I’ve seen a lot of sentiment around Lemmy that AI is “useless”. I think this tends to stem from the fact that AI has not delivered on, well, anything the capitalists that push it have promised it would. That is to say, it has failed to meaningfully replace workers with a less expensive solution - AI that actually attempts to replace people’s jobs are incredibly expensive (and environmentally irresponsible) and they simply lie and say it’s not. It’s subsidized by that sweet sweet VC capital so they can keep the lie up. And I say attempt because AI is truly horrible at actually replacing people. It’s going to make mistakes and while everybody’s been trying real hard to make it less wrong, it’s just never gonna be “smart” enough to not have a human reviewing its’ behavior. Then you’ve got AI being shoehorned into every little thing that really, REALLY doesn’t need it. I’d say that AI is useless.
But AIs have been very useful to me. For one thing, they’re much better at googling than I am. They save me time by summarizing articles to just give me the broad strokes, and I can decide whether I want to go into the details from there. They’re also good idea generators - I’ve used them in creative writing just to explore things like “how might this story go?” or “what are interesting ways to describe this?”. I never really use what comes out of them verbatim - whether image or text - but it’s a good way to explore and seeing things expressed in ways you never would’ve thought of (and also the juxtaposition of seeing it next to very obvious expressions) tends to push your mind into new directions.
Lastly, I don’t know if it’s just because there’s an abundance of Japanese language learning content online, but GPT 4o has been incredibly useful in learning Japanese. I can ask it things like “how would a native speaker express X?” And it would give me some good answers that even my Japanese teacher agreed with. It can also give some incredibly accurate breakdowns of grammar. I’ve tried with less popular languages like Filipino and it just isn’t the same, but as far as Japanese goes it’s like having a tutor on standby 24/7. In fact, that’s exactly how I’ve been using it - I have it grade my own translations and give feedback on what could’ve been said more naturally.
All this to say, AI when used as a tool, rather than a dystopic stand-in for a human, can be a very useful one. So, what are some use cases you guys have where AI actually is pretty useful?
Local models are really good at tokenizing the text and figuring the intent in the user input. Not perfect, but much better than any possible regexps you can think of. And it’s a trivial operation you can run even on a CPU model.
When troubleshooting, it’s nice to be able to ask copilot about the issue in human language and have it actually understand my question (unlike a search engine) and pull from and reference relevant documentation in its answers. Going back and forth with it has saved me several hours of searching for something that I had never even heard of a couple of times.
It’s also great for rewriting things in a specific tone. I can give it a bland/terse/matter-of-fact paragraph and get back a more fun or professional or friendly version that would feel ridiculously cringe if I attempted to write it myself, but the AI makes it work somehow.
I use it for little Python projects where it’s really really useful.
I’ve used it for linux problems where it gave me the solution to problems that I had not been able to solve with a Google search alone.
I use it as a kickstarter for writing texts by telling it roughly what my text needs to be, then tweaking the result it gives me. Sometimes I just use the first sentence but it’s enough to give me a starting point to make life easer.
I use it when I need to understand texts about a topic I’m not familiar with. It can usually give me an idea of what the terminology means and how things are connected which helps a lot for further research on the topic and ultimately undestanding the text.
I use it for everyday problems like when I needed a new tube for my bike but wasn’t sure what size it was so I told it what was written on the tyre and showed it a picture of the tube packaging while I was in the shop and asked it if it was the right one. It could tell my that it is the correct one and why. The explanation was easy to fact-check.
I use Photoshop AI a lot to remove unwanted parts in photos I took or to expand photos where I’m not happy with the crop.
Honestly, I absolutely love the new AI tools and I think people here are way too negative about it in general.
Timing traffic lights. They could look down the road and see when nothing is coming, to let the other direction go, like a traffic cop. It would save time and gas.
Or, here me out, we could use roundabouts/traffic circles. No need for AI or any kind of sensor, just physical infrastructure to keep traffic flowing.
Absolutely, but there are a few problems with this. First, I live in the US. Americans do NOT know how to negotiate a roundabout. There is a roundabout near my house. The instructions of how to use it are posted on signs as you approach. They are wrong. They actually have inside lanes exiting across the outside lanes that can continue around. So not only is it wrong but it’s teaching the locals here what NOT to do at a normal roundabout.
Second, they don’t fit at existing intersections.
Third, I think they would be more expensive than just a piece of tech attached to traffic lights that already exist.
I mean the best solution would be some good public transportation, but I’m trying to be more realistic here. That’s for more civilized nations. In the US the car rules. And the bigger, the better.
I live in the US. Americans do NOT know how to negotiate a roundabout.
As do I, but I think the main problem is that we don’t need to properly learn to use a roundabout, because the only times we have roundabouts are when they’re completely unnecessary/unhelpful. The three roundabouts I use most often are:
- right next to a stoplight, so they get jammed when there are a lot of cars waiting
- middle of a residential area, with stoplights/stop signs a block or two away preventing too much contention
- in a somewhat busy area where most of the traffic is going straight, so it functions like a speed bump
If we can figure out those continuous flow intersections, we can figure out roundabouts. We just need to actually use them.
Second, they don’t fit at existing intersections.
They absolutely do, especially at the ones where they’d make the most impact (i.e. busy intersections with somewhat even traffic going all directions). You may actually save space because you don’t need special turn lanes. They are a little more tricky in smaller intersections, but those tend to have pretty light traffic anyway.
I think they would be more expensive
Initial cost, sure, because the infrastructure is already there. But longer term, it should reduce costs because you don’t need to service all of those traffic signals, you need fewer lanes (so less road maintenance), and there should be fewer accidents, which means less stress on emergency services.
Putting in a new roundabout vs a new signal is a different story, the roundabout is going to be significantly cheaper since you just need to dump a bit of concrete instead of all of the electronics needed for a signal.
In the US the car rules
Unfortunately, yes, but roundabouts move more traffic, so they’re even better for a car-centric transit system. If we had better mass transit, we wouldn’t need to worry as much about intersections because there’d be a lot less traffic in general.
If we go with “AI signals,” we’re going to spend millions if not billions on it, because that’s what government contractors do. And I think the benefits would be marginal. It’s better, IMO, to change the driving culture instead of trying to optimize the terrible culture we have.
I think it’s useful for spurring my own creativity in writing because I have a hard time getting started. To be fair to me I pretty much tear the whole thing down and start over but it gives me ideas.
