It's interesting to me

[674169d48f3affox]

Yep, I think even if we factor out some of the oversized rosy tinted nostalgia glasses and some age related excitement, it overall was a much nicer time to live in for people from rich western countries. I'd go so far to say that 9/11 still feels like a turning point to me and probably many people who lived their youth in the 90s - a somewhat wilder yet more "innocent" and overall more optimistic era. I guess you can partially also "blame" the internet becoming an omnipresent part of daily life for changing that.

Hate to turn into one of those "everthing was better before"-guys but I genuinely do feel like it in recent years.

Edit: Of course, objectively it is somewhat perverse for spoiled westerners with relatively high living standards to complain about things "not feeling nice" anymore when at the same time there are large numbers of people living under much harsher conditions (and one might be closely connected to the other too). Still, the depression is real enough for people who believe that their foreseeable future is becoming far bleaker than what they are used to. It can certainly help to reframe things to a bigger picture but it's a mistake to think that people with comparatively high living standards subjectively suffer less - or that it could be avoided by contstant reframing.

[674169d48f644voodoo47]

well, we were still pretty much slaves over here just 30-something years ago (sure, there was no official owner that would hold a slave certificate with your name, but I don't care, if you can't move, choose your school or job unless whoever is in charge allows it, then you are a slave), so if 100 years is too much, then sure, lets do 35, still the same thing.

so unless you have a man pointing a gun at your head, you can always roll up your sleeves and make your life less miserable. it's difficult, and you don't want to do it? sure, your choice, but don't complain then.

"I have no choice, I have to work this crappy job because xyz" sorry, I just don't buy that.

[674169d48f7a9Pacmikey]

.

[674169d48f987voodoo47]

well yes, if your situation is "komrad, you now work in the lithium processing plant or we shoot you, and send your mother the bill for bullet", then you obviously can't do anything, but that's not what is going on in the civilized world. yet.

and yes, fixing your life is difficult, it requires willpower, discipline, and hard decisions, and it's going to be pain, and everything is probably going to explode in your face a couple of times, and you will have to start over. but still, if I think about what my grandparents had to deal with.. well, suddenly things don't seem so bad. makes the "but average miserable man had it somewhat easier a few decades ago" argument seem kind of weak.

anyway, either do it and enjoy whatever little success you may achieve, or don't and live in the hell of your own making, at least knowing you made the choice yourself. that's the deal, a bad one but half the planet would take it if they had the choice. they do not.

[674169d48fbe1Nameless Voice]

I think it's more of a case of some things get better and other things get worse.  So overall, your quality of life might be similar, or might even have improved, but the sad thing is that if humans weren't quite so horrible, it could have been a lot better.

A couple of decades ago, a man could make enough money to house and support his family, while women generally couldn't work most jobs.

Now, women are free to work too ...  but it takes two incomes to afford to support a family.

Before the industrial revolution, the amount of work people had to do varied by season and demand.   People in farming communities had a fair bit of free time when it wasn't sowing or harvest season, artisans and craftspeople would work when they had orders to fulfil, etc.

Now, everyone has to work 40+ hours a week, regardless.
But they get to buy lots of amazing devices with their money, which could only have been dreamed about a century ago.


Also, this of course conjecture.  Even historians don't know exactly what an average person's life was like hundreds of years ago, so ignorant me may be completely wrong on some of this.

[674169d48fd0evoodoo47]

no, that actually sums it up nicely.

it is what it is, and all you can do is make the most out of whatever situation you are currently in. what I'm trying to add here is that you should perhaps appreciate that nobody is going to shoot you/ship you off to gulag while/because you are doing that - a relatively new thing not all people can enjoy.

[674169d48fde9Pacmikey]

.

[674169d48ff24voodoo47]

simple - it's hope, most people need it to go on.

also there is no such thing as existence without suffering, so that entire argument about bringing new life into this world being unethical is just completely flawed. I mean (again) sure, if you are in a gulag, then perhaps don't spawn kids there, but any other place, well, you will need new people to continue trying to make things less of a hell once your time is up.

the only way of ending suffering is to erase all existence. wanna go that way?

[674169d48fffaPacmikey]

.

[674169d49012fvoodoo47]

fair enough - I don't really care too much for the real world, so whatever ends up happening, I'll be just grabbing popcorn either way.

not too big on the hivemind deal - I don't really want to know what is going on inside peoples' heads, thank you very much. and they probably wouldn't want to know what is going on in mine.

[674169d490272Nameless Voice]

What I don't get is how many people have kids and then don't care about doing anything about the huge upcoming climate catastrophe that will make the planet, at the very least, much less pleasant for those children than it was for our generations.

[674169d490a89sarge945]

@sarge945 You're still not convinced about AI in games? Watch this and tell me if you've changed your mind.

https://www.youtube.com/watch?v=d6sVWEu9HWU

Nope, haven't really changed my mind.

While the ability to generate dialogue is somewhat impressive on it's own as a technical feat, the dialog is (as expected) hollow and empty. It's essentially as if someone ad-libbed quick answers to questions without any regard for if they actually contribute any worthwhile information to the conversation.

As I pointed out before, it's great for generating data that fits with the initial dataset. But it's awful for doing anything that requires real intelligence, because that's not how machine learning models work. So of course you get a coherent but meaningless conversation.

I can see "dynamic AIs" being a passing fad in games. I think gamers are going to realise that endless dialogue with an NPC has little to no value outside of "immersion" gimmicks if it doesn't actually affect anything or present any worthwhile information. Then people will go back to preferring dialogue trees, because at least with those you know when you've exhausted all the useful information.

Ironically, games used to have "type to talk to NPCs" style dialogue systems, and they largely got replaced with wikipedia dumps (in games like Morrowind), and then dialogue trees.

This is yet another case of someone doing something cool (and it is undoubtedly cool) with no regard for it's actual value or practicality.

The only real value game developers are going to get out of this will be adding a small amount of dynamic flair to conversations. NPCs won't be able to generate new, interesting topics because ML sucks for that, but they can at least change up the way they talk about existing things, so the conversation can possibly flow slightly better. Speech generation might be especially useful for developers to be able to "slot in" different information, like "Hey, I see you're wearing the <insert players armour here>" without having to record separate lines for each item. But that's a small benefit for essentially either thrashing your CPU or hooking up to an external server.

[674169d490bc6Pacmikey]

.

[674169d490f1dsarge945]

Yeah, I would wager there are already some good speech synthesis algorithms that don't require ML which plenty of old games used to great effect. There were also some not-particularly-good ones, like the ones based purely on keywords.

Part of the ML hype train has been getting things we can already do reasonably well with a simple algorithm, and re-doing them badly with a complex Machine Learning model rather than improving the algorithm. When the ML model falls short of the traditional approach or produces wrong answers, we chalk it up to "the tech is in it's infancy" without any tangible reasoning for how this issue will be resolved, so nothing gets done. It's a perpetual, self-referential cycle where the technology is simultaneously impressive and "going to change the world" while being feeble, slow, and unreliable under the basis of "but it's not there yet".

I remember when Dominos advertised they were using AI to "optimise delivery times" (Side note: Notice how articles talking about it always fail to mention what it is they are actually doing or what tangible benefits they are seeing outside of a vague description? It reminds me of all the crypto articles talking about how "Blockchain will improve your business" but giving no tangible reasons why) by "predicting when a pizza will be ready". We don't need AI learning models for that. Plenty of other places can predict when food will be ready with a high degree of accuracy (far more accurately than I have seen from dominos) by using simple sensor readings or using heuristics based on cooking timings and customer numbers. Dominos has a history of using technology to cover it's own incompetence, like the DOM Pizza checker. Not only does it do a horrendous job (I consistently get pizzas with severe issues), but that's something an employee should be spending 3 seconds doing when putting the pizza in the oven and when taking it out. Dominos has gone AI-mad, and are looking for AI solutions to problems that are relatively simple, and wasting a lot of money in the process.

I can't wait for someone to "train an AI" to sort a list of numbers and then write a long paper and create a website about how smart they are and how revolutionary this breakthrough is. And I have no doubt it will do a slower, less accurate job at sorting numbers than a quicksort, with the caveat that "more research is needed".

People are already asking "How many a's are there in in Hawaiian" and the models are getting the answers wrong. Hmmm.....if only there was a fast, efficient algorithm for determining the number of characters of a certain type in a string.

[674169d491040voodoo47]

..anyway, facebook today - people with horrible birth defects, naked anime chicks, and morons who scoop butter with a sieve.

whatever.

[674169d49138ficemann]

I read all that as: "AI is good because people don't deserve to be paid for their work anyway."

Try working in the education sector. Not being paid for hours of work done is rampant there. For example an assignment may take an hour to mark but nope "we think that should only take 15 minutes", unfair class unpaid preparation time required, admin work required etc etc.

[674169d4914fdvoodoo47]

it's more like "if someone invents something that will make your work worth (next to) nothing, then that's what you are going to get", and that is going on ever since the wheel was invented (yaay wheel, we can move stuff around much more easily now, but all the people who are in the sleigh business are out of luck almost completely, can only get sales during winter from that point, and only in areas where there's snow).

doesn't matter how hard you work, if nobody wants/needs what you are doing, then you are not getting paid. you don't determine the worth of your product (isn't that right, Matrox, Ageia?), the (potential) buyers do. well, mostly, but you get the point.

[674169d491617Nameless Voice]

I looked at a review of a laptop on YouTube the other day.  Now it's started suggesting video reviews of random other laptops.

Feels a bit like the AI logic fail where, if I buy a processor on Amazon, it then keeper suggesting that I want to buy more processors.

[674169d49197ffox]

voodoo47The "wheel vs sleigh business / Can't stop progress"-analogy is a flawed way of thinking and a euphemism. To me it's more a case of "the economy is already broken for many people, let's try to pro-actively avoid that new technology XY is screwing the majority of people, aka the current "working class" as well as the shrinking "middle class" even more into a total disaster that wouldn't even work for the elite soon enough." Yes, this is me being unreasonably hopeful that in the end enough politicians realize that there must be a limit to the madness, I know.

[674169d491abfvoodoo47]

it's just an observation. it is what it is, and we'll see the result soon enough, I'm guesting.

[674169d491bb0Pacmikey]

.

[674169d4920f2sarge945]

it's more like "if someone invents something that will make your work worth (next to) nothing, then that's what you are going to get", and that is going on ever since the wheel was invented (yaay wheel, we can move stuff around much more easily now, but all the people who are in the sleigh business are out of luck almost completely, can only get sales during winter from that point, and only in areas where there's snow).

doesn't matter how hard you work, if nobody wants/needs what you are doing, then you are not getting paid. you don't determine the worth of your product (isn't that right, Matrox, Ageia?), the (potential) buyers do. well, mostly, but you get the point.

They all had this idea in their head that the harder you work the more success you'll have. No.

Who gives a shit if you're an expert at stocking shelves and scanning inventory? You are as replaceable as the keyboard on my desk. There is no point paying you more money if some random off the street can do your job at 70% efficiency with no experience.

You both understand the issue, most people don't. People seem to have this amorphous view of work, as if doing something (regardless of value) automatically entitles you to compensation, as if the creation of value (either for you or your employer) doesn't factor in.

You don't earn money for working, you earn money for creating value. For most people, these two things go together - they get a job doing "work" that may seem pointless to them, but which their boss considers valuable, and that's why they get paid. If that value disappears, there's no point doing the work.

But this issue existed long before AI and isn't really related to AI. So many people are already doing bullshit jobs that AI "taking them all" wouldn't change much - large sections of the economy already shouldn't have jobs because they create no value, and are only employed because of the incompetence of their managers.

[674169d492284fox]

It boils down to the simple question of what you prioritize: people or profit/efficiency. I'm not necessarily a people lover but I sure like those who put peoples well-being before efficiency/profit/technolgical progress better. It's not a binary choice of course, the actual goal is to keep a healthy balance and constantly re-checking our motives - and also thinking ahead enough.

Also, kind of a key question: as government, what would you do with the newly unemployed people if automation was getting introduced everywhere it can do work cheaper an/or more efficient? You think the majority of these people, currently working away in factories will all be able to re-educate themselves to do jobs that are can't be automated yet and are actually needed in large numbers? Provide them with wellfare options to survive or just leave them by the wayside`?

[674169d492621sarge945]

It boils down to the simple question of what you prioritize: people or profit/efficiency. I'm not necessarily a people lover but I sure like those who put peoples well-being before efficiency/profit/technolgical progress better. It's not a binary choice of course, the actual goal is to keep a healthy balance and constantly re-checking our motives - and also thinking ahead enough.

The issue is that even the most die-hard communist understands that nobody should be doing pointless work. This isn't a "people's well-being" problem.

We aren't seeing a new luddite movement. People don't want to "get rid of automation", regardless of their politics, and people who are trying to make this be about capitalism vs AI are arguing in bad faith - the actual argument here is if we should automate things that can be automated - a question which is essentially a no-brainer when it comes to anything else.

If automating factory work is OK, then so is automating art. In both cases you're taking away jobs. If the factory workers recovered, so will the art sector.