leo

in my software homesteading arc

Advice API - inspired by emacs `defadvice`! Thoughtful set of primitives + concepts that work together.

@chadptg @nicdunz @thsottiaux This should be fixed now :)

@nicdunz yeah same. it’ll just say it’s blocked and then fallback to commands even though everything is enabled. if i start a new thread it works for a little. hopefully @thsottiaux can fix soon

i dont understand why i cant use the codex app with computer use

@Daeshawn This should be fixed now :)

Codex Computer Use is amazing but man this blocker is annoying ... Anyone else running into it? "I tried. Computer Use can see that Safari is running, but access to the Safari window itself was blocked by the plugin permission gate, so I can’t inspect the open tab yet. What I observed: Safari is running Computer Use returned: access denied for com.apple.Safari"

@embirico queue by default, cmd+enter for steer

Do you prefer queuing or steering in Codex? What's your use case for each?

@SIGKITTEN yes - with model accessible tools I hope

im gonna be adding "apps" to kitty litter. they'll have persistent storage so you can come back to use them later and will have your stuff saved

in my software homesteading arc

This talk by @andy_matuschak is the clearest articulation I’ve seen of where I believe personal computing should go. The hard part isn’t generating more apps. It’s defining the "physics of a medium" where new behaviors can compose and compound. I believe that’s the path to personal dynamic media, and the thread running through my personal runtime experiments like vrs and eden. Must watch. I’ll be coming back to this one often: andymatuschak.org/tat/

I didn't use my terminal *once* today. Just my browser, slack, and Codex app.

emacs, hammerspoon, popclip, we are soooooooooooo back

Live-scriptable environments, like Emacs, are SO BACK in the LLM era. I’m excited for software with handles everywhere: things you can inspect, script, and reshape live. No sealed apps. No AI sidebars. Just hackable software. Great for humans, even better for models.

What if your whole computer were just pixels streamed to you from a model? I’ve been working with @zan2434 and @drewocarr to imagine a version of generative computing that’s much more flexible and visually rich than the GUIs we have today. (Video is sped up and edited)

@usereditable Love! The hierarchical lens is unique. Reminds me of Tcl/Tk scripting

Exploring a concept…

@lexrus @SIGKITTEN I dare you

If Codex had this theme, I'd be willing to spend an extra $10/mo

@synfinner Try `@` - computer use is a plugin, not a skill. We're working on cleaning this up :)

Still can't figure out why Codex can't see the computer use skill. But if I navigate to the plugin an click 'Try in chat', it works. :(

I’ll miss you, Bill!! I’m constantly thinking back on our chat about a future where “everything is in latent space” - it’s a reminder I carry with me at OpenAI to imagine bigger, still-unrealized futures. There’s still so much to push on + uncharted territory to explore. Excited to see what’s next for you!

I’ve decided to leave OpenAI. Below is the note I shared with my team. Building Sora zero-to-one with you all has been the honor and adventure of a lifetime. As this team knows well, one of the best parts of working on video is that you can see the scaling behavior with your own eyes, and there have been a lot of oh-shit moments over the years. About a month into Sora, way back when this was just a tiny two-man effort, we saw a sample where a land shark swam by a bunch of intricate cacti in a desert (it was a weird prompt), and the details of each cactus were perfectly preserved once the shark passed. We had never seen object permanence like this in any video model. That’s when we knew we were onto something. It’s amazing how much the Overton window has shifted on video models. OpenAI has a high tolerance for crazy moonshots, but even back in July 2023 there was a ton of skepticism that high-fidelity 1080p multi-shot generation was achievable within a year given the state of video in the broader industry. We managed to get there 7 months later. While the original Sora ignited a huge amount of investment in video across the industry, it took the next generation of models with Sora 2 for the broader public to understand the transformation happening. I’m proud of all the sleepless nights before and after the launch this team endured in order to deploy the technology in a responsible way and help steer societal norms. I am immensely grateful to Sam, Mark, Aditya and Jakub for fostering a research environment that allowed us to pursue ideas off-the-beaten path from the company’s mainline roadmap. It’s tempting in life to mode collapse to the most important thing, but cultivating entropy is the only way for a research lab to thrive long-term, and Sam deeply understands this. Sora was a project that could not have happened anywhere but OpenAI, and I will always deeply love this place for that. I’m going to miss this team a lot, but there are great things on the horizon for all of you. I will always be in your guys’ corner cheering you on. Love, Bill

@SIGKITTEN cat pfp TL takeover soon.

why arent u following this kitty?

@sughanthans1 it's realtime!

@leoshimo is this sped up?

Codex computer use + Spark absolutely ripping through Universal Paperclips:

@fat Great cap + great neighborhood

Sometimes we make hats