Living Action Role Player

For the record, my name is not Arjun

@Teknium @bradmillscan Good update.

@bradmillscan From the openai staff I talked to today.

Hermes doing the exact same nonsense as OpenClaw did. Gateway full of errors. I'm barely doing anything. All I did was run hermes setup and add ChatGPT oauth as a model provider. Switched back to Grok and asked the agent to diagnose. and it's acting very incapable & erroring as well. Is this the UX most people have when switching to Hermes? Just as terrible as OpenClaw so far.

@marclou what counts as shipping?

The first crewmate just shipped their startup 🏴‍☠️ He built an AI food tracker, but with a twist: It’s not for fitness people. It’s for parents to track their baby’s first foods and detect possible allergies or reactions. Fair winds, Captain Nil 🫡

@thereallisaann what do you think the underlying premise of onlyfans is?

@steipete why not use cloudflare for everything?

What do people use for SSO/SCIM/Endpoint Security in 2026. As we're hiring people for the OpenClaw Foundation, I gotta level up.

@cremieuxrecueil that's because you added "make no mistakes" to the prompt. smart man.

Absolutely incredible: I put Codex onto completing a fairly hard task that I never completed some years ago and it understood my vision and finished up like 99% of the work. It did this over the course of 48 hours, and it doesn't look like it made many mistakes.

@venturetwins Most people write a lot worse than AI. The author is deluding himself. Good writing will either become obsolete, or a rarer commodity; same branch for most of media.

Unfortunate news for those who care about the AI writing debate: we're nowhere near saturation. If you're on X, you're likely an early adopter. But new waves of people are discovering LLM writing every week and get one-shotted by Claudeslop ("everything I say sounds so profound!"). You can see it happen - first X, then LinkedIn, now Threads. We've diagnosed the disease but have to let it ravage these communities before it burns out 🫠

@DeryaTR_ it just does the thing.

That’s the moment of truth I personally reached after few days during early testing of Codex GPT-5.5. Indeed, I first create my agents.md (nowadays HTML) file & have it plan it or GPT-5.5 Pro create the prompts, also spawn some agents, then it just keeps on coding!

@mark_k They probably use it but switched after the 2 message limit. Even tech COOs are complaining about usage cost / ROI

I don't know anyone who's actually using Claude Code. Must be a normie thing? I guess it's popular with frontend monkeys. Anyway I think this too will pass.

@rudrank Agree 100%. Have been using the same workflows and grinding my limits lower much faster. It seems to burn lots of tokens and time doing basic things - git checkouts, deploys, etc. anyone sharing this feedback with @thsottiaux ?

Am I using Codex more, or does anyone also feel the weekly limits are going down faster? 2x promotion will be gone in less than a week; $200 plan weekly limits will be finished in 2 day

@Teknium Like way way way too much of this ;-)

@RileyRalmuto Played around with it - really trying to understand what it is. It's unclear how to go back to the same place over again. And if you can't get there, is it... just like a chat model?

i wonder if anyone will believe me when i tell them i built all of this from a cabin in the woods, alone, eating apples, peanut butter and creatine for 6 months. and without writing a single line of code. ive been called crazy (i am crazy) ive sacrificed relationships, opportunities, hella brain cells ive burned through something like $25,000+ on API calls alone without making a penny (most of which on things youll never even see) my family is convinced none of it exists because i wont show them any of it yet (lol) and in this last stretch ive been reflecting a lot, and ive come to realize something about myself- some people channel grief into very destructive bullshit, and ive watched so many people i love spin from mismanaging it. ive realized that all of this - at least everything ive done since August 10, 2025 - has been my subconscious attempt to channel my own grief into something im proud of, even if in the end its just for me. ❦ mnemos

@theo Don't discount them. They're solving an insane infrastructure problem. Go slow to go fast, you know what I mean?

I'm not convinced that DeepMind ever actually figured out how reasoning works. The models have been useless since they introduced it.

@jackfriks You gonna help each other with user testing pre launch? If so in.

i made an app that feeds you to the sharks if you don't publicly launch your own product in 30 days. no more of this: "dude i just 100x'ed my workflow with this new AI model"... meanwhile... 0 projects launched 0 revenue 0 users 100 x 0 = still 0. it's time to go from 0 to 1. it's time to: SHlPORDIE.COM 🏴‍☠️. ship a new product every 30 days until one changes your life or... DIE, in the app, and get kicked from the community forever while being publicly humiliated. no refunds for those who fail to ship. custom trophies to be collected for those who succeed. if you DO ship, you also get to remain in a community of people who actually ship things and get users ++ revenue. sidenote: i'm really excited to see if this can be the push someone needs like how @marclou's shipfast project pushed me and is the entire reason i have a $35K MRR solo operated SaaS now and many other successful mobile apps GLHF, DON'T DIE, and KEEP GOING!! i've never taken a launch this legit so let's see how it goes :)

@CliftonSellers Paths unlikely to change in that time: 1. Selling (charm, EQ, people skills) 2. High End or Niche Services 3. Networks / Power / Access 4. Hobbies and Sports Pick 2 out of the 4 to develop She’ll need spiritual fortitude too so take her to church.

Parents, honest question for you all My kids (3 daughters) are 10+ years away from turning 18 and college is looking less and less like the path What’s the new plan?

@brexton @yitong Still recommend Evangelion as an adult? Anime always feels interesting to me but stunted developmentally, like Ayn Rand.

@yitong We ARE The only problem is that evangelion is so hard to understand/explain that there’s only like 10 people on earth you can have this conversation with

I have a whole talk about how we’re basically episode by episode speed running evangelion rn. Hard to unsee once you realize

@Teknium Thanks. All responses take 2+ minutes some take 15. Baseline just sends the same sample questions every day. Can send the stats but even to answer ‘hi’ it’s slow. Haven’t dug deep in the docs but any diagnostics you recommend?

@SecretArjun Sorry but what is it doing for me to understand?

@Teknium 30 day update: Hermes is 75% of workflows (from 0% vs. Openclaw). Hermes rocks, but the trackbaseline.com round trips are 2.5 minutes average (GPT 5.5 high, non-fast). Any speed optimization tips? Or is GPT 5.5 the wrong model? Local; Mac Studio M3 96GB

the algorithm changes don't bother me much. i'm just here for a good time. when it stops being entertaining, i'll bounce. but it's still hilarious that i can post whatever thought crosses my mind and actual important people are forced to scroll past it. more so soon.

Another great reason to build in public is: the more you ship, the more you realize what you used to call a finished product is not even close to what you are actually capable of. Not. even. close.

App building 101: - Find a gap nobody local is solving (Market) - Live inside the problem yourself (Research) - Ship something ugly that works (Product) - Put it where the right people already are (Distribution) - Sell the outcome, not the feature ($) Stop waiting until it's perfect.

@steipete Make sure you're logged in though :-)

Pro Tip on Adversarial Reviews: Use GPT-5.5 Pro it's by far the best model for this. Use @steipete's oracle CLI and log in. You're welcome.