This seems right.

Marco Rogers

I usually sound cynical about these things. And I know people can sometimes interpret this as saying there's nothing we can do. That's not what this is. I'm saying the opposite. I think there are lots of things we can and should do. Because we can learn from the history of what happens with disruptive technologies.

But I also don't think we can put the genie back in the bottle so to speak. We do have to deal with the reality of what has been unleashed.

Dave Alvarado

@polotek I don't think the genie goes back in the bottle, but since asking the genie a handful of questions uses 2L of water and literally nobody is breaking even, much less making a profit, the genie might atrophy and die before too long.

Tom Bellin :picardfacepalm:

@polotek My biggest frustration with the current moment is that the moniker "AI" is being slapped on any data processing more sophisticated than an Excel spreadsheet.

What your friend is describing is what the mapping industry went through ~20 years ago.

The difference now is that no one called Google Maps "AI." It was sufficiently compelling to describe what Google Maps actually did.

Marco Rogers

@dave that's a nice thought, but I don't think so. I think the return on the investment looks different than people assume. But it's not going to die. This stuff is actually useful. It's just that the way it was brought into the zeitgeist is not an accurate reflection of it's usefulness. And yes, it is wildly inefficient today, but that can change over time. I don't think the people driving this are stupid.

Marco Rogers

@tob this is interesting. How did people refer to the disruption of Google maps technology back then? What was the conversation?

Dave Alvarado

@polotek I don't think the people driving it are stupid, I think they're either sociopathic or in the thrall of FOMO.

I completely agree that bullshit engines are *useful*, but right now it's a race to see if they can get efficient enough to be profitable before they go under.

And don't forget that OpenAI hasn't even begun its legal battles. They have already conceded that they they were violating copyright. They gave up that point when they started licensing content from Reddit, etc.

Marco Rogers

@dave I'm not misunderstanding you Dave. I know a lot about the tech. I disagree with your assessment. Yes it's messy. But these companies have enough money to ride through these issues and make it work.

For the record, you are correct about sociopathy and FOMO. It just doesn't change my assessment of the likely outcomes. I would not bet on this dying out.

Tom Bellin :picardfacepalm:

@polotek I wish I could give you a more insider view of it. I knew a girl who knew a guy who worked for a map making company.

IIRC, he was fatalistic about it from the couple of times I talked to him. To paraphrase: "My bosses are idiots. Mapping technology has been getting better for years and my company just didn't react and now Google owns everything."

I also think he said it was overall a good thing because the map makers were basically operating as a cartel.

Marco Rogers

@tob fascinating. If there's one thing we've learned, it's that protecting a monopoly is a full time job. Not a lot of room for innovation.

Jason Martens

@polotek I’m not sure about the genie argument. Unlike many other technology shifts, AI is fantastically *expensive* to operate. Not just to create the models but also to query them. Of course if businesses feel they can replace humans, then it will look cheap. But the cost involved seems to make this shift different from others.

Marco Rogers

@bitmaker thats not a complete argument. Yes the tech is expensive. But these companies have massive war chests. The question is whether they can be made less expensive and more profitable before the money runs out. I’m betting they will.