The AI Thread

Can you elaborate on this?

Well, even if we just use simple math on my ridiculous numbers above.
"Assuming that a 'living income' is $50k USD annually. And assuming 10 billion people. That's $50 trillion annually."

Current GDP is is something like $100 trillion. The numbers aren't exactly comparable, but even if we pretended that we needed a 10x higher GDP to have sufficient income for everyone to retire (we have a LOT of alternative uses for GDP other than lifestyle for some time), that's only three doublings from where we are today. After that, it's literally just choosing to replace low-value labor with automation. The calculations are really hard at the stage we're at, because a large portion of the 'income' in the world is just transfer of assets (upwards). Like, if a city owns 100% of a bridge but then gives up 50% of its ownership in order to settle a debt, the transfer of the bridge hasn't created any new wealth even if we finagle it onto the books somehow as the rich getting richer.
Anyway, bad math aside, 3 doublings is faster than the 10 doublings. Maybe a lot faster, depending on how the exponent for growth plays out.

Robin Hanson is an economist who's mad fun to listen to on this topic. He lists many "Hell on Earth" scenarios in Age of Em, but also posits that doublings in productivity could be measured in time periods that meat sacks have no real conception of.
 
Well, even if we just use simple math on my ridiculous numbers above.
"Assuming that a 'living income' is $50k USD annually. And assuming 10 billion people. That's $50 trillion annually."

Current GDP is is something like $100 trillion. The numbers aren't exactly comparable, but even if we pretended that we needed a 10x higher GDP to have sufficient income for everyone to retire (we have a LOT of alternative uses for GDP other than lifestyle for some time), that's only three doublings from where we are today. After that, it's literally just choosing to replace low-value labor with automation. The calculations are really hard at the stage we're at, because a large portion of the 'income' in the world is just transfer of assets (upwards). Like, if a city owns 100% of a bridge but then gives up 50% of its ownership in order to settle a debt, the transfer of the bridge hasn't created any new wealth even if we finagle it onto the books somehow as the rich getting richer.
Anyway, bad math aside, 3 doublings is faster than the 10 doublings. Maybe a lot faster, depending on how the exponent for growth plays out.

Robin Hanson is an economist who's mad fun to listen to on this topic. He lists many "Hell on Earth" scenarios in Age of Em, but also posits that doublings in productivity could be measured in time periods that meat sacks have no real conception of.
But the idea that we can make human labor unnecessary and everyone can be paid to be idle is a bit fanciful no?
 
Imagine people running the analogous calculation before commercial use of steam or any other tech that changed the balance.
Some new tech will bring an end to the current situation, but that doesn't mean it will be a good end. Things will probably get more chaotic.
Is the system even ready for the needed (then) changes in password/other encryption protection, when new discoveries on primes occur?

It appears that the two main reasons for stability of the system is human inertia due to misery, and human inertia due to happiness - and the latter is vastly rarer.
 
But the idea that we can make human labor unnecessary and everyone can be paid to be idle is a bit fanciful no?
And there's the secondary consideration as to whether we would be more "self-actualized" without work. I know that I'm probably more fortunate than most in my employment, but I self actualize through my work. A good part of this is scope for creativity that my job allows me. But some part of it is labor as such; I feel good about myself when I've put in a "good day's work." And it don't think it would be the same if I at all felt that it was just make-busy work b/c some computer was now doing better what I used to do.

Now, by the time the world achieves any of these things we're imagining (machines doing everything, the rich coughing up 2% of their wealth), I'll be long dead. But on behalf of some future Gori, I'm interested in how these matters will play out: self-actualization in a post-work world.
 
It'd be pretty good if people could earn a living by working only a couple hours per day. Spending half your life earning the ability to stay alive, is only cheered by the worst elements in society.
 
That's a good question. I wonder if there's research on how much of one's day people would like to spend working. I think that for most, it would not be zero hours. But perhaps for many it would be just something like 2 hours per day. It will of course depend on the kind of job they have. That's why I acknowledged that I am probably more fortunate than most.

But your work, Kyr, is writing stories (to the degree that you're able to support yourself doing that). And that's something you also love doing.
 
That's a good question. I wonder if there's research on how much of one's day people would like to spend working. I think that for most, it would not be zero hours. But perhaps for many it would be just something like 2 hours per day. It will of course depend on the kind of job they have. That's why I acknowledged that I am probably more fortunate than most.

But your work, Kyr, is writing stories (to the degree that you're able to support yourself doing that). And that's something you also love doing.
My paid work isn't writing stories, though :)
 
But you've been paid for your stories, no?
And you'd like to make a career as a writer, no?
(And do I feel confident that you will? Yes.)
(Remember the old joke that overnight success takes about twenty years. Calculate that date from the first story you got paid for. My money is on you making your living as a creative writer before that twenty-year mark.)

If scholarship has been done on this question, the way I would want it answered is "If you had the choice of working more or less, in your present job, how many hours per day/week would you work?" I'm putting the question to the forum, for unscientific results.
 
But you've been paid for your stories, no?
And you'd like to make a career as a writer, no?
(And do I feel confident that you will? Yes.)
(Remember the old joke that overnight success takes about twenty years. Calculate that date from the first story you got paid for. My money is on you making your living as a creative writer before that twenty-year mark.)

If scholarship has been done on this question, the way I would want it answered is "If you had the choice of working more or less, in your present job, how many hours per day/week would you work?" I'm putting the question to the forum, for unscientific results.
Paid, yes. But not a living wage/amount. My paid work has been in literary translation and seminars (online and offline) :)
I do have dreams, of course, but realistically one shouldn't aim to make a living as a writer in Greek (very few do, and it's not strictly about the work's worth as much as a market for it). I can always try English, or some other medium for that.
 
Your English is strong enough, Kyr, that you could write in that language (and frankly should, given the significantly larger market).
 
But the idea that we can make human labor unnecessary and everyone can be paid to be idle is a bit fanciful no?

Yes. A lot would need to go correctly. The tech curve needs to beat the environmental buffer we're eroding. The politics would need to distribute the resulting resources properly. The psychology would have to adapt to maintain the sense of productivity without actually contributing to the creation of essential goods.

But we are going to approach the era where it's faster/cheaper to build and train a robot to do an essential task than train a person to do it. And the people in regions with proper capital redistribution are going to be very tempted to ignore those that aren't. I don't feel wrong that '2% to charity' is a reasonable guess.
 
They curtailed bung :(
 
Bing? Did you read the NYT piece about it declaring its love for a particular beta-tester? Telling her she was in an unhappy marriage? All it took was asking it to adopt a Jungian "shadow self." (Goes and searches what that is, learns that it involves disclosing dark secrets, separately searches for what would count as dark secrets). Once released to billions of people, billions of people are going to find billions of different ways to get it to give such creepy answers (cuz that's what people do). For some stretch of time humankind is going to be collectively creeped out by the kinds of answers it gives.

The following is funny:

 
If the rich subsided everyone then they'd stop working which would mean the rich would lose their means of production.

Walmart shareholders need Walmart employees to feel the pressure to goto work.

I owe, I owe, it's off to work I go :dance:


Thanks Harv.
 
Most amazing feature of chatgpt is programing. Just asked it to write a script which would create a folder for every week since January 1st 2022 to current date and then move every archive in the root folder to each folder according to its modification date (I used approximately these exact words, no more, in Spanish)
Without any hesitation it created a little python script for me which worked flawlessly at the first attempt, except for some initial confusion with the year since chatgpt apparently thought we were still in 2022. (solved inmediately after I informed it we were already in 2023)
 
If anything I think professional programers have a problem. (and a huge help too)
 
Tried the same with the new Bing and it said it was a way too complex script and if I wouldn't prefer to learn python starting with something easier as an addition. I answered I didn't want to learn python only wanted the damn script, then a message appeared saying the conversation was over and I needed to move to a new topic.

What such a load of crap Google isn't in a hurry to launch Bard. :lol:
 
Hm, I just used Chatgpt.

I am wondering why I bothered talking to humans up to now ^^

The machine was actually quite good even in special topics - eg asked it about some plots in Kafka.

Obviously it is not thinking, but it does present a very believable (if at times comparatively trivial) perspective.

Will try again, 10/10.
 
Back
Top Bottom