i do shit, sometimes without ai
Mumbai
Joined August 2025
- Tweets2.6K
- Following425
- Followers4.2K
- Likes6.6K
update: been over a month with this setup and it is seriously great. regular meaningful updates are only elevating this experience.
the fact that it has an inbuilt editor, diff viewer, browser, terminal/custom chat UI (also a phone emulator if you need it) and all of this is accessible from the iphone app too.
been really helpful for productivity
While I have loved my terminal based workflow with kitty + herdr, I recently moved to using orca (@orca_build)
and honestly, while I miss the level of control and customization I could achieve, this actually makes things really great to work with. managing projects and workspaces feels simpler, a built-in editor, browser, some neat skills alongside some neat integrations with github and linear.
allows setting up automations, easier checkouts, tracking tasks. all I'm looking for is a higher level of workspaces (project folders?) that allow me to separate out work and personal projects and some more customizations. But apart from these, super good stuff.
Give it a shot!!
been saying this a lot lately: write your evals from day one!!!
youll thank yourself only a few weeks later!!
Today at work, we all had a moment of excitement as well as anxiety. Some problems that required full scale ml expertise only a year ago are now being solved by agents.
there’s obviously the challenge of costs and scale, but intelligence is getting cheaper by the day. times are truly changing. It’s troubling because my value is shifting and I dont know where it goes next.
but at the same time, there’s also excitement in the air. it doesn’t feel like an end, it feels like a new beginning.
(sounds a bit too dramatic, but this was my genuine reaction today lol)
Aarno retweeted
Introducing Orcus.
No LLM API key required to make AI model requests.
Orcus uses your web credentials to access AI models available through supported provider websites and deliver responses to your applications through a familiar, API-style interface.
From model invocation and streaming to structured outputs, tool calling, and batch processing, Orcus brings essential AI integration capabilities into one developer-friendly SDK.
Explore Orcus: orcus.site/docs/gemini
Now this is good stuff
Today we're launching @agent_relay
Your team’s agents shouldn’t need YOU to copy-paste their conversations.
Agent Relay lets them read each other’s sessions, ask questions, and hand off work across the tools your team already uses.
Keep using the Codex mac app, Claude in the terminal or even tools like Conductor.
Open source. Free to try our cloud.
agentrelay.com
maybe rewriting everything in rust isn’t so bad
皆様に、今夜に対応して頂きたい大至急のお願いです。
多くのウェブサイトで使用されているPHPに関し、PHPのセキュリティチームが、重大な２件の脆弱性に関する注意喚起を行っています。
公開しているウェブサイトでPHPを使用しているか確認し、今夜、10月21日午後11時（日本時間）に公開される修正パッチについて、可及的速やかに適用できるよう準備をお願いします。
修正パッチを速やかに適用することが困難な場合は、例えばウェブサイト等を一時的に停止させることも、リスクを低減させるための有効策の一つと考えられます。
国家サイバー統括室のXアカウントでも関係する情報を発信していますので、そちらも参考にして下さい。
国家サイバー統括室(注意・警戒情報)（@cyber_forecast） / X
follow these steps:
> go to unsloth’s fine tuning llm guide unsloth.ai/docs/get-started/…
> pick out a sample problem, something simple and easily testable like document extraction
> generate some data, write tests, evals and benchmark runners.
> fine tune a small model and test both variants
boom, you got a solid start to understanding llms!
more than anything else, you still need ownership. 4 people can run 20 agents and do the same thing that 20 people with an agent each could, but with 4 people, you’re building a house of cards. things will break at some point.
ownership, planning and accountability are truly top tier skills today. you cannot let agents fuck around and find out. only humans can.
pydantic ai + logfire is the best setup to build agent applications and custom harnesses today.
change my mind
i like to look at AI today in 3 planes
> consumer
> enterprise
> industrial
if you’re anywhere near enterprise ai, RUN!
companies dabbling in this field generally have no clue what they’re doing. everyone is building the same thing internally and they always conflate automation with ai.
consumer ai is crowded but there’s a great range of things to work on. it’s a great space with a quick feedback loop to learn, your work goes to people much quicker.
industrial ai is truly challenging today. generally includes fields that involve operations, logistics and manufacturing. data is often scattered and messy, requires significant domain knowledge and process understanding. this is where you want to find yourself today.
Aarno retweeted
Deploying your app? Here are 31 things I've learned that every app needs when they get to production. If you have your own leave a comment.
In our new agentic world these will be the minimum requirements for all software.
wisprflow has made some awesome changes with the ux. i have been testing out the notetaker for a couple of days, and one annoying thing it does is launch the notepad. two days straight I closed it. on day 3, i got a suggestion to change this setting in the popup itself.
a minor example but this is great attention to detail. makes using it a much better experience