I've been fascinated by elevator algorithms since visiting NYC as a kid. The interesting stuff starts to happen when you account for popular floors, people going to work, coming home at the end of the day, dog-walking times, subway arrivals, all the semi-deterministic behavior we see in real life.
All of this, the news articles, the social media discussion, this very discussion, will be part of the training set for future AIs. What will they learn from this?
I just got an S25 Ultra. How do you use the S-Pen with termux? I do have handwriting enabled in gboard and the S-Pen works for writing regular text. But can it write special characters, ctrl/alt, or something else to go beyond what regular gboard gives?
(I also have Unexpected Keyboard for Emacs, but S-pen input seems like a cool idea if it can work around limitations of gboard)
This is the right answer. Don't just feed the data to a chatbot; have it write code to do what you want, repeatably and testably. You can probably have working python (and a docker container for it) in under 30 min.
It's happening again now with robotics, self-driving vehicles and RL. Factory workers, truck drivers, construction work, order fulfillment, machinists, farm work, medical technicians and more are all very much at risk (same thing as OP: mostly junior roles getting automated). Some info at https://arxiv.org/pdf/2510.25137
I just this week vibe-coded a personal knowledge management app that reads all my org-mode and logseq files and answers questions, and can update them, with WebSpeech voice input. Now it's my todo manager, grocery list, "what do I need to do today?", "when did I get the leaves done the last few years?" and so on, even on mobile (bye bye Android-Emacs). It's just a basic chatbot with a few tools and access to my files, 100% customized for my personal needs, and it's great.
Cool! It would benefit from better physics though, maybe supersampling the position in time especially when moving fast. Each ball can't push to its edge fully, for instance.
And as a manager/CTO, the way to do this is to give the devs time to think about what they're doing, and reward implementation clarity (though it's its own reward for Amandas).
Clearly they were missing Amanda, the engineer who's had to review others' terrible code (and her own) for 20 years, and has learned the hard way to keep it simple. She knows she's writing code mostly for people to read, not computers. Give me a small team of Amandas any day.
I prefer Claude Code (the `claude` cmd line version, with Sonnet 4) because it's more like an actual pair-programming session. It uses my claude acct rather than costing extra per token. It also hooks into all my MCP tools (shell (restricted), filesystem, ripgrep, test runners, etc. etc.) which makes it pretty amazing.
After turning off its annoying auto-commit-for-everything behavior, aider does work OK but it's harder to really get it to understand what I want during planning. Its new `--watch-files` thing is pretty darn cool though.
Update: I refactored it to remove d3 and use solidjs for fully reactive SVG generation. I'm happy with how that came out. Panning and zooming are smoother now on all platforms.
Be a craft programmer. Create or join a small lifestyle tech company that understands the beauty of software, the importance of keeping your tools sharp and your technique polished. Avoid the giant soul-sucking companies where you're just a cog in a giant machine.
Ah, good old Ted Nelson. His "Computer Lib/Dream Machines" captivated and inspired me as a counterculture kid in the '70s, and played a big part in me becoming a graphics nerd today. (https://en.wikipedia.org/wiki/Computer_Lib/Dream_Machines)
No argument here, and I've been using emacs since around 1983. It's a hobby and a way of life. But like any way of life, it's for those who like that sort of thing.
This is fine; there are lots of similar "intro to quaternions" pages around.
One thing I haven't seen yet is someone showing how to interpolate between multiple quaternions over time, smoothly. (Hint: it's not slerp.) It's a non-trivial problem.
I tried to do a clockless fully async bus interface in around 1988 in a chip I was designing at Masscomp for a fast data acquisition system. Never got built, but it was fun trying, and it would've been really fast. "Lower design complexity" though: hahaha! Nope.
One thing not noted in the article is how the albedo reduction from greening in the arctic affects warming, ice melt, etc. It is good that greening has a negative feedback effect, but albedo reduction likely reduces that benefit in those regions.
That's a really cool application of Blender! I'm also using Blender headless in docker containers on Google Cloud, though not using Cloud Run. I'm doing it for self-service molecular visualizations -- "create your own molecule video". It's at https://bioviz-studio.com. But I run up to 50 dual-T4 instances for speed (Cycles X helps a lot!). And I agree the python API is a bit weird, and hard to use, but once you get the hang of it you can create really complex and interesting scenes just using python.