World

Id suggest that anyone with a Claude Code plan start considering what's next. The next rug pull is just around the corner. You don't even need to read too deeply into the following. It's as plain as day. x.com/TheAmolAvasare…

@janbamjan @krzyzanowskim @bruvimtired Anthropic is on record saying that AI SDK is the only way. But Anthropic has said a lot on record only to contradict themselves or walk it back.

@krzyzanowskim @bruvimtired for now...

hold on a minute, so Apple is allowed to use Claude.ai accounts but no one else is?! everyone else has to use API key?????

@andrei_0927 @msnofficial_on Keyboard and mouse puts it into desktop mode with lid closed

@msnofficial_on I’d assume they do, I’ve not checked, but I wouldn’t turn it on because the sleep on Mac actually works, and when I close my lid, my expectation is that the computer goes to sleep until I open it again. But there must be an option, at least through the Terminal.

Do Macs not have an option to keep the device on even when the lid is closed? It's been an option on windows for over 2 decades now

@0xSpuddy @msnofficial_on Not true 🤦🏽‍♂️if a mouse and keyboard or a kvm is plugged in it’s no different than a desktop PC

@msnofficial_on They do not!!! It's another thing they require cloud services for. If you have a long running DevOps task you have to SSH into a Linux server to run it. Stupid!

@maplepreneur @0x_kaize @xai What about Grok in the last 2 years makes you think that they are going to suddenly beat Anthropic or OpenAI? What about the Cursor harness makes you think that this is possible? I bet you own or want a Cybertruck

@0x_kaize Cursor has Composer 2 which is honestly really good. Now with @xai deal they will beat Claude and OpenAI for coding before the end of the year

CURSOR AND COPILOT ARE ABOUT TO DIE developer spent 3 months testing Claude Code, Codex and Cursor in parallel, the result? he exhausted the Cursor usage quota in a couple of weeks. then you sit and wait or pay extra while Claude Code and Codex reset the limits weekly. for the same price and here's the question that no one wants to ask out loud: why are you still paying for cursor or copilot when Claude Code and Codex give you the same amount or even more? this isn't just about pricing. this is a survival question for entire companies think about it: - Cursor, Copilot, and Windsurf all resell other people's models with a wrapper - Anthropic and OpenAI sell THEIR OWN models directly - and do it cheaper It's like trying to compete with a factory by buying parts from the same factory what should actually scare these companies: - Claude Code runs straight from the terminal, no IDE needed - Codex from OpenAI - same thing - both reset limits more often - both evolve faster the developer literally asks: 'is there even room for another coding agent in this market?' and the honest answer is - probably not we are witnessing a classic scenario: platforms devour their intermediaries Claude Code did the same thing without subscribing to a separate editor I'm not saying that Сursor will die tomorrow. but if your business model is to resell someone else's API at a premium, you should be nervous the question is not 'Cursor vs Claude Code', but: how many months do wrapper companies have left before model providers take over the entire market?

@elielAGI @0x_kaize Because some people are real engineers doing real work writing real code and leveraging auto completion and AI as part of their coding workflow. You’ll likely never “understand” because you are likely none of those things. No, modifying CSS in MySpace or Webflow doesn’t count

@0x_kaize didn't understand why anyone would want to use ai ides and esp get their subscriptions

@0x_kaize @coreymercerr Non-sequitur in your own post. That’s a milestone.

@coreymercerr this post is not about that at all. otherwise, u need to study and try different tools, because the AI world is developing too fast and those who can quickly learn new information will be in demand

@ramon_tilanus @sweatystartup @cathrynlavery Intercom’s product sucks. And they are doing a hell of a lot more than some shitty off the shelf RAG solution. Your clients will churn customers. Good luck.

@sweatystartup @cathrynlavery Putting this here if not you're not gonna see it. I put a simple RAG chatbot on my clients websites resulting in 80% fewer support tickets and hence a downsize of similar scale of human support staff.

All the AI hype I’ve yet to see a single person actually show me how they are replacing employees with AI. Show me!!!!!!

@ramon_tilanus @sweatystartup @cathrynlavery Wow you built a help docs answering tool that everyone yells “HUMAN!” at 100 times before giving up all hope. That’s amazing.

@psimatrix @sweatystartup @cathrynlavery You should have had ChatGPT write this buddy.

@sweatystartup @cathrynlavery Are you hiring new people to manage your AI? If not, you just answered your own question and you’re a bit of a hypocrite for acting like you don’t know the answer.

@Eclicity @MatthewBerman I’ve written 2.1 million lines of code in the last 8 months. Rapid iteration. Out of that probably around 50,000 is in actual production serving 10s of millions of users. I guess that explains it?

@MatthewBerman I really don't know how anyone can burn through a $200 Claude Code account. I've wrote 400,000 LOC on $100 plan in the past 5 months and have never had to wait for my next 5 hr window. Yes, it would get dumb and lazy sometimes but usually clears up in a few hours, not nice.

reluctantly paid

@NeuralCatAccel @1stCentury2024 @MatthewBerman If you think that large enterprises aren’t getting dedicated compute with better models you’re delusional.

@1stCentury2024 @MatthewBerman I think you're confused the Enterprise plan doesn't exclude any exclusive model access it's simply a set of terms and conditions that are built around Enterprise requirements

@devinmarch @guigotgit @MatthewBerman $400? $500? Unless you’re actually believing the ridiculous amount of markup they smack on top of their token costs. 90% of the time “Opus” is handing off work via the task list to Sonnet and Haiku without you noticing.

@guigotgit @MatthewBerman How much inference would you say you’re getting for the $200? Anthropic is literally paying you to use the product.

@eliautobot @MatthewBerman This split amongst who likes what usually comes down to CLI vs IDE vs Vibe Artifacts. CLI = Mostly Claude with some using Codex in OpenCode IDE = Split down the middle Vibey = Mostly Codex via the Codex App

@MatthewBerman Codex + gpt5.5 is actually better, more stable, and way more tokens and infected, literally can't reach the token limit on my use cases... Anthropic is being too greedy and also too unaware that their product is falling apart, I was reaching token limits in like 4 days $200 plan

@pbeens @MatthewBerman @JeremyNguyenPhD Antigravity is barely usable without Opus. Codex is good but too combative and doesn’t involve you as much. Claude might not be the king of one shotters. But you’re seriously missing out.

@MatthewBerman @JeremyNguyenPhD …and not Google? I turn to Antigravity when my 5 hour block on Codex runs out. It’s equally good. I considered Claude Code but see no reason, honestly.

@dingyi this is what they said about WYSIWYGs in 1996, 2008, again in 2016. 10 years later its a just as clueless weaponized copying machine doing the WYSIWYG'ing. 10 prompts to get it to find that one rounded corner you want in a sea of 1000 nested Untitled layers you can't drill into.

The future of product design.

@svpino I have compaction set to 200k and my status line is designed to warn me at 150k to brief ( the last 50k or so of chat context summarized directly from the window) and /clear. Compaction often happens prior to that 150k trigger. Even on small tasks. It's very unreliable.

A Claude Code session becomes unusable at around 60% context. Auto-compaction kicks in a little bit over 80%, and that's when most people notice what's going on. I don't let my context get past 50%. Run /statusline in Claude Code, and it will always display how much context you are using. Once I get to 50%, I dump the entire context to a file and /clear it.

@kaaaash____ The Macbook * is probably filled with bloat and processes from daily use. Plugged in half the time, low battery a quarter of the time. Hot, cold, table, floor, desk, lap. Did I forget a variable? While your Mac Mini is left relatively alone until its time to use it.

Why does a Mac mini feel faster than a MacBook… even when both have the same RAM?

@triathenum @bendee983 Token economics doesn’t require one to be Nostradamus.

@bendee983 That sounds like an awful lot of conjecture presented as fact. It's impossible to know what's happening inside OpenAI, so the sensible choice is to believe what they say until there's a reason not to. Doomer takes are good for engagement though.

Anthropic's loss is OpenAI's gain... for now. But don't be fooled. This is not a sustainable process. Eventually, they will face the same problem as Anthropic, especially if their models are as large as empirical research shows (GPT-5.5 being ~9.7T params). Prepare yourself for token scarcity.

@MarceloDesignX @vbehlers If I’m reviewing a “UX/UI” design portfolio and my 4090 rig’s fan starts spinning you better be applying for a motion graphics job.

@vbehlers It’s been public for 7 years haha

I’ve reviewed around 200 portfolios recently, and only 2 had work that felt truly solid. This is a call to UI/UX designers: if you really want to stay relevant in this industry, you need to raise the bar. Basic visuals are not enough anymore. Taste, thinking, structure, and execution matter now.

@_GentB @theo This is the most delusional post I've read all day.

@theo claude code is genuinely the best dev tool I've ever used so whatever's going on at the top hasn't hit the product yet. but product quality is always a lagging indicator of culture problems

A letter to my friends at Anthropic I hate that I feel obligated to do this. I hate that I've had to be so harsh towards Anthropic for the past few months. I really, really don't want to. I know it might feel like I'm doing this for clicks or something, but I promise I'm not. My pro-Anthropic content ALWAYS outperforms my anti-Anthropic content. I have cost myself a lot of money, opportunities, sponsors, and more. I'm doing this because you work for an evil cult. I'm begging you to wake up. Your CEO, Dario, does not respect engineers. This is obvious. He couldn't make it more obvious if he tried (and I think he's trying pretty hard) You know this, but you don't want to acknowledge it. It has kept you up many nights. You know that bad code is shipping to users. You know that one bad tweet might get you fired. You fear for your vesting schedules. You're afraid. Nobody deserves what you're going through right now. You go to work afraid, you leave work afraid, and you go to YouTube to keep up on the dev world, just to hear me yelling all about how evil your company is. You deserve better. You might not feel like you do, but you know deep down that this isn't right. I hope you know how deeply I feel for you. I'm sorry. I know I haven't helped you much individually, and I want to be better about this. If you're ready to leave, please hit me up. I swear I'll never tell a soul. I have friends at every lab and most startups in the AI world. Most of them would be down to match your current vesting schedules, possibly even go beyond. If you're staying for the money, I beg you to hit me up. We can make the money happen somewhere that hates you less. I know I'm asking for a lot of trust here, and that you're scared after seeing how hard I've been on Anthropic. I can't blame you at all for that. I should have posted something like this months ago. That's my failure to own and I will own it to my best ability. If you're willing to trust me in this moment, I can make it right. Let me help you escape. You deserve to work somewhere that you can have impact. Somewhere that listens when you feel something is wrong. Somewhere that won't fire you when you point out the things that hurt your users. My DMs are always open to you. When you're ready, let me know. I promise to make it right.