It's interesting. I've been tinkering with an article summarizing/highlighting browser extension, and realized that I don't want the end-user to have read AI-generated content because it's not as high-quality as I'd hoped. But on the flip side, I'm loving having the AI write most of the code for me.
You only have to change every car, truck, tractor, water heater, clothes dryer, lawn mower, leaf blower, stove top, etc to be electric instead of using natural gas (ie electrfy everything), which likely requires all the changes that the parent mentioned. Also, not everyone has enough land to install enough PV to power their entire homes + transportation, so a lot of this is going to have to come from the grid, which requires changes in transmission which are being actively blocked by the current administration (see Grain Belt Express).
And throw it all away before you fly home? That's like saying "Why wash dishes when you can just throw them away and buy new ones"?
As an aside, you should absolutely wash your clothes after buying them before wearing them. They're covered in chemicals that aren't great for your skin. I usually pack a week's worth of clothes and do laundry once a week while traveling. Doing laundry more often is a drag.
Hey, thanks for building this. I'm trying to use it for this usecase. I'm trying to pause the script during the manual steps. How would you go about this? Example:
1. Change working directory block
2. Interactive terminal
3. Manual Step
4. Interactive terminal
5. Manual step
If I press run at the top, then it doesn't stop for me to do 3 before executing 4. Should I run each step individually for this usecase? Will changing the directory in step 1 apply to steps 2 and 4 if I don't hit run at the top?
This is not recycling. Recycling implies that you can produce the same product again many times; it's a sustainable practice. This is repurposing or upcycling. It's cool they're getting a second life, but they won't get a 3rd, 4th, Nth life unless the batteries are actually recycled into their component materials at end-of-life. It's kind of like the plastic brick companies: cool that plastic is being turned into a construction material, but it doesn't mean we can stop mining for the primary source material any time soon.
I'm not a Python dev, but had to write a script the other day and got all cought up with the virtual env stuff. Why can't `uv` just infer the dependencies from the `import ...` line? Why declare the dependencies twice?
Python import names are not necessarily unique or the name of the package on pypi/pip. Something like PyYaml is imported as yaml, but potentially other packages could supply a slightly different yaml to import
I had the same thought with a recent NewScientist article. Signed up, read it, and tried to cancel to avoid the recurring fee. There's no click to cancel; I had to submit a form + request and repeatedly check my email to see if my request had been honored. I'm still waiting for the perfect pay-per-article platform to show up.
As someone who has built a timer-based procrastination browser extension, I'd like to add that it would be a nice touch if you could stop playback between videos (or maybe between chapters on long videos) rather than cut people off right as the timer goes off. It's a bit jarring to be in the middle of something you're enjoying and for the screen to go blank.
This isn't entirely on-topic but I've been trying to understand why AI video editing isn't more common, and thought you might know. I've had an idea for a while to make tennis match highlight videos that show every single point of the match. Tennis has a lot of downtime between points (and even more between games and sets). I just want to tell an LLM: here's a two-hour long video of a tennis match. Strip out all the gaps between points. I'm guessing this would a very expensive frame by frame analysis of the video right now and that's why it's not done. Is that right or are there other reasons?
Yeah, it's probably too expensive and complicated to send all the frames to an LLM. I only send about 30 images at 576x324 which is around 15000 tokens a video, and comes to about $0.045 per video. First I save one frame every second and then loop through them comparing each to find the differences, and send just screenshots which have changed significantly, up to a max of 30. Claude only allows 100 images per API call, so it would be a bit fiddly and costly to handle 7000 frames.
Though, now that I'm thinking about it, you could probably do this locally and just look at the part of the image that has the current score, do some local OCR on it to check if the score has changed each frame, if it has, store the timestamp and then use ffmpeg to extract the correct parts. Probably wouldn't need an LLM at all.
As for editing, one thing I do in my videos is audio keywords so my app can do specific things. For example, I can say "AI, mark what I just said as important." Then when it transcribes the audio and the LLM processes it, it will mark that part as a Distinct Moment with a start and end timestamp, a title and description that will show in my app as a clickable link to that part of the video. I'm thinking of adding more commands for more complex editing too.
