自分はビジネスに楽しみを見出したいタイプなんだけども、「楽しければいい」という考えと「勝たないとだめ」という考えで競争が起こった時、かならず「勝たないとだめ」と思っている側がビジネスとして最終的に勝つという原理原則だけは忘れないようにしている。勝つことにこだわっている組織は強いし、怖い

2025 AI predictions 1. Superbowl 2025 is AI's mainstream moment. Every fourth commercial brags about AI, marking the exact moment AI becomes too mainstream to be cool. The backlash starts Monday February 10th. 2. A Fortune 500 company loses billions when their AI agents make thousands of bad decisions in minutes. Boards scramble to create "AI governance" roles. AI corporate crises become a thing in 2025. 3. "AI-free" becomes the new organic. Products start advertising themselves as "made by humans" or "AI-free" as a premium feature. Not because they're better, but because humans start craving the imperfection and authenticity of human-made things. 4. We move away from text to prompt interfaces. It starts to get cumbersome and new, interesting ways to prompt with AI get built. Our grandchildren will laugh we did text to prompt. 5. Consumer apps finally crack personalized AI. Not through better models, but through better understanding of human psychology and social dynamics. 6. The rise of outcome arbitrage. Companies start buying cheaper AI-generated outcomes instead of expensive human labor. This becomes commonplace. 7. Voice AI creates the first post-smartphone killer app. Something that makes typing feel as outdated as dial-up internet. 8. AI creates a content crisis so severe platforms have to completely rethink distribution. The volume of AI-generated content forces a return to human curation. 9. The rise of micro-companies - one person plus AI agents generating millions in revenue becomes normal, not exceptional. 10. The first AI-native retail brands emerge. Not just AI marketing, but entire supply chains and product designs driven by real-time consumer behavior prediction. 11. Education splits into two tracks: AI operators and AI creators. Traditional coding bootcamps die as the market realizes what skills actually matter. 12. Video generation kills the influencer game. When anyone can create perfect videos instantly, authenticity becomes the only metric that matters. 13. Browser extensions become the new App Store. The best AI tools aren't standalone apps but extensions that enhance everything you do. 14. AI makes startup ideas worthless. When anyone can build anything, distribution and timing become the only things that matter. 15. Custom AI training becomes the new digital moat. Not through better algorithms, but through better data collection and user understanding. 16. The end of linear software development. Products stop having "versions" because AI agents continuously evolve them based on usage. The idea of "shipping updates" becomes as outdated as physical software boxes. 17. Code becomes worthless but software still gets harder to build. AI writes perfect code, but system design and product sense become the new bottlenecks. 18. AI creates a new class divide: those who know how to orchestrate agents vs those who get orchestrated by them. Middle managers don't disappear - they become agent supervisors. This starts in tech jobs and moves beyond that. 19. Startup teams flip to 80% product people, 20% engineers. When AI handles the building, knowing what to build becomes everything. 20. The rise of "authentic" influencers who prove they're human. Verification switches from blue checks to "human-generated" badges as audiences crave real people. 21. Personal AI agents start making most of your small financial decisions. Your money moves automatically between assets based on your spending patterns and market conditions. 22. The fastest path to wealth creation in 2025 is understanding which assets are mispriced because of AI. 23. The great talent flip. Top engineers stop optimizing models and start building products. When AI can write good code, the most valuable skill becomes knowing what to build, not how to build it. These engineers become harder and harder to retain. 24. The "AI copycat economy" creates a new type of competition. Products get cloned instantly, but instead of killing companies, it creates weird new dynamics where being copied actually increases value. Original creators become more like "product DJs" - their value is in the mix and timing, not the individual features. 25. The first "AI-native" social network explodes. Not another chatbot - think dynamically generated worlds that adapt to each user's interests in real-time. 26. Private investors start investing in solopreneurs/mini teams + agent networks instead of traditional companies. The best returns come from owning pieces of AI agent networks that generate value 24/7, not traditional startups. The definition of "company" fundamentally changes. 27. SaaS and agents merge completely in 2025. Every SaaS product becomes an agent platform, and every agent platform builds SaaS features. The ones that don't adapt die or get bought for pennies.

Multimodal models often employ primarily numerical data. Today we measure their performance on understanding time-series data (a fall detection task) when presented visually as plots compared to numerically, finding significant improvements. Learn more at: