dgritsko

Was thinking of that exact article while I was reading this, we must be following the same folks. Highly recommend folks reading it, even if it feels insane in some regards it's a fascinating look at what the limits of this stuff could be and I found myself with a lot of new ideas to consider after reading it. Interesting times.

Edit - after writing "we must be following the same folks" I realized that you are, in fact, Simon Willison - I saw this article because you posted about it on X, haha.

Reminds me of this... https://xkcd.com/221/

Your comment implies that this is all that they're working on, which does not seem substantiated.

At least for now, Jev is not multimodal. So a screenshot alone wouldn't cut it.

I think I've got the same three boxes. HDMI cables and adapters fall into the Venn diagram overlap of the last two IMO. You could also make a strong argument for USB also falling into "PC Accessories". But there's nothing quite as satisfying as having the _exactly right_ cable or adapter on hand to solve a problem. Makes it all worth it!

Yeah, this is my experience as well. It's mostly fine, and I've been using it nearly every day for years.

TL,DR: No, unless you're doing it so hard that you're straining your joints. But don't crack your neck.

Recently stumbled across a YouTube channel (no affiliation) where a guy is documenting his process of building a working physical replica of Space Cadet Pinball. I found the process pretty interesting. If you're intrigued: https://www.youtube.com/watch?v=BU_02HABZ4s

This isn't quantized, right? Just a smaller context?

Must ensure that this quote is present whenever Cromwell's Rule is mentioned - "I beseech you, in the bowels of Christ, think it possible that you may be mistaken."

You dropped your "/s".

I use Conductor pretty much exclusively and it makes it incredibly easy to try different models, even within the same workspace - definitely recommend giving it a shot. Whenever I'm forced to use the Claude Code app directly it just seems woefully inadequate compared to Conductor

This sounds fascinating, are you documenting this in any public way for those of us that are inevitably curious on the details?

What a brilliantly written piece. Maintaining one's integrity is unfortunately rare enough that it makes Watterson's story so remarkable. I completely respect and admire his dedication to doing something for its own sake, for holding himself to the highest standards imaginable, and from walking away from it all for his own reasons - even if selfishly I'd rather him keep writing so that there would be more to enjoy. Time to go pull some old volumes of Calvin & Hobbes off the shelf for the hundredth time, I suppose.

Exactly. LLMs at their core are just fancy autocomplete. Extremely fancy, to be sure, and the output that they predict can be very useful - but people who anthropomorphize them or ascribe higher significance to the generated output seem to be missing this.

These blog posts are fascinating to read. I don't have a personal blog, but if I did I'm sure I would've written a very similar post as I've been wrestling with similar thoughts over the last few weeks. I have the distinct sense that I will look back on February 2026 as an inflection point, where AI crossed over from being an interesting parlor trick to something that fundamentally and irreversibly altered what I do day-to-day. It's bittersweet, for sure - it feels inevitable that the craft of software development that I've loved for years will be seen as an archaic relic at some point in the not too distant future. It may be several years yet before the impact is broadly felt (the full impact of today's frontier models has yet to be felt by the general public - to say nothing of models that will be released in the next few years) but this train doesn't seem to be slowing down anytime soon. This post was a helpful reminder that who I am is not defined by the code I write (or don't write) - there's so much more to life than code.

I'm genuinely curious as to what the key differences are (especially those that would cause someone to switch), as someone who is pretty tech savvy but whose use of Linux as a daily driver is admittedly pretty weak.

What made you switch from Pop OS? I just installed it on a couple of old PCs I had lying around for my kids to play around with/learn from.

Surely there's gotta be a better term for this. Recency bias?

Fellow SourceTree apologist here. It remains one of the first things I install on a new machine. I'll do simple stuff directly in the CLI, but stick with SourceTree for anything moderately complicated (as you've mentioned).

FWIW, this is what I wound up with - keeps the original branch name but ensures that it's based on the latest from the "dev" branch:

orig_branch=$(git branch --show-current) && git checkout dev && git pull && git branch -D "$orig_branch" && git checkout -b "$orig_branch"

Ah, excellent - appreciate the help! I'm already getting a ton of value out of this tool, thanks for sharing!

At the moment it's mostly Cursor or VS Code, but I was actually thinking of SourceTree. I'd like to look at the pending changes and manage the commits myself, and I could do that if I could add "open -a SourceTree ." as a custom command. I didn't see a place to edit a setup script, is that just on the filesystem?

This is awesome. A couple of suggestions:

- It'd be great to change the default branch used for creating new workspaces.

- I'd like the ability to add custom tools to the "Open in..." menu.

And if you're one of today's lucky 10,000 and haven't heard of the concept of "lucky 10,000", you can read the relevant XKCD here: https://xkcd.com/1053/

For me, it's the fact that content generated by an LLM is fundamentally different than content that comes directly from a search index, but displaying them alongside each other conflates the two. Most people don't know the difference, and place the same level of importance (or maybe even more importance) on AI-generated content. Yes, this content is convenient. However, if the content isn't accurate or correct (which it may or may not be, given that it's just a statistically likely sequence of tokens) then is it actually beneficial as a whole?

That's a big part of what makes this game enjoyable - a clue that is very obvious to one person might not even cross the mind of someone else. To anyone reading this who hasn't played, it's definitely worth giving it a try.

Great short story. Several times while reading it, I wished that I could download Abelique on the app store and try it out - I guess I'll have to settle for picking up my sketchbook instead.

Apple recently "pulled the plug" on their car project, apparently. https://www.npr.org/2024/02/27/1234315814/apple-cancels-elec...

Was out to dinner with the family, and my kids were enthralled with the "word search" that was on the kid's menu. I thought it'd be cool to build a custom word search generator - given an arbitrary list of words, it'd spit out a grid containing them. The part that made it fun was trying to figure out a layout for an arbitrary list of words that would be as compact as reasonably possible. I was able to get something working in just a couple of hours, and my kids loved being able to do word searches with stuff that was relevant to them, like names of family members. Of course, there are tons of similar generators freely available online, but it was very satisfying to figure it out for myself and come up with something that the kids enjoyed.