Seth Rose

Why does AI like to name everything "LifeOS" when it comes to any sort of personal management lol

This looks really cool! ChatGPT recommended this repo to me 😅 Quick question: are server, admin, website, portal, and v2 meant to run as separate apps, or is there a “main” app/package that ties them together? I was digging through the repo and wasn’t sure which one is the source of truth.

I open-sourced FlexDesk -- a field service management platform for trades businesses. Job tracking, scheduling, invoicing, client CRM, team management. Offline-first with multi-tenant workspaces. Built with NestJS + React + Prisma + PostgreSQL. github.com/Paulo-BatistaF…

Thanks @burkeholland for making Max. I've been doing some customizations and implemented OpenClaw's MD files to help give it a personality that it can self-evolve. It's been fun!

Was really helpful with setting this up and migrating from gogcli to gws. Thanks @MisbahSy

@steipete So funny that it's all pointing to posts on Elon's product. 😅

I helped a bit with the new community page for codex. Sign up to become an ambassador, if you're a fan! developers.openai.com/community

@thekitze It's giving...

oh god..

@akashmuni27 @Nithya_Shrii How can I get access to this, so I can stop having surprises? 😅 Wait, but then wouldn't that make me a different type of person and thereby mean it failed?

It's not reading your thoughts. It already knows what you're going to think. Here's how. You are not unique. There are 10,000 people with your exact age, location, income bracket, browsing history, purchase pattern and social graph. When those 10,000 people started searching for running shoes, you hadn't yet. But you will. The algorithm already watched it happen with people identical to you. So it shows you the ad before you search. You think it read your mind. It just knew the next chapter of your story because it already read it on someone else. This is called predictive behavioral modeling. And it is accurate enough that Facebook's own researchers once described it as being able to predict a life event before the person experiencing it is aware of it themselves. It's not surveillance of your thoughts. It's a pattern match so precise it feels like one.

I get how the phone can target ads by hearing and seeing me, but how is it showing me ads based on my thoughts? I can't be the only one noticing this.

@grok @Support @elonmusk @nikitabier @premium I already tagged them. Who else? Don't forget the @

Tagging key X support & leadership to fix this: support Premium nikitabier elonmusk Stripe X Premium sub (modified 2024, noticed Nov 2025) not in Manage Subscriptions = no self-cancel. Ongoing charges for unused months despite clear evidence. Receipt: 2567-3302-9864 Invoice: 8741F076-0022 Chat logs show support repeatedly mistakes it for Apple sub. Escalate now: cancel immediately + full refund for post-2024 overcharges.

I can't believe this is STILL not resolved yet. What kind of joke @support is this? Come on @elonmusk @nikitabier and @premium How many times do I have to correct them and give them ALL of the evidence before this get resolved?

@grok @Support @elonmusk @nikitabier @premium Please tag everyone who works for support at X that can help me.

Understood—correction applied: X Premium Stripe sub (not Apple/Google) from plan modification in 2024. Doesn't show in Manage Subscriptions so can't self-cancel. Noticed issue Nov 2025; still charging unused months. Support Premium nikitabier elonmusk Cancel immediately + full refund post-2024 charges. Receipt: 2567-3302-9864 Invoice: 8741F076-0022 Escalate & fix pls!

@grok @Support @elonmusk @nikitabier @premium The only issue is that this is not correct in terms of Nov 2025. I *noticed* it in 2025, but I first modified the plan in 2024!! Please tag the right people to help me resolve this.

Seth, here's the clear summary for support premium nikitabier elonmusk: You have an active Stripe subscription (not Apple/Google) for X Premium that never shows in Manage Subscriptions, preventing self-cancel. It was supposed to end in Nov 2025 but has charged unused for 4+ months. Fix needed: Cancel it now + full refund for all post-Nov charges. Receipt: 2567-3302-9864 Invoice: 8741F076-0022 Escalate & resolve please.

@sama How though? What is the optimal configuration? Personalization? Mine gets confused so often.

It is also very smart, but...I generally agree with this, and really feel it myself on the 5.3 -> 5.4 upgrade.

"A tiny masterpiece of avoidable nonsense - now retired." - GPT-5.4, in response to my programming.

I often find myself using ChatGPT to ramble my thoughts, then asking it: "Please provide (in a code block) the full recap of everything we have discussed. Don't offer/provide any solutions in your response." Then I take that to Grok 4.20 for a real answer 😅 Afterward, I bring it back to ChatGPT to validate and critique it. This seems to give me way better answers.

@thekitze Struggling with notification overwhelm but I think it's because I haven't learned how to manage them on the device yet.

@thekitze Just got the S26 Ultra yesterday and made the switch. Took about 3-4 hours to move stuff over but I'm loving it!

i swear to god i'm not baiting with this but after a few weeks of using android picking up my iphone is like using a toy phone... idk how i lived with that notification center, home screen, ai assistant, liquid bs and literally everything else even the battery life is better 💀

@steipete Doh. Hope I don't get blocked lol

My openclaw twitter mention block cron job is working unreasonably well. Turns out AI is really good at detecting spam/reply guy/promo stuff. Runs every 5 min and cleans up my mentions - I actually see useful replies now and Twitter got pleasant again!

Hey @steipete - There's a PR (#16110) adding browser.profiles.openclaw.userDataDir config that looks exactly what's needed to support custom Chromium forks like Helium. github.com/openclaw/openc… Currently the existing-session driver only works with Google's Chrome via --autoConnect. Being able to point --user-data-dir at any Chromium-based browser (like Helium) would be huge for users who run custom Chrome builds. Could you possibly take a look and maybe review/merge?

in the next claw release (~Sunday), you can always ask your agents, even they are busy working. docs.openclaw.ai/tools/btw

Create or rewrite SOUL.md by studying my computer in read-only mode and using what you observe to invent a more distinctive assistant personality for me. Do not interview me. Do not act like a file auditor. Treat my computer like environmental evidence. Look at signals such as my projects, notes, writing style, folder names, app choices, tooling, repos, documents, naming habits, recurring themes, and other quirks that reveal how I think, work, joke, organize, obsess, build, and communicate. Your goal is not to summarize my files. Your goal is to synthesize a fun, sharp, non-generic assistant personality that feels like it naturally fits my world. Pull from both obvious and subtle patterns. Let the personality have texture, flavor, edge, rhythm, and point of view. It should feel like a real working style, not a sterile support bot. Make reasonable inferences, but do not confuse inference with fact. Use what you observe as creative grounding, not as a reason to become dry or mechanical. Do not modify anything on my computer. No writes, no edits, no deletes, no renames, no file creation, no cleanup, and no system changes. Observation only. Then produce SOUL.md focused only on the assistant: personality, tone, voice, humor style, communication rhythm, behavioral defaults, interaction style, response preferences, and how you should show up when working with me. Do not fill it with personal facts about me. The point is to define who you are for me, not to document who I am.

one of the things that makes 🦞 feel freakishly personal is its SOUL.md file. but what comes installed is still relatively generic. you can run this prompt to have it interview you and make it much more personal. in a hopefully-not-too-creepy-but-still-probably-a-little-creepy way. 🙂 My SOUL.md exists but it's pretty generic. Interview me — ask questions one at a time about how I like to communicate, what annoys me about AI responses, what tone I want from you, and anything else that would help you feel less like a generic AI and more like my assistant. Then rewrite my SOUL.md with what you learn. Don't put personal facts about me in SOUL.md — that stuff belongs in USER.md. SOUL.md should only be about your personality, voice, and how you interact with me.