An AI model can solve a difficult problem and still be impractical to use. It might spend too much time reading context, require every researcher to repeat the same expensive computation, or need a human hovering over every action. Capability is only part of the engineering problem. The machinery around it determines how much useful work actually gets done. Three September releases make this concrete. DeepSeek V4.1 Flash changes the economics of processing long histories. Google DeepMind's AlphaGenome Atlas makes billions of biological predictions available for reuse. Meta's Muse gives an agent a persistent computer and a controlled route into other applications. Together, they suggest that some of AI's most consequential progress is happening in how intelligence is organized and deployed....

' Hey, Linas here! Welcome back to a ' weekly free edition ' of my daily newsletter. Each day, I focus on 3 stories that are making a difference in the financial technology space. Coupled with things worth watching & most important money movements, it's the only newsletter you need for all things when Finance meets Tech. How to Use GPT-6 Astra, the OpenAI Model Jensen Huang Called AGI ' [8 GPT-6 Astra workflows, copy-ready prompts, and practical settings for founders, builders, investors, and entrepreneurs who want to build products, research investments, and get more value from AI] The Ultimate Guide to Grok Bot: The First Real AI Agent OS ' [The complete guide to building a persistent AI agent operating system with Grok Bot ' setup, working prompts, 8 end-to-end workflows, multi-agent teams, security, and the economics of AI teammates] GLM-5.3-Flash: Local AI Goes Multimodal ' [Why the first open multimodal AI model that rivals Claude Opus 4.8 and GPT-5.6, runs on a single Mac, and was served without NVIDIA chips resets the math for founders, builders, and investors]...

Imagine bringing a robot into your kitchen and saying, 'Help me clean up after dinner.' You have just compressed a remarkable amount of engineering into six words. The robot must distinguish leftovers from rubbish, discover where plates belong, and work out why a drawer refuses to close. Eventually, someone will hand it a wineglass. Everyone will suddenly become very interested in the quality of its training data. That kitchen captures the promise of a ChatGPT moment for robotics. We would be able to give a machine useful new work through conversation and examples, with sufficiently little setup that teaching it becomes an ordinary activity. Generalist robot models are making this prospect more credible. The remaining distance involves learning, control, and the economics of getting a machine to work somewhere new. ChatGPT made broad capability easy to explore. You could ask for a poem, then a program, then an explanation, and discover the range yourself. Robotics needs an equally persuasive encounter with versatility. The complication is that its answers have mass....

If your pitch still leads with 'we use AI,' you're describing infrastructure, not a business. Consider that 97% of the products nominated for this year's Products That Count Product Awards are deeply integrated with AI, highlighting the extent to which the days of AI as a differentiator are officially gone. We analyzed Crunchbase data on 576 venture-backed, AI B2B companies that raised $50 million-plus rounds since the start of 2025 using Hamilton Helmer's 7 Powers framework and layered in insights from Products That Count's 600,000-plus product leader community. Counter-positioning is what happens when a newcomer builds a business model so structurally different that the incumbent can't copy it without destroying their own economics. Netflix versus Blockbuster is the canonical example. Blockbuster could have matched the subscription model, but doing so would have gutted late-fee revenue, which kept their stores alive. So they didn't, until it was too late. In the AI era, this power is rare and underutilized. Only 5% of companies in our dataset leverage counter-positioning. Investors price that scarcity at a median enterprise value of 5.3x per dollar raised ' the highest multiple of any power in the analysis....