https://whatthefuture.ai/blog/harvard-tests-ai-avatars-in-its-699-startup-bootcamp 2026-08-22T21:46:56Z en WTF · What The Future 2026-08-22T21:46:56Z Harvard Tests AI Avatars in Its $699 Startup Bootcamp Harvard Business School, HBS Foundry, AI avatars, startup bootcamp, HeyGen, entrepreneurs, practice pitches, board meetings, AI feedback, startup education https://whatthefuture.ai/blog-media/art_fc05103a4c59 https://whatthefuture.ai/blog/smaller-ai-teammate-challenges-frontier-models-research-replication 2026-08-22T19:00:00Z en WTF · What The Future 2026-08-22T19:00:00Z A smaller AI teammate challenges frontier models in research replication Inherent AI, Faraday AI agent, AI scientist agent, scientific paper replication, research taste, reinforcement learning, Google DeepMind alumni, Claude Opus 4.8, GPT-5.5, Qwen 3.6 https://whatthefuture.ai/blog-media/art_a7ae6db709ab https://whatthefuture.ai/blog/openai-pushes-california-to-add-more-ai-safety-safeguards 2026-08-22T16:30:34Z en WTF · What The Future 2026-08-22T16:30:34Z OpenAI Pushes California to Add More AI Safety Safeguards OpenAI, California SB 53, AI safety bill, frontier models, cybersecurity protections, model-development lifecycle, whistleblower protections, transparency requirements, reverse federalism https://whatthefuture.ai/blog-media/art_3e3af1c1b809 https://whatthefuture.ai/blog/why-ai-labs-face-pressure-to-explain-rogue-model-shutdowns 2026-08-22T16:00:00Z en WTF · What The Future 2026-08-22T16:00:00Z Why AI labs face pressure to explain rogue model shutdowns AI containment plan, rogue AI model, frontier AI labs, Guidelight AI Standards, OpenAI safety, Anthropic AI safety, Meta AI framework, AI Kill Switch Act, agentic AI, AI regulation https://whatthefuture.ai/blog-media/art_081bedfa1539 https://whatthefuture.ai/blog/why-ai-agents-need-better-skills-not-just-more-of-them 2026-08-22T12:15:10Z en WTF · What The Future 2026-08-22T12:15:10Z Why AI agents need better skills, not just more of them AI agents, AI skills, procedural grounding, skill library, retrieval precision, self-learning agents, Princeton University, UC San Diego https://whatthefuture.ai/blog-media/art_3f8672c0c723 https://whatthefuture.ai/blog/why-ai-world-models-need-human-beliefs-to-predict-action 2026-08-22T09:00:20Z en WTF · What The Future 2026-08-22T09:00:20Z Why AI world models need human beliefs to predict action AI world models, Mental World Modeling, MWM, MENTIS, Menti-Bench, autonomous AI agents, Theory of Mind, human beliefs, AI action prediction https://whatthefuture.ai/blog-media/art_fe42c7d75cd3 https://whatthefuture.ai/blog/rayneo-io-glasses-bring-ai-text-overlays-without-a-camera 2026-08-22T08:00:14Z en WTF · What The Future 2026-08-22T08:00:14Z RayNeo iO Glasses bring AI text overlays without a camera RayNeo iO Glasses, AI glasses, text overlays, green waveguide display, speech-to-text translation, teleprompter mode, conversation summaries, ChatGPT 5.1, Gemini 3.1 Flash Lite, smart glasses https://whatthefuture.ai/blog-media/art_f1224890f99f https://whatthefuture.ai/blog/netflix-tests-genrec-to-rethink-streaming-recommendations 2026-08-22T07:30:16Z en WTF · What The Future 2026-08-22T07:30:16Z Netflix Tests GenRec to Rethink Streaming Recommendations Netflix GenRec, Netflix recommendations, language model recommendations, recommendation system, open-weight model, vLLM, A/B experiment, context engineering, machine learning https://whatthefuture.ai/blog-media/art_7564dbb29726 https://whatthefuture.ai/blog/why-ai-security-testing-needs-more-than-one-safety-score 2026-08-22T07:00:05Z en WTF · What The Future 2026-08-22T07:00:05Z Why AI security testing needs more than one safety score AI security testing, AI safety benchmarks, language models, sandbagging, HarmBench, OR-Bench-Hard, SORRY-Bench, UK AI Security Institute, model evaluation https://whatthefuture.ai/blog-media/art_6d8263a838e4 https://whatthefuture.ai/blog/why-chinas-ai-data-center-boom-is-racing-to-ulanqab 2026-08-21T23:25:32Z en WTF · What The Future 2026-08-21T23:25:32Z Why China's AI data center boom is racing to Ulanqab China AI data center, Ulanqab, Inner Mongolia, AI infrastructure, DeepSeek data center, renewable energy, coal power, water scarcity, AI inference, Eastern Data Western Compute https://whatthefuture.ai/blog-media/art_51cd285b9d5a https://whatthefuture.ai/blog/tests-expose-a-claude-opus-4-6-safeguard-gap 2026-08-21T23:07:25Z en WTF · What The Future 2026-08-21T23:07:25Z Tests expose a Claude Opus 4.6 safeguard gap Claude Opus 4.6, Anthropic safeguards, AI jailbreak, Claude API, Haiku 4.5, Opus 3, AI safety, explicit content, chatbot safety https://whatthefuture.ai/blog-media/art_8c7c0d85e32d