{"slug": "10-new-businesses-unlocked-by-thinking-machines", "title": "10 new businesses unlocked by Thinking Machines", "summary": "Thinking Machines, led by Mira Murati, released Inkling, a 975B open-weights model that companies can fine-tune via its Tinker platform without building GPU infrastructure, enabling startups to own specialized AI models. Bridgewater has already used the platform to train a model on financial knowledge, reportedly outperforming proprietary models on financial reasoning. This accessibility allows founders to create businesses previously reserved for well-funded AI labs, offering better margins, control, and defensibility.", "body_md": "# 10 new businesses unlocked by Thinking Machines\n\n### 10 Businesses Unlocked by Training Your Own AI Model\n\nMost AI products being launched today rent intelligence from the same handful of GPT and Claude APIs. That makes them fast to build, but it also leaves them with **similar capabilities, permanently variable costs and limited control** over the technology at the centre of their product.\n\nThis week, **Mira Murati’s Thinking Machines released Inkling, a 975B open-weights model that companies can fine-tune through its Tinker platform without building their own GPU infrastructur**e. Bridgewater has already used the platform to train an open model on its financial knowledge and reportedly outperform proprietary models on financial reasoning.\n\nFine-tuning is becoming accessible enough for startups to own the intelligence they sell. **Founders can now build specialized models, license them, deploy them inside enterprise environments and create businesses that were previously reserved for well-funded AI labs**.\n\nIn this issue, I break down:\n\n**How owning the model can give startups better margins, control and defensibility** Why o\n\n**pen weights turn AI intelligence into an asset your company can own****How startups can now train specialized models without becoming AI labs****Why Inkling’s capabilities matter for products, customers and unit economics****10 business ideas unlocked by accessible fine-tuning** How to decide whether owning a model should become your startup’s advantage\n\nHey, welcome to AI Market Fit!\n\nEvery week you’ll get two things, fast:\n\nWhat’s happening now (and where the money’s headed).\n\nWho’s already building it (so you can invest, partner, or steal their playbook).\n\nHowever, most of the content from this newsletter will be only accessible to the paid subscribers.\n\nIf you are serious about building or investing in AI, you can subscribe here:\n\nHere’s a good place to start:\n\n↳[Alex Karp’s Playbook for Which AI Companies Survive](https://www.theaiopportunities.com/p/alex-karps-playbook-for-which-ai)\n\n↳[19 great new details about Cursor’s rise](https://www.theaiopportunities.com/p/19-great-new-details-about-cursors)\n\n↳[Perplexity’s CEO 2026 AI Pitch](https://www.theaiopportunities.com/p/perplexitys-ceo-2026-ai-pitch)\n\n↳[The First 48 Hours of Claude Fable 5](https://www.theaiopportunities.com/p/the-first-48-hours-of-claude-fable)\n\nTry premium free for 7 days. Or get 50% off this week only.", "url": "https://wpnews.pro/news/10-new-businesses-unlocked-by-thinking-machines", "canonical_source": "https://www.theaiopportunities.com/p/you-no-longer-need-openai", "published_at": "2026-07-19 14:00:28+00:00", "updated_at": "2026-08-16 15:41:59.107496+00:00", "lang": "en", "topics": ["artificial-intelligence", "large-language-models", "ai-products", "ai-infrastructure", "ai-startups"], "entities": ["Thinking Machines", "Mira Murati", "Inkling", "Tinker", "Bridgewater"], "alternates": {"html": "https://wpnews.pro/news/10-new-businesses-unlocked-by-thinking-machines", "markdown": "https://wpnews.pro/news/10-new-businesses-unlocked-by-thinking-machines.md", "text": "https://wpnews.pro/news/10-new-businesses-unlocked-by-thinking-machines.txt", "jsonld": "https://wpnews.pro/news/10-new-businesses-unlocked-by-thinking-machines.jsonld"}}