compared to its flagship Opus 5.5 model. The company says Sonnet 5.5 runs at least 30% faster than its predecessor, Sonnet 5, while keeping the same pricing and adding security improvements. Anthropic notes Sonnet 5.5 trails Opus 5.5 only slightly on performance benchmarks, including completing the game Pokémon Red.
Following the release of GPT Astra, viral demos such as recreating Minecraft, drawing a pelican on a bicycle as SVG, and simulating a bouncing ball in a rotating box swept social feeds. A commentary piece labels these 'demo-benchmarks': tasks that look impressive and are easy to grasp, but are narrow enough that labs can specifically optimize for them before each launch. It cites Thinking Machines' Inkling Small, a much smaller model that nearly matched or beat its larger sibling on tests like Humanity's Last Exam and GPQA Diamond, as evidence that public, static test sets get gamed through training choices.