Today, private market revenue multiples are nearly 10x those in the public markets (for infra companies). What does the gap mean for founders? Here’s a few takes from Sai Senthilkumar and Jordan Segall: ➡️  The private market ARR multiples should get you excited, but the public market

Lots of conflicting takes about gpt-oss (yay open-source in the spotlight)! We’re powering the official @openai demo gpt-oss.com with HF inference providers thanks to Fireworks AI, Cerebras, Groq Inc and Together AI so we have a front-row seat of

BOOOOM! You can now run OpenAI gpt-oss 20B natively in Colaboratory T4 for FREE! 🔥 Powered by Transformers ⚡ The setup takes a bit since everything is bleeding edge, but once done it should work as expected Link to our cookbook in comments 👇

What OpenAI's $500 billion valuation means for stock investors trib.al/3gKCLyt

What a week ! techcrunch.com/2025/08/07/ope…

OpenAI's GPT 5 System card (a 59 page doc) Key Takawsys 👇 🚦 The model router The chat system holds two main models. A quick one answers easy stuff. A deeper one opens only when the router sees tricky logic or the user types “think hard”. The router trains on real clicks,

Lots of conversation on what the future of software looks like in the enterprise. Here’s how I think it plays out. For deterministic workflows where the cost of getting something wrong is high, enterprises will have a tendency to pick core platforms for their most common,

💰💰💰One-third of global VC funding in Q2 -$91B -went into AI In Mary Ann Azevedo’s Crunchbase News piece VCs including Menlo’s mmurph weigh in on: - where AI is delivering - where markets are overcrowded, and - which sectors could define the next wave mnlo.vc/41yYz13

It's clear that we're on a trajectory right now of AI models continuing to improve in capability across math, reasoning, logic, tool calling, and various domain-specific tasks that will get better as more training data continues to get generated. While there will be debates

« Work has evolved from linear processes like Waterfall to iterative teamwork with Agile, and now to rapid, personalized cycles between humans and AI. » open.substack.com/pub/climbingth…

Introducing Jan-v1: 4B model for web search, an open-source alternative to Perplexity Pro. In our evals, Jan v1 delivers 91% SimpleQA accuracy, slightly outperforming Perplexity Pro while running fully locally. Use cases: - Web search - Deep Research Built on the new version

This is actually a pretty surprising and something that should lead companies to change how they are thinking about hosting. Model performance for the open weights GPT model vary by meaningful amounts depending on who is hosting it, with Azure & AWS being low. Worth watching.

Updates to ChatGPT: You can now choose between “Auto”, “Fast”, and “Thinking” for GPT-5. Most users will want Auto, but the additional control will be useful for some people. Rate limits are now 3,000 messages/week with GPT-5 Thinking, and then extra capacity on GPT-5 Thinking

The Human–AI Loop isn’t just theory — it’s already reshaping how tech teams work. My latest Substack explores this open.substack.com/pub/climbingth… This Scott Hanselman 🌮 episode with rodric from Postman shows it in action. open.spotify.com/episode/433CAD…

What AI tools should you be using to be more productive? It's my job to test new products - here's the 10 I'm sticking with ⬇️ 0:00 Intro to AI productivity 0:20 Comet – AI browser 1:09 Julius – Chat with your data 2:08 Happenstance – AI LinkedIn 3:10 Granola – Bot-free AI

Buildathon: The Rapid Engineering Competition livestreams this Saturday, August 16. Top developers will compete to build 5+ products in a single day using AI coding assistants – projects that traditionally took weeks. Watch live as they advance through semifinals and finals, and

we’ve only scratched the surface between what current models and the right app layer can achieve 👀

They got GPT-5 to perform better on security tasks than OpenAI’s own numbers. How? Well, that's what agent engineering is about. It's creating an environment and a harness that is best suited for the task and use-case. This is a very specialized domain knowledge that it's

New podcast episode: Dylan Patel's hot takes on NVIDIA, OpenAI, Intel, Meta, Apple, and more. (1:11) Reactions to GPT-5 (7:34) The Economics of AI (12:30) Advice for Sam Altman (14:18) NVIDIA’s Growth (21:27) Threats to NVIDIA (26:09) The Silicon Startup Boom (45:28) Data