I know this won’t be popular here, but I still think it’s worth reading Ed Zitron’s take on AI and its future. His articles are really interesting, but this is a nice summary of what he’s been saying over the past year:
expected an op ed, got a dissertation of opinion. Not sure you can call something that long a “summary”.
I think perhaps his opinion is, despite his protestation, the doomsayer side of it, the AI companies are spinning the glorious side of it, and the truth of the future lies somewhere in the middle.
I’m hardly surprised that AI modelling is plateauing, but that doesn’t necessarily mean its a dead end with no avenues of expansion that is going to immediately collapse. He derides the “scuzzy mindset of billionaire shitheels” (his words), but drives a nail in the coffin because of the costs of the technology (Try naming a technology that didnt lose money during its development? If we gave up on anything that didnt make money, you wouldnt have anything that derived from the space programs of the world…)
Will we hit the limits of AI? Sure. Will it be a disaster? Probably not. To quote the end of Deus Ex…
(Eliza, somewhat ironically, is a video game depiction of an AI.)
2 Likes
I just read the article and much of it sums up my own thoughts and fears about the generative AI hype train.
For example, that ads could possibly be coming to ChatGPT is exceptionally bleak!
However, like Marc says, the future likely lies somewhere in the middle and one cannot but help feel that Mr Zitron is something of a doomsayer. Overlooking AI’s transformative potential disregards its proven impact across a wide range of industries, from medicine to engineering and beyond.
What is your take on things, Ralph?
1 Like
I’ve been reading Zitron’s writings on this for about a year now, and I called this a summary (perhaps wrongly) because he links to some of the other stuff he’s written on the topic.
The overall impression I get from his articles is that (he believes) AI isn’t really able to deliver on what it’s promising. Even though ChatGPT and similar things are popular, they aren’t really able to generate any profit. He seems to be arguing that people who are still investing billions into AI are being conned by people like Altman who have no real idea how to get AI to turn a profit, and that the bubble will pop sooner or later. And of course, there are other issues like energy consumption and the problem that AI will run out of it sources of information, etc.
I guess we’ll see, but it’s not as if we haven’t seen tech bubbles burst before. But the rush to turn everything towards AI certainly seems like a display of herd mentality without much thought about the cliff in the destance.
I think GenAI will eventually destroy itself as more and more content online is generated by GenAI and then GenAI will train on the content it itself produced earlier, basically watering down the information and introducing more hallucinations.
I just hope it doesn’t do too much damage in the meanwhile. It only enforces the trend I’ve been observing that people aren’t willing to invest time and energy into properly rather things, but rather want simple answers to complex problems, which then eventually blow up in their faces.
2 Likes
That is indeed a worry. And it seems that people are betting so heavily on AI now that we’re in some kind of self-fulfilling prophecy where the pursuit of AI advancement justifies sidelining critical global challenges like climate change.
A case in point is former Google CEO Eric Schmidt. He thinks it’s time for us to fully invest in AI infrastructure because climate goals are too lofty to reach anyway.
1 Like
I find myself reading that, nodding along, and then coming to the end and going “Yes, and?”
Again, name a “new technology” that DIDN’T overpromise and underdeliver, that some people didnt overinvest and end up eating losses, and that people thought would consume the world’s resources.
I’m sure the WWW didnt have those sorts of problems/concerns, right…?
Right…? 
3 Likes
Eek. What a way to view the situation!
Haha, that’s another story. Much of our life has become dependent on on this very fragile infrastructure, which makes me quite sad. There is a strong push to get rid of cash where I live, which is insanity. I recently was in a supermarket with the week’s shopping when it was announced that systems were down so payments could not be taken. People were leaving trolleys in the aisles and walking out in a bad mood. I just went up to the checkout, paid in cash and walked out smiling. I got a lot of dirty looks from other customers, though!
1 Like
I’ll be honest and say I stopped carrying cash years ago. I’ve never run across this situation, but i’ve also never been so desperate to buy groceries that if it happened i’d be overly put out. That said, I also live in a place where walking out of the grocery store would mean crossing the road to the other grocery store and checking there. It’s an annoyance, but not a disaster. Which is much like i’d imagine AI suddenly “collapsing” would be. If ChatGPT stopped existing tomorrow we’d… keep doing what we were doing for a decade before it existed…?
This is obviously a small-case scenario, but it’s when it happens on a larger scale that the real trouble starts. We’ve increasingly based our infrastructure on very feeble foundations. One cheeky Chinese cargo ship can take out Europe’s internet, for example. And so many essential services depend on these systems now … making us very vulnerable to attack or disaster. We kind of rely on experts and politicians to ensure there are backup options, but in reality they have their heads in the sand just as much as we do.
At the moment, the disappearance of AI wouldn’t be much of a problem, but if we get all excited and start to build more critical infrastructure on top of it, well …
Speak for yourself. I would be furiously Googling for how to center a div.
4 Likes
Isn’t that what you’d have done before GPT (over the last ten years? 
)
Anyhow, here you go! https://www.sitepoint.com/css-grid-center-element/
2 Likes
Sand wasn’t my first thought!
1 Like
Exactly so. I still do it, even with GPT. So for me, nothing changes
2 Likes