antirez

Btw that's less than 1000 lines of C code, plus my botlib (in C too). Uses MacOS low level APIs to access windows and send keystrokes.

Remembering to run stuff with tmux, tunnels, termux or another SSH client on the phone is ways too much. So I thought I could build a Telegram bot that gives me access to terminals running on my computer when I'm just with my phone around. Security must be handled. I believe I'm going to use it.

Yes, so I use Claude Code for the 80%, Codex to reach the 100%.

Typo, should be "especially if the make the core much more complicated". Fixing. Thanks.

Anthropic may be near to release a new model, however, as things are today, after weeks of testing, I'm totally sure Codex with GPT 5.2-xhigh is definitely a model capable of handling way more complex tasks.

My AGENT files are create iteratively by concatenating things I needed to keep telling the agent in order for it to behave, so it would be really strange if once in the AGENT file those things would no longer work.

YouTube video by Paolo Caressa Norbert Wiener: scienza e vita di un grande matematico sfogliandone qualche libro

Un video sulla vita e le opere di Norbert #Wiener, grande scienziato, inventore della #cibernetica e matematico sopraffino, raccontate sfogliando qualche sua opera.

youtu.be/dSPd7excxC4?...

The new AGENT markdown file in the Flux2.c project -> github.com/antirez/flux...

As I did time ago with the Claude account, let's stop also the ChatGPT subscription I did for an error via iPhone to redo it via the web site. I don't think the 30% cut is fair and, as a customer, I want to express it with my choices.

Second time my MacOS crashes because of a combination of Ghostty memory leaks + MacOs kernel terrible handling of OOM conditions.

"What's your workflow with coding agents?"

Well, this, among the many others.

Still a work in progress, but, what do you think? Redis documentation for agents: redis.antirez.com

New blog post: Automatic Programming -> antirez.com/news/159

Recently I started using the term "automatic programming" in my YouTube channel, and somehow I'm sticking to it more and more.

CLAUDE_CODEX_SKILL.md CLAUDE_CODEX_SKILL.md. GitHub Gist: instantly share code, notes, and snippets.

Here is the skill file that lets Claude use Codex.
gist.github.com/antirez/2e07...

If you explain how to do it, Claude Code is able to use Codex when it is incapable of solving certain issues. This way you have the best of both worlds.

The email is IMHO mostly "ingestion". It is useful to forward emails to agents saying remember me this stuff or alike, or send me an invite for the calendar, ...

We will need AI personal assistants. But why provide access to your email, calendar and files? Your secretary (if you have one) has access you your email and calendar? Usually not. An AI assistant entry point should be WhatsApp/Telegram and email. It will then have it's calendar, files, WhatsApp.

Yep that's possible indeed, even if I think that's a field where there are big diminished returns after a certain level. People buy things following the most obscure impulses that are rarely manifested on social networks.

Yes, but likely no need for that amount of GPUs to do segmentation and other cool stuff they released. And: right now the most actionable / frontier AI is LLM driven, no doubt. Not referring just to the API, but to the actual usefulness and business value.

Twitter error page in 2010

Twitter is over capacity.

Too many tweets! Please wait a moment and try again.

#InternetHistory

Apple didn't even try since the board is only focused on pumping the stock price, without any plan for the future.

I wrote this in April, before Claude Code and Codex. Now I am changing my mind: The Carpenter can indeed just ask for a nice table.

I don't understand how META AI ambitions are still trusted by markets. It is the company with the worst ever AI spending compared to the results they got. Partially because of the LLM skepticism inside, and partially because of wrong picking of the heads of the AI divisions.

Imagine doing this refactoring of files by hand... So much saved time.

My kind of brother suggestions to coding agents.

Sometimes during Redis command design work I feel that CISC / RISC tension inside myself :D

GitHub - antirez/whisperbot: Telegram bot that transcribes audio messages via whisper.cpp Telegram bot that transcribes audio messages via whisper.cpp - antirez/whisperbot

I have had enough of Telegram services run by unknown individuals to do the audio transcription task in my Telegram channels. Wrote one using my botlib, ffmpeg and whispher.cpp. May be useful for you as well, perhaps? github.com/antirez/whis... Transcription quality is WAY better.

90% of people in tech and elsewhere follow what all the others do. It is built in in the human race social behavior.

My blog at antirez.com used to run well in a very old Linode instance. Now I dismissed Linode/DigitalOcean in favour of @Hetzner_Online and the page generation is now 5 milliseconds (I mean, local HTTP request to completion). No web framework + Redis performs well.