Generative AI is not smart to begin with. LLM are basically just compressed versions of the internet that predict statistically what a sentence needs to be to look “right”. There’s a big difference between appearing right and being right. Without a critical approach to information, independent reasoning, individual sensing, these AI’s are incapable of any meaningful intelligence.
In my experience, the emperor and most people around them still has not figured this out yet.
You use LLMs for everything? Seems strange, as they don’t reason. They are specifically designed to mimic human speech. So they are great for tasks that require presenting information that looks intelligible, or at least are very easily testable, but beyond that you run into serious issues with hallucination fast…
Or do you mean “AI” as in data science and automation? That’s a very different thing which is a bit off topic. That Kind of “AI” is neither new nor has the hallucination/ecological/cost/training effort issues associated with it
I dunno dude, all your answers talk about “AI” in suspiciously vague terms. “I use AI to …” is the new “built with blockchain”. Skip the marketing terms and talk shop.
Sounds like neither of you watched the video. Fortunately, I did so here’s a quick summary. The thesis is that music is getting worse, for a few reasons. Author argues:
The first point has been touched on by many other people. It’s a common trend in a lot of places outside of music too. People are replaced with machines and processes in a lot of settings especially in corporations and commerce, and while that’s great for efficiency and predictability, it creates a sterile landscape devoid of human expression. This is not to say all music has this. But mass market music is a chief culprit.
The other point really resonates with me with videogames and videogame sales. You can get a dozen great steam games for the same price as a single Nintendo title, yet I probably put 10x the time into that one Nintendo title than all the other steam games combined. Had to get every bit of value out of that expensive Nintendo purchase. YMMV on this point though. I don’t stream music so I can’t say how it has affected me personally.
Terminals are powerful and flexible, but still slower than a dedicated UI to see states at a glance, issue routine commands, or do text editing.
Terminal absolutists are as insufferable as GUI purists. There is a place and time for both.
The benefit of AI is overblown for a majority of product tiers. Remember how everything was supposed to be block chain? And metaverse? And web 3.0? And dot.com? This is just the next tech trend for dumb VCs to throw money at.
Just wait till the advertisers find out the eyeballs they are paying for are also just AI sock puppets. Enshitification strikes again.
The wording of the article implies an apples to apples comparison. So 1 Google search == 1 question successfully answered by an LLM. Remember a Google Search in layspeak is not the act of clicking on the search button, rather it’s the act of going to Google to find a website that has information you want. The equivalent with ChatGPT would be to start a “conversation” and getting information you want on a particular topic.
How many search engine queries, or LLM prompts that involves, or how broad the topic, is a level of technical detail that one assumes the source for the number x25 has already controlled for (Feel free to ask the author for the source and share with us though!)
Anyone who’s remotely used any kind of deep learning will know right away that deep learning uses an order of magnitude or two more power (and an order of magnitude or two more performance!) compared to algorithmic and rules based software, and a number like x25 for a similar effective outcome would not at all be surprising, if the approach used is unnecessarily complex.
For example, I could write a neural network to compute 2+2, or I could use an arithmetic calculator. One requires a 500$ GPU consuming 300 watts, the other a 2$ pocket calculator running on 5 watts, returning the answer before the neural network is even done booting.
Article summary: Japan’s system is not interchangable with systems outside Japan, which is a friction point for export.
I’m pretty sure Windows is a key part of their “cloud stuff” strategy. You are right that consumers are not the direct focus of Windows, since they are not the direct paying audience, and that shows in the direction Windows is going, but getting consumers to use Windows is a big part of creating corporate buy in for Microsoft cloud services. Corporate environments will shun Microsoft cloud services if employees can’t use Windows, or Windows features run afoul of corporate policies (like blanket LLM bans).
Just yesterday here on Lemmy, I mentioned the dangers of violating privacy, and some commenters went on about “what dangers?” Implying there were none…
Is it not enough to gesture broadly?
I have noticed that the quality of results on Google and DDG and others have been declining steadily over the last few years, and I think this is mostly a result of click farms generally getting better at gaming the system. Genuinely quality content is just being drowned out by crap.
ChatGPT doesn’t really address this. I also don’t see ChatGPT as a genuine replacement yet because 1) hallucination is still too big of a problem and 2) the value add of using natural language for queries doesn’t seem all that beneficial to me. Sorta like, how IF you are already used to a terminal, it will be faster or just as fast as a GUI for many things.
The only real value I have seen from ChatGPT, is for complex boilerplate generation that is very easy to verify. ChatGPT is fantastic for generating regex, for example. Or poems, if you prefer.
The insulation was supposedly improperly installed. There, saved you a click.