
End of 2025: Opus 4.5 & GPT 5.2, coding solved End of 2026: WoW Forever, Ocarina of Time Remake, GTA VI, gaming solved What’s next?

Developer Relations × API Design × AI Agents | @temporalio | Ex-@blocks/@square | I help teams build platforms developers love Tweets are my own.

End of 2025: Opus 4.5 & GPT 5.2, coding solved End of 2026: WoW Forever, Ocarina of Time Remake, GTA VI, gaming solved What’s next?

Friday in October’s Product Edition: AI Framework Integrations. New framework. Same reliable execution. Temporal’s AI integrations are now generally available. Build agents with Google ADK, OpenAI Agents SDK, AWS Strands, LangGraph, and more—without rebuilding how they recover from failures and resume their work. Explore the final October Product Edition story: temporal.io/blog/ai-framewor…

I’m a recent convert but also believe, @temporalio is the way 🙏
I noticed an enterprise pattern this year, where most companies are still struggling with AI in the SDLC, but some are succeeding with rolling out AI, even open-weight models, in Customer Service departments. The world-class shops use @temporalio and classic data science evals. I call this pattern the Counting House, and wrote a post on it: yegge.ai/essays/the-counting… The post is way too long to read in one sitting so don't bother. It's more of a research project. There are some jokes buried in there, but few and far between. Just skim it, look at the pictures, and you'll get the basic idea of how it works. I think it's pretty neat, and we'll see a lot of them roll out next year. Any VCs who've been following me know I haven't believed in any ideas enough this year to want funding. But this one might be different.

I noticed an enterprise pattern this year, where most companies are still struggling with AI in the SDLC, but some are succeeding with rolling out AI, even open-weight models, in Customer Service departments. The world-class shops use @temporalio and classic data science evals. I call this pattern the Counting House, and wrote a post on it: yegge.ai/essays/the-counting… The post is way too long to read in one sitting so don't bother. It's more of a research project. There are some jokes buried in there, but few and far between. Just skim it, look at the pictures, and you'll get the basic idea of how it works. I think it's pretty neat, and we'll see a lot of them roll out next year. Any VCs who've been following me know I haven't believed in any ideas enough this year to want funding. But this one might be different.

Speaker submissions are officially open for Replay 2027! Replay, our annual conference, is coming back to San Francisco May 17-19, 2027. We'd love to invite you to share hard-earned wisdom from putting the next generation of software into production. Submit a session: pulse.ly/mkxirmy3xq


I am SO tempted to build a Magic the Gathering benchmark for LLMs. I tested Sol 6.1 vs Opus 5.5 on each building a Brawl (EDH/Commander) deck. Sol 6.1 actually 1-shot it better conforming to target. Opus 5.5 got there but had 1 illegal card for the format and was off a little

I'm hiring a Social Media Manager for the Agentic AI Foundation (@AgenticAIFdn). We have a LOT of exciting stuff happening around the world and are looking for someone who can capture that motion socially. If you enjoy a fast paced environment where you can be fun and creative, I'd love to chat! The role is fully remote with the opportunity for occasional travel. Please share with the most creative social media strategists you know. careers.linuxfoundation.org/…

@thsottiaux what am I doing wrong? I'm just using the Codex desktop app. Why am I not thirsty for tokens like everyone else?

I have a reset that expires on the 3rd and another that expires on the 4th. I have been running Astra Ultra with Fast enabled and I've still only burned like ~30% of my availability. I fear I'm nmgi

Skeuomorphism strikes again. But I suspect it’s more likely to lure users into “texting” with their dot more like a friend or colleague, and have the experience feel less like our usual LLM chat
why do dots have read receipts. like why are they just straight up copying the imessage UI without any thought to whether any of these details even make sense it's an on-demand ai chatbot ofc it's going to read my message as soon as I send it. what tf else could it be doing

And don't you forget it

accidentally said you need to build better APIs instead of MCPs and they kicked me out of SF
accidentally said automated tests instead of evals and they kicked me out of SF

Nice case study from PayPal on WebMCP: - Reduced Agent Compute Resources: Approximately 4.6 times less costly in model-inference usage when completing checkout. - Faster E2E Checkout Completion: Approximately twice as fast at the median with WebMCP enabled. - A More Consistent Experience: Using three structured tool calls via WebMCP instead of more than 30 interface-interaction steps with the feature disabled.

I was not expecting things to go this way, but I think MCPs are better than CLIs for most integrations. The models have gotten much better at tool calling, we can defer tools & MCP is now stateless. If you need to compose/filter data, add params like query to your MCP tools.

I can’t stress enough how little an idea matters compared to the agency of the people executing the idea. I have had the privilege of knowing and sometimes even working with some of the most successful people (by various metrics). The difference between mediocre and excellent work and outcomes is predominantly one of agency. In practice this means: they dont wait for things to happen to them they go out and make things happen for them. They don’t wait for someone else to do something, for someone to teach them, for someone to give them the path, etc. They just go out and find a way to do it. I think the single biggest superpower these people have is the realization/belief that the world around them is completely mutable. Most everything that happens is because a person made it happen. I used to tell people to look around the room you’re sitting in. Look at everything. Every noun. It almost all exists because a person willed it into existence. Nothing is stopping you from doing the same. I see people online all the time dismissing someone else’s success because “I had that idea first” or whatever. I mean… yeah? If so then the difference is… you. So a bit of a self own whenever I hear that. Number one tip: act with agency.