There's no inherent difference. It's people mistakenly generalizing after hating on current image and text generation AI models. And there's obviously a discussion to be made about how these models were trained unethically by stealing data, but a) that is irrelevant to their usefulness as a technology b) it was recently shown that you can train them with purely public domain data with good results (yogthos posted an article from Stanford the other day).
GenAI is not (inherently) a grift, unless you believe the following applications are nothing but grifts:
Just because AI became a buzzword doesn't mean that these technologies were not under the AI umbrella. I have an AI book from 2003 that mentions solutions to problems such as object detection, obstacle avoidance, machine translation, language generation and so on.
In my experience, it doesn't matter if they open at 10 or 11, because people will start showing up at 12 or 1am. And there's some bar/night club hybrids (bar floorplan but packed with standing people and the music is at night club volumes) that for some reason society has marked as "after" places -- where you go after your "main" club. These don't peak before 2-3am, it's insanity.
Aren't the Triads in HK operating with the help of CPC, or at least with them looking the other way? It's not like the entire game praises China. Although I don't remember the bartender interaction.
Although explicitly asking for hentai to assist you in language learning is a bit bizarre, he has the right mindset. No one is going to get proficient by reading hentai alone but if you have to read porn, might as well do it in your TL. Bonus points if he finds one that's 80% "story" 20% actual porn.
ć: A soft ty sound as in "Katya" or "feature"; occurs nearly exclusively in the combination ić at the end of family names. F Radić, Pavelić, Ranković, Milošević.
How do you expect people to know how to pronounce this without having studied the language before hand? It's a pretty stupid thing to be angry about. People are raised with native language(s) and they can't pronounce sounds or combinations of sounds not found in them without some training.
Japanese is easier to pronounce for English speakers, because it consists of simple syllables that map almost 1:1 with English ones. And people still mess up pronunciations, because of course they would, that's how languages work, unfortunately.
It also helps that some Japanese words appear in English as their phonetic pronunciation and not their literal transliteration. E.g. tofu is actually written 'toufu' in hiragana, with 'ou' being a long 'o' sound.
Why do people comply with England's honors system and always call knighted people 'Sir'? I see it all the time with popular figures (e.g. Attenborough, Ferguson) on social media comments and I'm not from the UK.
You can also notice this in wikipedia. The page for a knighted person will always start with "Sir X is/was a ..." even in languages other than English. The only exception I found to this was French wikipedia. I hope their journalists and general public do not comply with this idiotic nonsense as well.
The one time i used the web search thing it copied and pasted stuff directly from the sources, so maybe I made some wrong assumptions about that. But limiting their replies to what exists in a given web page shouldnt be too hard
I haven't watched Black Mirror so I can't really compare.
we don’t need to accurately predict all the exact ways it will go wrong ahead of time to make a point about how capitalism interacts with technology
I agree with this, that's what I said at the part where you quoted me. But I think there should be some thought behind the satire. You could complain about:
The energy costs of running these models
People getting displaced because of new data centers using up all the water/electricity in an area
People treating LLMs as oracles
People using LLMs instead of actually learning the thing they're studying
And so on. These are more fundamental problems than a server slowdown, an LLM alarm clock or the canned "As an LLM I cannot..." response.
This is too over the top. I understand the anger towards LLMs and the market hysteria to shoehorn them anywhere ... but alarm clocks? Maybe someone will try to grift silicon valley with an idea like that, but I'm sure it won't have widespread (or any, really) success, similar to the IoT SaaS juicer.
There has been a lot of meaningful improvements in the points made in the text recently. I don't use LLMs frequently, but I used ChatGPT for something the other day and was surprised to find that it started replying instantly, and the speed of the text generation was much faster.
You can also have them reply by searching the Web first. If you do so, they will reply with sources for every claim. I assume a similar feature where they search PDFs/documentation is already in the works or released, so if we ever get to the point where we have AI assistants in cars, they will provide information based on your model only.
Also, I think we're past the point where self driving cars are so useless that they end up looping in the parking lot. I wouldn't be surprised if in 5 or 10 years they're super reliable. An older relative of mine drives an EV (not a tesla, thankfully), and he has no complaints from the assisted driving features (not fully self driving though). For example, he says that if you overstep your lane, the car gradually corrects its position.
I don't believe you have to write a satirical piece that's 100% accurate with the latest models/technology, but right now you're attacking a strawman
Crest and Sensodyne are named in the article. Oral-B, Colgate and Aim are also owned by American companies. The only other brand I know of is Aquafresh, whose parent company also owns Sensodyne.
The american urge to unleash nuclear armageddon