Dylan
992 posts
Dylan
@dylanjcastillo
🙋♂️ About: https://t.co/zEVJvdkpb3 🛠 Side projects: https://t.co/jpXZeVmfqN
Madrid, Spain Katılım Mart 2020
456 Takip Edilen465 Takipçiler
when you use codex do you want to choose the model and reasoning effort yourself?
should the product just pick based on the task?
what would make you trust that choice?
Are you someone who always uses extra high, or do you think about using low when you don't need that much thinking
English
@HamelHusain I've done that for very small, one-off analysis, but when I have to come back it's painful to make changes.
I'm using @marimo_io more and more these days
English
I prefer html to notebooks now for many tasks including most data analysis
English
Dylan retweetledi
Today I want to share one of the main projects I have been working on: TabulAI. Tabular data runs much of the business world, yet it has not received the same sustained research attention as images, text, audio, or code. TabulAI exists to change that.
1/9
English
Dylan retweetledi
Dylan retweetledi
Sci-Hub is an evil website that pirated 85M+ research papers and made them freely available
And now they've added AI to their database to make Sci-Bot.
It answers your questions using latest, full-text articles.
But DO NOT use it. We should all try to make billion-dollar academic publishers richer.
I'm putting the link below so you know how to avoid it.
English
@LangChain congrats on the rebrand!
(i wasn't initially sure what the new logo represented but a bird taking flight makes sense)
English
Dylan retweetledi
It never fails to delight me when I see cool projects built with Textual.
The author of this project has clearly got a good eye for design in the terminal.
Check it out. Give it a ⭐️
github.com/Kohei-Wada/tas…
English
Dylan retweetledi
We extract nearly all (95.8%) of Harry Potter and the Sorcerer's Stone from Claude Sonnet 🤷🏻♂️
Anthropic@AnthropicAI
We’ve identified industrial-scale distillation attacks on our models by DeepSeek, Moonshot AI, and MiniMax. These labs created over 24,000 fraudulent accounts and generated over 16 million exchanges with Claude, extracting its capabilities to train and improve their own models.
English
Dylan retweetledi
@levelsio Tech gurus on X are always like “voice is the new interface” or some shit and then silently doomscroll their phone for 8 hours.
English
Dylan retweetledi
@shl @daviduche03 People act as if using agents to build software removes all the work; it accelerates the process, but it also upped expectations.
English
Dylan retweetledi
JUST IN: The odds of Jesus Christ returning this year have surpassed Kamala Harris’s chances of winning the next election.
English
@jxnlco Not really for notebooks, but I've been experimenting with building a better harness for agents to do data analytics at kasipa.com.
I'm planning to build an Electron app soon, so that you can connect to any DB and use more compute
English
@intellectronica @opencode Interesting. Why? I use it all the time.
My workflow is usually think of feature, iterate plan with CC, then execute.
English
I really don't like "plan mode". I never use it in any of the agents. It only annoys me when I select it accidentally. At least @opencode allows me to disable it easily.
English
Made a small enhancement to Toad's Diff view.
The edge of the diff has a semantically colored line. i.e. red for deleted, green for added, or grey for no change.
It is subtle, but I think it helps.
WDYT?
English