Rick
2.7K posts

Rick
@pan3sar
Microelectronics & Agentic Engineering.
Bali, Indonesia Katılım Ağustos 2011
600 Takip Edilen215 Takipçiler

For my non-tech friends, this is like Ronaldo joining Manchester City
Andrej Karpathy@karpathy
Personal update: I've joined Anthropic. I think the next few years at the frontier of LLMs will be especially formative. I am very excited to join the team here and get back to R&D. I remain deeply passionate about education and plan to resume my work on it in time.
English

This is wild. I'm not using my account atm
I have no 3rd party apps installed. I have nothing scheduled.
Still something is hammering my current session. I've already used 51% of my weekly usage and it's Monday.
@bcherny @trq212 are you guys seeing this bug across accounts?

Rick@pan3sar
There is a bug that is consuming weekly usage without any actual usage or minimal. 7 hours ago, at 2am local time to me limits got reset and there’s already usage on my account. Anyone else seeing this?
English

Paid $100 for Claude Max 5x, forgot to turn off my VPN, used it for ~30 minutes, then got logged out and hit with a “your account was suspended” email.
Already submitted an appeal.
For people who’ve dealt with this:
Does Anthropic actually reverse these suspensions?
Do they issue refunds?
How long do appeals usually take?
Because paying $100 for 30 minutes of access is brutal.

English

There is a bug that is consuming weekly usage without any actual usage or minimal.
7 hours ago, at 2am local time to me limits got reset and there’s already usage on my account.
Anyone else seeing this?

ClaudeDevs@ClaudeDevs
Happy Friday! We've reset everyone's 5-hour and weekly rate limits.
English

@bcherny @Taniyatweets_ Can you help with any info on this? I can't work out how i've used my week's usage on a max20 account. I've put code to work on extra usage to investigate but he cannot find anything.


English

@Taniyatweets_ A few prompts sounds very low. Would you mind running /usage and pasting the full output here? Happy to debug together
English

Anyone else experiencing this? I genuinely can't work out what to use my usage.
Had Claude dive in and nothing found. @ClaudeDevs any advice?

English

Introducing SubQ - a major breakthrough in LLM intelligence.
It is the first model built on a fully sub-quadratic sparse-attention architecture (SSA),
And the first frontier model with a 12 million token context window which is:
- 52x faster than FlashAttention at 1MM tokens
- Less than 5% the cost of Opus
Transformer-based LLMs waste compute by processing every possible relationship between words (standard attention).
Only a small fraction actually matter.
@subquadratic finds and focuses only on the ones that do.
That's nearly 1,000x less compute and a new way for LLMs to scale.
English

@dpSingh702 Getting my Claude account banned. 🫠 I had Claude build my Gstack for the claw agents but I should have publicly checked this.
English

@pan3sar building in public forces you to question everything — because strangers will question it for you whether you like it or not 😅
the "didn't question it" phase is real though. you just want things to work so you accept the defaults and move on.
what made you finally stop and q it?
English

This Claude course can save you 20+ hours a week on GTM work you are currently doing manually.
Over 3,000 founders and agency owners requested access last month... but I couldn't send it to everyone.
So I'm releasing it again today.
It reveals the exact Co-work setup needed to turn Claude into a system that pulls reports, triages your inbox, monitors your tools, and runs scheduled tasks every day without you triggering a single thing manually.
Like + comment 'COWORK' and I'll share the full course for FREE.
Follow me to receive the DM.
For the next 48 hours only.

English

Great news.
I've officially moved the first of my 5 OpenClaw agents to Claude Code.
It has the same functionality as my Claw and all of my Claw's memory, context, knowledge base, and personality, but writes to my second brain MD files instead of to its own separate files like a Claw does. This means that I can port the second brain into any agentic harness I want (Gemini, Codex etc) and have access to all these models without high switching costs.
When I'm sitting at my computer, I use Claude Code via CLI or the Claude Desktop app, and when I'm away from my computer, I can continue working with my agents via Telegram.
Best thing about this is that I get to use my Claude Max subscription with the latest and greatest Opus 4.7 and I get Anthropic's enterprise grade security.
English





