In reading this article, I noticed that there is a distinct lack of a specific goal beyond simply to build agent-based models.
This reflects the state of AI research today: it has crossed the state of post-usefulness. In other words, new computing has crossed the threshold of being useful (as in genuinely improving human lives ON AVERAGE) and is now about creative perceived value so that we can spend money on it, merely to develop it further. It's only powered by immediate, short-term gains whose positive value to society don't last when compared to the long term societal detriment.
If we were smart, we would destroy this technology and cease research on it because it will only lead to human beings not having much of a role in keeping society going, so that in turn we will become too independent from each other to care about each other. But the addictiveness of using AI will likely make those who develop this technology blind to this simple fact...
We are like children playing with fire at a gas station.
>now about creative perceived value so that we can spend money on it, merely to develop it further.
>in turn we will become too independent from each other to care about each other. But the addictiveness . . .
>We are like children playing with fire at a gas station.
All this could have been said about the early work on "personal" computers by Xerox which never did do them very much good at all, using 20/20 hindsight looking at what followed after.
>It's only powered by immediate, short-term gains whose positive value to society don't last when compared to the long term societal detriment.
Looks to me like all most of them have is short-term losses, and for many that will be as far as they get.
For some of them also, Societal-Detriment-R-Us, no doubt about it, so good call overall from a reasonable point of view.
Thats a pretty cynical take. The likely overly optimistic take would be that it would provide a level of usefulness of a highly paid and highly competent personal assistant.
If the overly optimistic upside is a quicker, more knowledgeable, higher-performance-information-digesting personal assistant at lower cost than ever before, it would still be out-of-reach for almost as many ordinary people as the top professional human personal assistants are now.
Unless they never manage to reign in the costs, then it will be more out-of-reach than a real human, disregarding the potential game-changing element of real human empathy.
I would expect those that have funds to spare already have both, regardless of the limitations of each.
If you can come up with a possible downside, that would be something to compare it against.
Everyone might not have to be cynical for their opinions of the potential outcome to be either positive or negative.
It doesn't. It's not meant to be a conclusion. But it's a fact of AI: it reduces the need for human beings. That, combined with its lack of usefulness, means we should destroy AI as it only brings us down and makes the world worse.
To be clear, the resentment and public statement is important and relevant IMHO.
> combined with its lack of usefulness
the language capacity of these systems makes them useful to interact with humans. The purpose of that interaction is a question for legitimate social theorists, policy and law enforcement.
An emotional "gut reaction" of anger and disgust is not misplaced, but it is essentially personal, therefore in the category of drama more than policy.
This might be true, but might also have been said by agricultural workers about mechanisation of farming. 30% of Europe was in agriculture 200 years ago[0] - new jobs get created, and even if they add very little value, upward pressure from minimum wage laws means the people with money end up paying for them.
Well, that is why I caution against reasoning by analogy. AI is different and should examined based its characteristic: speed of development and scope of applicability.
Agreed, not "it'll put people out of jobs". That's not guaranteed at all (see also: taxi drivers should be nonexistent by now according to the same prognosticators), and retiring a profession doesn't mean that people won't have jobs.
The name is literally "H". First X, now this. The difference is that (former Twitter) X can afford this. I'm very unsure of the choice of name, regardless of their merit.
Weird thing is that this article (or rather press release) doesn't even acknowledge how unorthodox the name is.
This reflects the state of AI research today: it has crossed the state of post-usefulness. In other words, new computing has crossed the threshold of being useful (as in genuinely improving human lives ON AVERAGE) and is now about creative perceived value so that we can spend money on it, merely to develop it further. It's only powered by immediate, short-term gains whose positive value to society don't last when compared to the long term societal detriment.
If we were smart, we would destroy this technology and cease research on it because it will only lead to human beings not having much of a role in keeping society going, so that in turn we will become too independent from each other to care about each other. But the addictiveness of using AI will likely make those who develop this technology blind to this simple fact...
We are like children playing with fire at a gas station.