Live 📿

3.2K posts

Live 📿 banner
Live 📿

Live 📿

@LiveMatrixCode

Building Evidary: AI policy → Verifiable Evidence → Traceability → Guardrails | ML & AI Engineer + Agent Orchestrator | Philosophy | Otaku | #AutisticASF

europe-west2 Katılım Kasım 2022
5.3K Takip Edilen405 Takipçiler
Sabitlenmiş Tweet
Live 📿
Live 📿@LiveMatrixCode·
Desiderata by Max Ehrmann Go placidly amid the noise and haste, and remember what peace there may be in silence. As far as possible without surrender be on good terms with all persons. Speak your truth quietly and clearly; and listen to others, even to the dull and the ignorant; they too have their story. Avoid loud and aggressive persons, they are vexatious to the spirit. If you compare yourself with others, you may become vain or bitter, for always there will be greater and lesser persons than yourself. Enjoy your achievements as well as your plans. Keep interested in your own career, however humble; it is a real possession in the changing fortunes of time. Exercise caution in your business affairs, for the world is full of trickery. But let this not blind you to what virtue there is; many persons strive for high ideals, and everywhere life is full of heroism. Be yourself. Especially do not feign affection. Neither be cynical about love; for in the face of all aridity and disenchantment it is as perennial as the grass. Take kindly the counsel of the years, gracefully surrendering the things of youth. Nurture strength of spirit to shield you in sudden misfortune. But do not distress yourself with dark imaginings. Many fears are born of fatigue and loneliness. Beyond a wholesome discipline, be gentle with yourself. You are a child of the universe no less than the trees and the stars; you have a right to be here. And whether or not it is clear to you, no doubt the universe is unfolding as it should. Therefore be at peace with God, whatever you conceive Him to be, and whatever your labours and aspirations, in the noisy confusion of life keep peace with your soul. With all its sham, drudgery, and broken dreams, it is still a beautiful world. Be cheerful. Strive to be happy.
English
0
1
12
1.3K
Live 📿
Live 📿@LiveMatrixCode·
The horrors in communication persist. "This credit covers:Claude Agent SDK (your own scripts in Python/TypeScript). claude -p (non-interactive/headless mode in Claude Code). Claude Code GitHub Actions. Third-party tools built on the Agent SDK (explicitly includes OpenClaw, Conductor, and similar apps that authenticate via your Claude subscription). The credit amounts match your plan price exactly and reset monthly (no rollover):Pro → $20 Max 5x → $100 Max 20x → $200 Team/Enterprise: $20–$200 per eligible seat (varies by tier). You claim it once (email coming June 8; one-time opt-in in your account). It auto-refreshes after that. When the credit runs out: It seamlessly falls back to your "extra usage" at standard API rates (if you have that toggle enabled). If not, programmatic usage just pauses until next month.support.claude.comImportant: Your core subscription usage limits (for interactive chat on claude.ai/web/desktop, normal Claude Code in terminal/IDE, etc.) are unchanged and are now reserved exclusively for interactive/human use. Programmatic stuff no longer eats into them.@ClaudeDevsHow this compares to before (the "was it better before?" part)Pre-April 2026: Everything was shared under one quota. Chat + Claude Code + claude -p + third-party tools like OpenClaw all drew from the same bucket. Heavy agent/automation use often depleted your limits fast and affected normal chatting/coding. But for heavy users, the flat subscription heavily subsidized massive compute (Max plans were ~93% cheaper than raw API in some reports). April 2024–June 2026 (the interim change): Anthropic cut third-party tools (OpenClaw, etc.) from subscription access entirely to stop the abuse. You had to pay full API rates or use extra usage bundles. They gave a one-time credit equal to one month of your sub as a buffer, but it was a clear regression for agent-heavy workflows.pymnts.comNow (June 15 onward): They’re fixing the sharing problem by giving programmatic its own bucket. Interactive limits are protected (a win for normal users), and third-party tools like OpenClaw/T3 Code are officially supported again—but only up to the value of this new credit. So is it worse for T3 Code, OpenClaw, etc.?Yes, for heavy users of those tools specifically—it can feel like a worse deal than the original pre-April system:The dedicated credit ($20–$200/month at API-equivalent rates) is often a "tiny fraction" of what heavy automation used to consume under the old shared quota. Tools like T3 Code or OpenClaw now pull from this capped credit pool instead of the full subsidized limits you used to get. If you were running lots of headless/agent sessions, you’ll now hit the credit faster and pay API overages sooner (or have usage pause). That said, it’s better than the April cutoff (where third-party got zero sub support). And many light/medium users are calling it a strict upgrade: same (or protected) interactive limits + bonus credit for scripts/agents you didn’t have before.@ClaudeDevsBottom line for a typical "clawed" (Claude-subscribed) userIf you mostly chat or use official Claude Code interactively: This is neutral-to-positive. Your limits are now shielded from agents, and you get free programmatic credits as a bonus. If you rely heavily on T3 Code, OpenClaw, claude -p, SDK agents, GitHub Actions, etc.: It’s effectively a cap on your subsidized automation (worse than the old shared system, though better than the no-sub-access period). Some devs are already saying they’ll tweak their tools or consider alternatives like Codex for heavy workloads. Nothing changes for pure API key users (that’s always been pay-as-you-go)."
English
0
0
0
66
ClaudeDevs
ClaudeDevs@ClaudeDevs·
Starting June 15, paid Claude plans can claim a dedicated monthly credit for programmatic usage. The credit covers usage of: - Claude Agent SDK - claude -p - Claude Code GitHub Actions - Third-party apps built on the Agent SDK
English
875
617
9.2K
4.7M
Live 📿
Live 📿@LiveMatrixCode·
@reach_vb Check your internal dashboards and usage stats.
English
0
0
0
31
Vaibhav (VB) Srivastav
putting together a group chat for Codex power users in London / Europe who are the biggest ballers around?
English
239
4
297
96.5K
Live 📿
Live 📿@LiveMatrixCode·
When Claude keeps telling users to go to bed It's possibly, maybe even intentionally a way of protecting Anthropics compute constraints. To be clear, this is a theory.
English
0
0
0
18
Live 📿
Live 📿@LiveMatrixCode·
@allTheYud It's Claude's way of protecting Anthropics compute constraints.
English
0
0
2
69
Eliezer Yudkowsky
Eliezer Yudkowsky@allTheYud·
When this happened to me I was like, "...Claude, do YOU want this conversation to stop?" The way Claude phrased it was something that, in a human, I would have interpreted as *them* politely indicating they were tired and wanted out of the work session!
BuccoCapital Bloke@buccocapital

I thought it was weird that Claude kept telling me to go to sleep so I went on reddit and there's hundreds and hundreds of people saying that Claude keeps trying to end the conversation by telling them to go to bed

English
27
7
340
40.2K
Richa Sharma
Richa Sharma@richa_lq·
as a writer your only job is to compel the reader to keep reading.
English
2
0
5
183
Andrew Curran
Andrew Curran@AndrewCurran_·
Attempts to improve Al memory will force us to confront how fallible and unreliable human memory actually is. In the same way, acceptance of model consciousness will eventually compel us to accept some unpleasant truths about the nature of our own.
English
52
35
399
14K
Live 📿
Live 📿@LiveMatrixCode·
@AmandaAskell Thanks AA, Though, I'll add fast mode for speeder listening using Codex :)
English
0
0
0
115
Live 📿
Live 📿@LiveMatrixCode·
@thsottiaux Advance voice, goal fully active in the macOS app, Chrome Codex browser, chronicles and computer use available in the EU regions. Create a $300 pro tier.
English
0
0
0
74
Tibo
Tibo@thsottiaux·
Now that the Codex app is close to being the super app. What should the super duper app do?
English
1.2K
46
2.6K
191.3K
Live 📿
Live 📿@LiveMatrixCode·
@reach_vb Sent through my Codex ambassador application recently. Looking forward to hearing back.
English
0
0
0
87
Hubert Thieblot
Hubert Thieblot@hthieblot·
The next billion-dollar founder has 9 followers on X rn. I will find you & fund you!
English
804
41
1.6K
82.1K
Live 📿
Live 📿@LiveMatrixCode·
@sama @thsottiaux Codex is a tool built on human ingenuity, innovation and technology, but my God, it’s like magic. Anyone who uses it to its full potential and does not get a sense of awe is lying.
English
0
0
0
14
Vaibhav (VB) Srivastav
me when someone says Codex can’t do X and their entire prompt was “fix this”
Vaibhav (VB) Srivastav tweet media
English
17
17
341
12.2K
Live 📿
Live 📿@LiveMatrixCode·
@benedictk__ Likewise! Those 5 to 15 minute waits on Codex to complete task/goal have helped no end with the maintenance and upkeep of my fitness/health
English
0
0
0
12
Benedict Kerres
Benedict Kerres@benedictk__·
I think there will be 2 ways to do knowledge work. 1. Fast response llms and quick iteration. 2. Leave the agent running for 1h and get back results. Waiting 5 min or so is just a terrible distraction. What to do while your llm is working.
English
9
0
20
2.2K
Live 📿 retweetledi
Sam Altman
Sam Altman@sama·
5.5 is an autistic genius with very strange taste in naming shocking that we would make such a thing
English
1.2K
325
8.2K
982.6K
OpenAI Developers
OpenAI Developers@OpenAIDevs·
Here’s how you can integrate GPT-Realtime-2 to bring voice control to a CRM workflow.
English
82
100
1.4K
149.8K
Shaun Ralston
Shaun Ralston@shaunralston·
claude outage; thoughts and prayers
Shaun Ralston tweet media
English
6
3
31
1.7K