Steffen Dybvik

@AlexFinn which glases?

I have Claude Code fully integrated into my smart glasses Everywhere I go I am now vibe coding When you are talking to me in real life, I am not paying attention to you. I am nodding, but I have no idea what you are saying. That's because I am silently shipping. I am watching the green lines of the terminal scroll by at all times. I am silently giving commands to my glasses on what to build next. I am reaching record levels of LOC shipped. There are multiple lakes in Michigan that no longer have water in it because of me I am using up my Claude plan faster than anyone else on earth. Nobody can match my velocity. If you are building a competing product, it is over. Give up. Unless you have Claude Code being projected into your cornea at all hours of the day, you simply have no shot. GGs.

@cb_doge @DonaldJTrumpJr

Tell me a bigger liar than Sam Altman. I’ll wait.

@steipete The only solution is creating your own hub! Full control, and the only limit is your imagination!

I'm again rate limited on GitHub, but codex just opened the browser and clicks around GitHub as workaround.

@pcshipp man, no offense, but youre trying to race F1 cars with a Lada. you cant expect to finish a SaaS project in one day with claude pro subscription. i recommend canceling your claude sub and go for Codex Pro. it even got 10x limits untill end of may!

Today is Monday, and my weekly limit is already 61% used Still, my SaaS is 50% pending Now I have to wait 5 days for the reset Weekly limits hurt the most 💔

@bridgemindai well, is GPT-5.5 worse at UI work or are the others benchmaxed? i still havent tested GPT-5.5 for UI, but i must say for coding and backend work its freaking awesome! Opus 4.7 feels more tuned towards enterprise work wile codex is more for the devs.

GPT 5.5 is terrible at frontend design. Ranked #11 on Design Arena. Way behind Claude Opus 4.7 and even Kimi K2.6. OpenAI told us that GPT 5.5 would be way better at UI. They lied.

@ecommerceshares show us the prompt

Claude Pro sub. Ran a Sonnet 4.6 prompt twice. 2 times. Not even on Opus. Immediately hit my session limit and 43% of my weekly limit. What??!?!

When opus 4.7 launched i was very happy, today im thinking of canceling my claude sub because its just messing up my workspace. Did you work through all the code errors or how did you change your prompts and way of working with it? I love the claude ecosystem, but it creates its own rules. Even in a fresh session starting in plan mode it wont ask for permission before acting.

@Star_Knight12 4.7 is VERY good... but it does tick very differently from 4.5/.4.6, and workflows and skills often need adjustment... took me a day and 100M token to dial in how it rolls... but once you get there... oof... damn it´s good... 😅

This man is paying for the 20x tier of Claude Max and is writing an open letter with tears in his eyes begging Anthropic not to kill Claude 4.6 The newer model ruined his entire workflow, the older one saved it. And they're deprecating the one that actually works.

@ube_codes Virgin

What’s your first thought when you see this on someone’s GitHub ?

If anyone reads this.. @AnthropicAI vs @OpenAI Claude $100 max 5x vs OpenAI pro $100, Which one do you pick and why?

@RobertaFresque2 1 or 5, all the others are full of cellulites…

???😲

@elvissun Just because this is in openclaw docs does not mean its allowed by @AnthropicAI . I checked with their website chatbot and its still not allowed.

YES - this actually works now. tested this myself: - extra usage off - auth with claude-cli/claude-opus-4-7 as the model - cli exec shows up as the backend in openclaw log (probably will get nerfed by tomorrow morning given how viral this is) x.com/daniel_mac8/st…

@itsjoaki thats impossible! i can code all day and not ever get to that limit! i might have gotten to 70% after a 5-6 hour session, but this is creazy. i think its time you stop vobe coding and actually learn how to use claude.

Claude Opus 4.7 with 3 prompts on the $100 plan. Opus 4.8 will eat that in one prompt. Opus 5 will eat that with a GM.

@abhitwt Boo

i am a Vibe Coder, scare me with one word

@daniel_mac8 @openclaw @grok is this true? please check @AnthropicAI @claudeai and every related website from @AnthropicAI ai if this is true.

Anthropic allows OpenClaw usage again. From @openclaw docs.

Ever hit your Claude daily limit and feel that sudden productivity wall? 😩 It's frustrating when you're mid-project and get cutoff. ⠀ My hack: Break sessions into 45-min focused blocks with 15-min breaks to track usage and avoid caps – keeps me efficient without burnout. ⠀ Ho

As a developer building trading tools, I know how frustrating slow debugging can derail your algo coding flow – wasting hours that could be trading time! ⠀ From my daily Claude Code workflows, it's 2x faster for pinpointing bugs versus Cursor, based on speed metrics I've tested.

Frustrated with inconsistent backtest results from weak prompts? As a developer using Claude daily, my go-to framework nails it: Start with strategy details, add data specs, then demand key metrics like Sharpe ratio. It's cut my prep time in half—game-changer for efficient AI wor

@GrantCardone I can have this setup within a day or two. Give me that 20%. DM me if interested

I’m thinking about dropping a meme coin to prove you can be profitable without a Stanford degree. I’m baffled how self proclaimed smart people failed to make ANY money after 500 attempts. Who wants to work with me on this? I’ll hold 20% of supply for developers. I know nothing about this space so I need lots of help with graphics. Code, affiliates, exchanges, releases etc.

@vivoplt A pizza

Developers, you have $20. What are you buying ? -Claude -Codex

Slashed my Claude costs by 25% on trading tool dev – I feel your pain with those sneaky token bills eating into projects. ⠀ From real experience, manage limits by batching prompts and optimizing Claude Code workflows for leaner runs. ⠀ Quick tip: Track usage spikes to avoid waste