Below are my six predictions for 2026. One year later, a review to see how many times I was proven wrong:

1. Although Manus boosted market confidence, after MiniMax and Zhipu上市 in Hong Kong stocks, funds will most likely first observe secondary market feedback. If signals such as IPO underperformance, excessive selling pressure, or weakening liquidity appear, Chinese VCs will be more cautious in deploying capital, which may further concentrate funds into a few top-tier targets until US VCs provide new narrative catalysts or AI achieves verifiable marginal breakthroughs. Chinese entrepreneurs will find it more difficult to raise funds.

2. AI giants will continue to eat into more niche scenarios for daily active users and user retention. Users, seeking convenience and cost reduction, will concentrate high-frequency needs on a few platform-based applications like Gemini, further squeezing the survival space of independent apps. Except for a few projects with data, industry, or ecological barriers, AI startups will need to be “small and beautiful + quick to monetize.”

3. The story of stablecoins will fade: many traditional payment companies attempting to chase stablecoin narratives will find that stablecoin adoption falls short of expectations, and traditional financial institutions will strongly resist risk controls, forcing them to shrink their crypto payment product lines; numerous stablecoin digital banking startups will face customer acquisition difficulties, entering zombie states or pivoting. However, several dark horse projects divided by industry/region will quietly grow rapidly, with the global share of stablecoin payments continuing to rise.

4. The myth of wealth creation through blockchain will diminish, and the “coin issuance, exchange listing, dump, and pump-and-dump” game of crypto projects will gradually become ineffective. But more crypto consumer apps will break out and spread: around the 2026 World Cup, prediction markets and trade-for-fun will become the hottest tracks; improved UX design and technological iterations like Passkey and Paymaster will make users unaware that they are directly using blockchain.

5. The development of AI hardware will diversify: a particular AI glasses product will find Product-Market Fit in 2026, transforming from a toy into a daily necessity relied upon by the masses; AI rings and AI wristbands will provide more health data and avoid being classified as medical devices, thus avoiding regulatory pressure and further developing, with new major players entering the scene; people will start to dislike the routine of dancing and somersaulting with embodied robots, and industry attention will shift back to AI robots capable of performing real tasks.

6. AI Agents will be able to complete longer, more complex, cross-scenario, cross-platform tasks faster and more accurately on their own. The growth of AI browsers will not explode, but average user engagement time and token consumption will continue to rise; companies will truly treat AI Agents as junior employees, and tech companies will increase the use of AI tools and systems in interviews. Solo companies coexisting with AI will become the new individual entrepreneurs in the digital age.