Frühere Suchanfragen
Suchoptionen
The emerging picture is that, for all of its possibly beneficial uses, #AI is fast becoming the tool tech-oligarchs use for strip mining humanity’s future.
I am increasingly uncomfortable with using it.
I just honestly don't understand what *I* would use the various generative methods for. With my job, with my interests and skill set, they don't offer anything useful. Nor are they AI in the sense that I've been using the term for 40 years.
The most useful task that I've seen generative methods perform is creating very poorly optimised, confusingly commented (if at all) computer code for tasks that have already been mastered.
This has largely been my experience as well.
Also, a lot of the hype around what is referred to as ai pretty much boils down to marketing. It feels like every algorithm and Perl-Script is now being sold as “AI”.
A lot of decision makers also appear not to be technically inclined and lack understanding. It’s like a buzz word, the new thing everyone feels they need to participate in without knowing what it is, what it does, how it does it, or considering if it makes sense to use.
@mrundkvist @Morgunin FWIW I use a few LLM tools through their APIs(so I do pay for them), and for my use I do find them useful.
The new reasoning models are really good to have something approaching a socratic dialogue with. I rarely agree with them, but in thinking about why I disagree with what they say I get a better idea of my own opinions
Are those dialogues the use that is useful to you, or are there others?
@mrundkvist I find them useful. The other application I built that I find super handy was a bot which read the abstracts of all the papers on the ArXiv (where physicists post paper pre-prints) and rank them based on relevance to my research.
Also for my personal website, while I could learn CSS and JavaScript and such, but realistically od never have the time. So using LLMs to do the CSS on my website (according to my specifications) let's me do things which I never would have otherwise found the time to do.
Ultimately I think that a lot of the yapping about it replacing workers is over-played, but it certainly has its place.
The algorithms that video games use to play against a human player have been referred to as "AI" for decades. That's funny, because usually they just involve cheating. If you turn up the difficulty level in Civilization, the game doesn't play smarter. It just gains the ability to create expensive troops out of thin air.
Civilization's algorithm once created a huge Mongol army out of thin air and invaded my empire. Using railroads, I sent all my troops to the first city they would reach. And they weren't smart enough to march around it and take my other 19 now undefended cities. That army crushed itself against the first extremely well defended city it found!