> AI is a world-changing technology, just like the railroads were
This comparison keeps popping up, and I think it's misleading. The pace of technology uptake is completely different from that of railroads: the user base of ChatGPT alone went from 0 to 200 million in nine months, and it's now- after just three years- around 900 million users on a weekly basis. Even if you think that railroads and AI are equally impactful (I don't, I think AI will be far more impactful) the rapidity with which investments can turn into revenue and profit makes the situation entirely different from an investor's point of view.
Railroads carried the goods that everybody used. That’s like almost 100% in a given country.
The pace was slower indeed. It takes time to build the railroads. But at that time advancements also lasted longer. Now it is often cash grabs until the next thing. Not comparable indeed but for other reasons.
> just three years- around 900 million users on a weekly basis.
Well, I rotate about a dozen of free accounts because I don't want to send 1 cent their way, I imagine I'm not the only one. I do the same for gemini, claude and deepseek, so all in all I account for like 50 "unique" weekly users
Apparently they have about 5% of paying customers, the amount of total users is meaningless, it just tells you how much money they burn and isn't an indication of anything else.
Sometime you have to force the trickle down economy a bit, these people are destroying my industry I might as well cost them as much as possible before I have no choice but to move on.
It's also literally 0 effort, click > sign out > click > sign in. It saves me $200 a month, that's not too far from half of my rent
I can understand the spirit, though this reinforces my impression that the product is so good that people jump through hoops to use it, even if they hate it in principle. If they suddenly cut off any free access to it, how much would you be willing to pay per month to keep using it? One dollar? Ten? Twenty?
Also, maybe I'm missing something, but no amount of free accounts on ChatGPT gives you what you get with a paid subscription, especially with a $200 one; and there's paid plans from just $8/month.
I like movies, I still torrent them, if tomorrow a police officer is being my back 24/7 I will just stop torrenting, but I still won't pay $20 a pop to go to the cinema.
> Also, maybe I'm missing something, but no amount of free accounts on ChatGPT gives you what you get with a paid subscription, especially with a $200 one
These days I'm mostly running opus 4.5 through "antigravity" and I'd rather become a potatoe farmer than give $8 to Altman
It's a really tiresome conversation, and I should just stop replying, but...
If you have to stop torrenting it doesn't mean that you have to pay $20 per movie. There is a price >0 that you're willing to pay to do something you love. On youtube there's a lot of movies for 4 or 5 dollars.
I'm also using Claude, both through Cursor (paid by my company) and privately (paid by me, $20/ month).
I'm going to go out on a limb here and say that users who put that much effort into using this stuff for free, using a dozen different accounts, are very rare.
> user base of ChatGPT alone went from 0 to 200 million in nine months, and it's now- after just three years- around 900 million users on a weekly basis.
Doesn't have anything to do with AI itself. Consider Instagram then TikTok before this, WhatsApp before, etc. There is a clear adoption curve timeline : it's going WorldWide faster. AI is not special in that sense. It doesn't mean AI itself isn't special (arguable, in fact Narayanan precisely argue it's "normal") but rather than adoption pace is precisely on track with everything else.
is not correct IMO. Those are two very different areas. The impact of railroads on transport and everything transport-related cannot be understated. By now roads and cars have taken over much of it, and ships and airplanes are doing much more, but you have to look at the context at the time.
It's meaningful because it shows that people like the product a lot, and for a lot of different reasons. There are only few products that can reach such market penetration, not to mention in only three years. As the quality of AI increases, people will quickly realise that they are willing to pay for it as much as they pay for electricity. And the same goes for businesses.
Isn't that akin to a 1990s tech model like CompuServe or AOL? "Let's create this awesome environment where people will want to pay us for this wonderful service, we'll send them a CD in the mail to get them started withh a free month, then charge $0.30/minute. We'll make a fortune!"
Indeed, but in the end they all have to cover their costs. People are already getting real, measurable value out of them and they will be willing to pay for it like they pay for utilities. Though I'm not excluding that the AI companies will manage to create some kind of moat to keep their customers (such as personalisation, memory, etc.).
At this point I'm taking the word "slop" as a sign meaning "I really didn't think this through and I'm just autocompleting based on a gut feeling and the first word that comes to mind".
Well unfortunately for you, it has a precisely defined and well-understood meaning for all those not covering their eyes and ears in denial. Quoting Merriam-Webster:
> Digital content of low quality that is produced usually in quantity by means of artificial intelligence.
Chosen by the editors as word of the year, by the way.
"Slop" in English means "liquid junk, rubbish, tripe". No need to call for Merriam Webster's help.
The point is that AI can produce slop (as people do, too), but it's just silly to imply that everything it can produce is slop. That's just lazy, sloppy thinking.
Sure. I'm fully aware that AI can be useful, especially once we move past LLMs.
However, I do think that the majority (or mainstream) use of GenAI today is indeed not very useful or even harmful. And I do think that something like railroads are more useful by orders of magnitude.
This comparison keeps popping up, and I think it's misleading. The pace of technology uptake is completely different from that of railroads: the user base of ChatGPT alone went from 0 to 200 million in nine months, and it's now- after just three years- around 900 million users on a weekly basis. Even if you think that railroads and AI are equally impactful (I don't, I think AI will be far more impactful) the rapidity with which investments can turn into revenue and profit makes the situation entirely different from an investor's point of view.