Human evaluation asks people to score, rank, compare, or annotate AI outputs. It is essential for qualities such as usefulness, tone, creativity, cultural appropriateness, factual support, and safety when no automatic metric fully represents the goal. A reliable study defines clear criteria, randomizes presentation, includes representative tasks, trains evaluators, and measures agreement. Reviewers can still be inconsistent, fatigued, biased, or influenced by style, brand, and answer length. Preference results also depend on who participates and what context they receive. Human evaluation should protect worker well-being, especially for harmful content, and be combined with automated tests that provide scale and reproducibility.
Odyssey-3 generates worlds that respond to actions, leads a published physics benchmark and powers robot demonstrations. Flash is available to try in a browser.
ChatGPT can now build interactive diagrams, maps, calculators and games inside answers. Intelligent UI is rolling out with GPT-6 across free and paid plans.
OpenAI’s annualized revenue was nearing $50 billion, below earlier reports near $70 billion. Different treatment of partner sales helps explain the gap.
Anthropic’s new policy bans extreme, repeated abuse of Claude from November 12, while excluding ordinary frustration, disagreement, fiction and testing.
Jeff Bezos says AI could let families live on one income or work three days a week, but has offered no timetable or plan for sharing the gains.
Claude Haiku 5.5 cuts costs for everyday AI tasks, with API rates up to 20x below Sonnet and Anthropic-reported benchmark gains over GPT-6 Luna.
Google’s SynthID Detector is now available globally for image, video and audio checks. It detects supported AI watermarks, rather than proving a file is real.
Google Playground turns prompts into browser games without coding. Unity Spark will add a more advanced creation path, with a closed beta coming soon.
Google’s Nano Banana 2.1 improves image quality, text and editing consistency while cutting its image-output token price to a quarter of Nano Banana Pro’s rate.
OpenAI will watermark eligible ChatGPT and Codex text in the EU, but restricted detector access and editing-related limitations complicate the promise of identifying AI writing.