Yeah, that’s because that’s all that you’ve seen. Not because that’s all that there is.
The tip of the iceberg for someone who doesn’t know what is beyond the surface.
LLM’s have everyone’s focus simply because they got so popular and are the only AI’s that regular consumers can use. There are a lot of advanced machine learning models helping out behind the scenes that all fall under the original umbrella term of AI. For example in science and in medicine (especially medical imaging).
Microsoft isn’t working on those things. Consumer facing shit is what people are talking about when we bitch about AI. We’re not going to add an annotation with the few exceptions every time we post about how terrible it is. Besides that there’s plenty of LLM shit being put into medicine and science where it doesn’t belong too. People are going to die because of this crap if they haven’t already. For the vast majority of people there is no distinction and that’s dangerous as fuck. Better to conflate it all as shit and let the useful products rebrand behind the scenes than to keep shoving LLMs into everything because they’re riding on the coattails of a handful of good applications.
Who said anything about Microsoft working on them? But yes, Microsoft is probably also working on those…
And that is exactly the point. Don’t blanket bitch about AI when you only mean a tiny subset of AI, that happens to have become a fad.
Ok let’s say
Cars <-> cybertruck
Same as:
AI <-> LLMs
If Tesla makes a shitty car, does that make all other cars from other manufacturers that have come before it and will come after it also shitty?
Why do you say: AI is shit when LLM’s and the like are shit, and not cars are shit when the Cybertruck is shit?
(Ok not a perfect example because Cybertrucks aren’t as popular, but you get the idea).
And there is not just a handful of good applications. They’re just not consumer models and are used in certain fields as I mentioned before. As a whole, the number of ML algorithms still vastly outnumber the LLM’s and the like. That may shift now everyone and their mother is making an “AI”. But for now, let’s stick to the facts.
Yeah, that’s because that’s all that you’ve seen. Not because that’s all that there is. The tip of the iceberg for someone who doesn’t know what is beyond the surface.
LLM’s have everyone’s focus simply because they got so popular and are the only AI’s that regular consumers can use. There are a lot of advanced machine learning models helping out behind the scenes that all fall under the original umbrella term of AI. For example in science and in medicine (especially medical imaging).
Microsoft isn’t working on those things. Consumer facing shit is what people are talking about when we bitch about AI. We’re not going to add an annotation with the few exceptions every time we post about how terrible it is. Besides that there’s plenty of LLM shit being put into medicine and science where it doesn’t belong too. People are going to die because of this crap if they haven’t already. For the vast majority of people there is no distinction and that’s dangerous as fuck. Better to conflate it all as shit and let the useful products rebrand behind the scenes than to keep shoving LLMs into everything because they’re riding on the coattails of a handful of good applications.
Who said anything about Microsoft working on them? But yes, Microsoft is probably also working on those…
And that is exactly the point. Don’t blanket bitch about AI when you only mean a tiny subset of AI, that happens to have become a fad.
Ok let’s say
Cars <-> cybertruckSame as:
AI <-> LLMsIf Tesla makes a shitty car, does that make all other cars from other manufacturers that have come before it and will come after it also shitty? Why do you say: AI is shit when LLM’s and the like are shit, and not cars are shit when the Cybertruck is shit?
(Ok not a perfect example because Cybertrucks aren’t as popular, but you get the idea).
And there is not just a handful of good applications. They’re just not consumer models and are used in certain fields as I mentioned before. As a whole, the number of ML algorithms still vastly outnumber the LLM’s and the like. That may shift now everyone and their mother is making an “AI”. But for now, let’s stick to the facts.
This is the technology community after all…