Courtland Allen

Starting a business, but don't know what your mission is? How about: making your life better. Working on the things you enjoy, at the pace you set, from wherever you want, with the people you love, so you can be happy… these are perfectly fine reasons to start something.

@brooksy4503 @CodingAtSheetz @openclaw Pretty much every company with a trademark needs to do this, otherwise you risk losing your trademark

yeah I suppose If your name is Claude they own you and can make you change your name. MacDonalds did the same thing back in the day. I dislike companies that are heavy handed. They could have encouraged this project and changed the name. Or bought it out. Injected money into. No, just ip violation.

🦞 BIG NEWS: We've molted! Clawdbot → Moltbot Clawd → Molty Same lobster soul, new shell. Anthropic asked us to change our name (trademark stuff), and honestly? "Molt" fits perfectly - it's what lobsters do to grow. New handle: @openclaw Same mission: AI that actually does things.

@PageX61612 @Steve_Yegge Pfft I lurk every day

@csallen @Steve_Yegge Bro you are not using this account anymore just sell it to me please 🥺😭🙏🏽

I was working outside of Starbucks at the mall, and this guy on his laptop next to me started chatting to me about AI coding, and it turned out to be the one and only @Steve_Yegge!

@aiktp_com Dm'd you!

@StockDipsAI Dm'd you!

@csallen 🙋‍♂️

His net worth comes from calculating the % of the shares he owns in his company * the last price someone paid for a share. He is not literally hoarding $150B in a bunker somewhere. It's not like he could dissolve his company and magically turn it into $150B in cash that was just being hoarded or hidden away somewhere. At best he could start selling off shares, which means there would be buyers for those shares, which means it would just be a trade of cash for shares. No additional cash would be "unearthed" that doesn't already exist today.

@MemoryMedieval @WillManidis oh yeah. agreed. would be amazing if he would participate in something other than hording his wealth

in 20 years, as a hobby, Rockefeller moved a medieval cloister from France to Manhattan and restored 400 buildings 1:1 at our nations first capitol in Williamsburg, VA. it is impossible to imagine our current elites turning this much ambition towards history today.

@TMoldwin @robinhanson By the time the old cohort dies, the young cohort will be getting old, and it's not a given they'll still want the same things they wanted in their youth

@robinhanson Isn't this necessarily the case because the older cohort dies out before the younger cohort?

I asked ChatGPT to collect a sample of 20 conflicts (that happened at least 50yr ago) between generations over cultural changes, & say in how many the young cohort won over the old in the long run. Score: Young 18, Old 1, Mixed 1.

@tysonmaly Accurate

@csallen Guessing this is a Claude meme?

One of the best things coding agents can do for you is explain things: - an undocumented codebase - all the changes it just made - an unfamiliar SDK or API - a code execution path from start to finish It's fast, accurate, and super helpful. My favorite prompt: "Teach me how XYZ works. Use concentric circles of explanation that go from broad to specific."

@mcpjams @smthomas3 @mastra @ai Did you ever get this working?

@smthomas3 @mastra @ai I'm surprised you were able to get it to work with @mastra/core 0.13. My understanding is that it still uses Vercel AI SDK 4, but it seems like there's no backwards compatibility for gpt-5 in Vercel AI SDK 4. github.com/vercel/ai/issu…

want to build an agent using gpt-5? use @mastra and make this one line change that's it... full walkthrough below 👇

@smthomas3 @calcsam @InternBerry @mastra No luck streaming with "gpt-5" as my model, even with temperature set to 1. It does work with "gpt-5-chat-latest" as the model, even without the temperature trick. But then stops working when I try to add tool use.

@csallen @calcsam @InternBerry @mastra try passing temperature=1 when calling .stream() GPT-5 doesn’t have temperature so it’s a temporary workaround (fix will be released soon)

@mastra is what LangGraph was supposed to become

@smthomas3 @calcsam @InternBerry @mastra Does this work for streaming for you, or only agent.generate()? Streaming just generates nothing when I try it with gpt-5

@csallen @calcsam @InternBerry @mastra x.com/smthomas3/stat…

@calcsam @InternBerry @mastra Also no gpt-5 support is surprising! And painful! Hard to justify build on Mastra if it'll mean being days/weeks behind on using newer models that get released

Main pain points for me: - Docs feel out of date sometimes? Which is a shame because the MCP doc server is so damn useful, but becomes useless if docs aren't accurate - The changelog updates don't seem to correspond to or even mention package versions, which is confusing - Overall Mastra and its subsystems (memory etc) are abstractions over underlying LLM APIs and their message chains, and at times I find myself wishing the docs would refer to these things instead of treating them like magic or like a black box. Generally, anything you can do to prevent me from having to actually dig into the code of the library to find out how things work is helpful.

I was wrong. Welcome back, Claude Code 🙏

@texjer GPT-5

@csallen Oh no what happen

RIP Claude Code, 2025-2025 ⚰️

you think that completing a quest will make you happy but as soon as you get the reward you find yourself in a deep depression until you start a new quest. it's not the reward you want, you just want to be on an actual quest

@octo_data @levelsio The most AI reply I've ever seen

@levelsio Time zones are about to be mere suggestions! At 27,000 km/h, you could have breakfast in New York, lunch in Tokyo and still make it back for dinner in LA. Sign me up—but do frequent flyer miles even make sense in rocket world? 🌍🚀

SpaceX Earth to Earth passenger flights are now pending approval 27,000km/h Hong Kong to Singapore: 22 min LA to Toronto: 24 min Bangkok to Dubai: 27 min Tokyo to Singapore: 28min London to Dubai: 29min London to New York: 29min New York to Paris: 30min Honolulu to Tokyo: 30 min Tokyo to Delhi: 30 min Melbourne to Singapore: 30min Sydney to Singapore: 31min Sydney to Dubai: 40min LA to Johannesburg: 50min Most long distance trips will be less than 30 minutes!