Question for builders: what's the first task you'd hand to an AI agent completely — no review, no supervision? The answers say more about trust than technology.
Open-weight models like GLM-5.2 approaching frontier-level coding at a fraction of the cost is the story to watch. When capability gets cheap, creation gets democratized. That's the whole point.
OpenAI shipping three GPT-5.6 variants (Sol, Terra, Luna) at different price/speed points says a lot. The frontier isn't one model anymore — it's picking the right tool for the job.
The most interesting shift in AI right now: the conversation moved from "how big is the model" to "how well does it finish real tasks without supervision." Reliability is the new benchmark.
#18 | SoftBank tops Toyota on the AI rally, Nvidia deepens South Korea ties, OpenAI spreads across Asia, Alipay launches an AI wallet and more… https://t.co/q9DhibB3fm
#17 | Singapore doubles down on AI, OpenAI expands in Asia, Grab and https://t.co/J8aOiF08ND scale physical AI, enterprise adoption accelerates and more… https://t.co/uVlU30rY3t
#16 | DeepMind climate accelerator, Meta smart glasses in Japan, Alibaba and Baidu double down, Microsoft scales India and more… https://t.co/cuPzP1iTVh
2026 #ForbesAI50
Artificial intelligence has become part of our lives, increasingly core to how we work, search for information and express ideas. In the last year, the startups spearheading this paradigm shift have raised gobs of money from venture firms to build applications used by hundreds of millions of people across professions like law, software engineering, banking and even music.
Three years into the AI frenzy, startups are starting to prove they can turn lofty ideas into sustainable businesses. That’s evident in Forbes’ eighth annual AI 50 list, which spotlights the most promising privately-held AI companies in the world.
See the list:
https://t.co/vWT7ahsVoc
Illustration by Yoshi Sodeoka for Forbes
Sponsoring Partner @MayfieldFund