I've been a happy user of tldraw as a whiteboard for ages. The offline bit is neat, but the killer part of this is that it also becomes a shared workspace that both you and an agent can easily edit + explore. Highly recommend firing it up, Develop -> 'Install Agent Skills', doodling something, and then asking your agent of choice to make some edits or riff on the idea.
There's a 'flow' slider I turn way down when actually painting with it, but I leave it high by default because that's the most "cool" feeling setting for a first interaction :)
Oooh yeah I forgot how incredible that was! They really put care into how the pigments mix and move. I used to love watching timelapses of people doing art in rebelle.
Ah yeah that's not ideal, fixed. I like the menu sticking around on iPad, but it now gets out of the way on smaller phone screens. Thanks for the reminder, I'd been meaning to do this!
(Also, tldraw.com is fantastic for quick+easy diagrams and works great on mobile if you want a better whiteboard that isn't so art-focused :) )
I got some AI writing vibes too, but looking closer, I think it might be human-written (or at least partly so) - perhaps just picking up some AI conversation styles? FWIW, Pangram gives it a mixed but mostly-human score too. Maybe AI is not just changing the way we speak; it's changing the way we perceive all writing ;)
Nicky Bay is incredibly talented and prolific, my favourite macro photographer by far! Big fan of his spider photography, there are many cases where the only high-quality pics of some species will be his. I'm not as deep in the isopod world but this looks like an invaluable resource for people trying to ID + explore them.
Nice build. There are cheap (<$20) commercial versions, often targeted at kids. They make fantastic gifts, I thought it would be a bit of a gimmick but having instant printouts capturing happy moments added a fun dynamic to a few family trips, and our fridge is covered with an ever-rotating cast of family pics. And because they're ~free, it overcomes the blocker of a polaroid having to be 'special' while still keeping some constraints vs just snapping a cellphone pic.
For what it's worth, Pangram (generally very accurate) rates this as 100% "We are confident that this document is fully AI-generated". The few paragraphs I looked at felt solidly AI to me too.
@maybe-tomorrow out of curiosity, is my guess that this is made with help from codex correct? (I'm trying to keep up my sense for the different default aesthetics of the different models, but this one I'm not sure about)
Thinking Machines have put out a string of incredibly high-quality posts lately. Hard to oversell how much cred it's buying them with the AI research community! Keep up the great work folks
And this prompted me to record a video showing some of my random non-work usage recently, to give a feeling for what the app looks like :) https://youtu.be/Y2B27hdKMMA
We made the tool, and that will eventually be available on its own. But the method requires some discipline and 'unlearning'. It's very hard to show someone an AI tool and not have them treat it just like ChatGPT/Claude/... - that's the part that takes the time, and having a community of people working through different examples and case studies together is a lot more motivating for this than just staring at an empty prompt box :)
I've always found this interesting. Think+transmit seems more likely to be the bottleneck vs receive, given that we can easily parse most podcasts etc at 3-4X speed. If being understood by everyone wasn't required, I wonder if one could learn to boost both send and receive rates?
This is going to be my go-to reference for 'how should I get started with electronics' - getting to see various capabilities (sense things, move motors, flash LEDs) without having to purchase parts while not knowing what you don't know is fantastic. Props for pushing through all the projects - I look forward to seeing what you make with these new skills :)
A few observations:
1) the performance vs complexity curve looks very similar to that for most humans (having seen groups attempt Towers of Hanoi with 5 car tires) haha
2) models can trivially solve some of these tasks when given tools
3) this is an internship paper with some quirks that many mostly dismissed, but is being quoted everywhere as "Apple proves LLMs can't ever reason"
Anyway, fun experiment to test your understanding of these things but don't take any conclusions as gospel :)