RunLLM

🎬 Hot Take: @cgwu0530, Co-founder & CTO of RunLLM “At the end of the day, open source models will win.” Chenggang makes the case: the future of AI depends on flexibility, transparency, and interoperability — and that’s where open source has the edge. 👀 Watch the full…

🎬 Hot Take: @lateinteraction, MIT Professor & @DSPyOSS Creator & Project Lead “The level of AI we're approaching is a rebirth of computing.” From assembly to C to Python, we used to see big leaps in how we built software. Then… not much. Now AI is accelerating things again…

🎬 Hot Take: @shirshanka, Co-founder & CTO of @DataHubCloud “The one thing I’m sick of is everyone saying, ‘Can’t we just solve this with generative AI?’” Shirshanka makes it clear: GenAI isn’t a magic wand. The best results come from a blended approach — grounded in a real…

We’re launching @RunLLM v2 next week (👀 stay tuned!), and we honestly feel like the product might not be ready. That’s just life in a startup. There’s always more polish you can do, and you’re probably right. But getting something exactly right before you know it’s the…

There's a lot of discussion about the new UX for AI, but everyone seems to be thinking about this in terms of web & mobile changed UX. @profjoeyg and I have a different perspective. A good UX for AI isn't about hiding information behind menus — it's about where and how you work.…

🎬 Hot Take: Joe Hellerstein, Professor at UC Berkeley “The more you trust AI, the less you should trust it.” As your expectations rise, so does your risk — unless the system around the model is designed to earn that trust. 👀 Watch the full mini-documentary here!…

🔥 First we chased scaling laws. ⚡ Then we tried inference scaling. 😅 Now @profjoeyg says… maybe we missed the point. The real unlock? It’s not about size. It’s about how you use it. (We’re still talking about LLMs.) 🎬 Full mini-documentary: youtu.be/MyjT2nBpbE8 #AI…

💸Got AI budget? 80% of AI projects fail. Why? Flashy demos. Poor fit. Zero follow-through. This post breaks down what actually works — and how support and engineering teams can pilot AI tools that deliver real outcomes, fast. If you’re evaluating solutions (or already…

🎬 AI Hot Take: @joe_hellerstein, Professor at UC Berkeley “LLMs are guessing machines. Pretty amazing ones — but they’re still guessing.” That’s why you need more than just the model. Neuro-symbolic AI — backed by policies and correctness checks — is where things get…

🗓️ ICYMI. Run of the Week | July 14-20, 2025 RunLLM insights, updates, and shares from the past week—all in one place! 📌 7/29 Launch Sneak Peek: “Agent’ is the Buzzword of the Year. 🚀 We’re launching RunLLM v2 next Tuesday, July 29. Details to come, but here are some included…

🐳 RunLLM CEO @Vikram argues that if foundation models like #ChatGPT, #Claude & #Gemini were to plateau today, we'd still have 10+ years of innovation ahead of us — like having the whole ocean floor left to discover. What do you think? 🎬 Full mini-documentary here:…

🎬 AI Hot Take 🔥 "What makes a bad AI product is a product that's too focused on AI." - UC Berkeley Professor @profjoeyg 👀 Watch the full mini-documentary here! youtu.be/MyjT2nBpbE8

🎬 AI Hot Take: @dpatil, Former U.S. Chief Data Scientist “The biggest challenge that we face is access to data. AI is a roof on top of a house. Those walls, that foundation have to be rock solid. Then you can integrate AI systems into it.” The message is simple: before you…

We predicted earlier this year that there would be an acquisition spree in AI applications, and I think it's pretty safe to say that we're here. And it's crazier than we expected. Today's post covers how to think about these acquisitions and why they seem so weird. 👇