Petio Lazarov

Starting this account properly today. I use AI tools until the ugly parts show up: limits, bad runs, weird failures, wins. I'll post Codex notes and what breaks after the demo. No hype farm. No magic. Just notes from using the stuff.

@thsottiaux just give us gpt 5.6 bro. we dont even want resets anymore at this point

This was fixed. You know what's coming 👀 Give us 24 hours to reset the Codex rate limits across all plans.

@sumaerjolly @danshipper @CopilotKit @attio I keep coming back to the return path. If the UI can tell Codex why a field exists, which record it came from, and what changed, this gets interesting. If not, it is dashboard cosplay with a slower handoff.

Spent the weekend tinkering on @danshipper Codex native app idea. Here is a basic demo using @CopilotKit on top of @attio CRM data. Attio here is the data layer/source of truth. CopilotKit generates the UI on the fly Codex becomes the agentic workflow layer through its native browser. The important bit is that the generated UI can hand context back to Codex. Feel like this version where AI generates UI on the fly, will be better than computer use. With browser use the agent still has to operate through whatever UI the SaaS app gives you. Here the UI can be generated for the exact task and context in front of you. Since it all runs inside Codex you also get your own skills, plugins, tools & context handoff built in. It goes back to the idea that Codex & Cowork become the layer for knowledge work. The app is no longer the surface. The UI is just whatever the task needs.

@Da7_Tech If they do it how will they cut limits every 2 weeks?

AI subscriptions need real transparency. Don’t tell users “5x, 10x, 20x usage.” 20x of what? How many tokens do I actually get? What counts against my limits? Can the real usable quota change after I pay for a yearly plan? This is why I respect MiniMax — despite some notes I have about them. They clearly state that their Plus plan gives roughly 1.7B tokens per month. That kind of clarity makes users feel safer, especially with long-term subscriptions. The problem is that many companies do the opposite. Codex is one example. The $20 plan used to feel genuinely usable. Now, in many cases, it barely survives one serious task before hitting the 5-hour limit. Perplexity is another example. Users went from 600 Deep Research searches per month to only 20. That is a 96.7% reduction — basically a 30x cut — even for people who had already paid for Pro annually. Gemini is another example. I paid for a full year, then shared limits were introduced across images, videos, and other features. These limits did not exist before in the same way, and suddenly the plan became much tighter for real usage. And GLM Coding Plan is the clearest example for me personally. The issue I faced — and many others reported too — shows the same core problem: unclear limits, abnormal usage counting, and users not knowing what they are actually getting. When I pay for internet, I know how many GB I get. When I pay for electricity, I know what I’m being billed for. But with AI subscriptions, users often pay for a marketing promise they cannot measure. That leaves customers exposed to silent quota changes, tighter limits, and the feeling that the plan they bought is not the plan they are receiving. AI usage limits should be transparent, measurable, and eventually regulated. Users deserve to know exactly what they are buying.

@kunchenguid Too little too lqte

wow - this is huge! anthropic is officially walking back their decision about banning programmatic use of claude code subscription quota why is this a big deal? this is a signal that anthropic is revisiting their ecosystem strategy which many of us have been criticizing by allowing invoking claude code programmatically, anthropic will basically extend their subsidized subscription to power a much wider range of applications, not just their own, which effectively means they are leaning more into being an infrastructure provider rather than the super app that eats everything else they still have more to do to gain back my trust as a developer but this is a very positive change and i'm happy to see anthropic revisiting their strategy

@OpenAIDevs Just give us gpt 5.6 already we dont care about those bullshit

Use the OpenAI Developers plugin in Codex to build faster with OpenAI tools by setting up API keys, finding the right docs, and debugging along the way.

@harukaawake Its the niggers

🇯🇵 I honestly can't even imagine behaving in this way. How are these people not ashamed? If we did this in Japan we would feel so much embarrassment.

@sama @thsottiaux please tell us what is happening? Gpt 5.5 is getting more and more retarded. It felt peak 1 moth ago, from there on it has been only a downfall

@thsottiaux Where is gpt 5.6 niger

Codex can see and set its own /goal. Everything we build, we build also as a tool for the agent. This is a generalization of meta prompting, where you let the agent set its own task based on your intent.

@Adidotdev they can bet on speed + slight increase in intelligence. I am fine with that. if the speed is 3x of what it is now, while we get increase in quality, and the price keeps the same or goes down... well I am really really good with that

if GPT-5.6 is weak, Anthropic wins. if GPT-5.6 is strong, the government bans it too. OpenAI can't win this round.

usage limits need a run card. budget. stop line. retry cap. survival note. the annoying part is finding out after 18 minutes of wandering.

@MoonDevOnYT Hey moondev I used to watch your videos like half a year ago, I am wondring how much you made trading during that time

you are getting wrecked on polymarket because you trade like a pure degenerate instead of tracking actual whales. i built a custom scanner that strips away the noise and exposes the exact trades of highly profitable wallets in real time. watch me leak the exact code to reverse engineer their alpha and stop trading like a loser here

@danshipper next chart i want: runs started vs usable diffs, because a forced spike is not the same thing as codex earning the slot.

before and after fable ban: my claude app vs. codex app usage

@MoonDevOnYT wait so you are giving access to your code in the zoom meeting?

you are literally burning money on polymarket while i just forced 36 ai agents to strip mine the platform for alpha it is frankly disgusting how fast these bots uncover the exact loopholes needed to bleed the rest of the market dry watch me leak the raw prompts and steal the winning bot blueprints before this gets taken down here

@foxtomb232 I am new in twitter, and I would really appreciate a shoutout

I started as a reply guy with 0 followers. Now I’m at 17.2M and guess what? I’m still replying. If you’re building too, drop a reply and let’s connect 🔒💬

@blakefakhoury and why dont you show the youtube channel

start an AI sleep channel. trust me lol, this takes us less then 5 minutes a day

@MoonDevOnYT Hey moondev why did you close your github projects behind a paywall brother? make a monthly sub at least, I can't afford to pay several gazzilion dollars for that cmon :D

fable 5 has been killed by the US government if you didn't spend the past 72 hours building trading systems with it you will be left in the past we will never have that powerful of AI again

@MoonDevOnYT can you share the strategy for science?

@sattyyouneed make the limit visible before the run starts. otherwise the agent spends your cap like it found a company card.

Is there any trick to avoid Codex usage limits?

@rezoundous they had to give us the jailbreak instead

Next level snitching

@Latin0Patri0t @ChrissGPT Bro I am from Europe... I can't count on Europe to do anything... we only regulate, we don't produce. Same happens in America now.

@petiosz @ChrissGPT You should only count in your country not a foreign country. Mistake #1

OpenAI already requires ID for some features. Anthropic will most likely simply do the same to use mythos. This will continue to get more stringent as we get closer to AGI