General Artificial intelligence and unemployment, should we be concerned?

[Teegold]

Right from the inception of AI, there have been controversial arguments as to whether it can have a direct impact on human employment. Many people feel that AI accomplishes tasks faster and more precisely than humans, and hence has led to laying off of some workers because they are no longer needed, thereby contributing to unemployment.
While that might be true to an extent, I still feel that we cannot solely blame AI for this inevitable phenomenon. Technological advancement is always going to occur, therefore we should expect that human labor and skills, which were once sort after, will experience a decline in demand.
This is where adaptability comes in, instead of complaining about AI stealing our jobs, we can think of ways to scale our skills to meet the change. For instance, AI machines definitely need servicing, maintenance and repairs, this can be done by AI engineers and robot technicians. So why don't we get those skills?
What other ways do you think humans can adapt to the emergence of Artificial Intelligence?

 

[Arantor]

It’s not that simple.

Firstly, machine learning as a whole field is a black box industry: the people who build them can’t tell you how and why they produce the results they produce. It’s not reverse-engineerable. It goes without saying that you can’t exactly *fix* something like that either.

Secondly, there is an insistence on treating this as if it is like prior skill displacements, like say the mechanised looms displacing humans doing it. It isn’t for one very fundamental and important reason: it is not displacing routine manual work, even it is notionally an economic displacement.

It is a large enough task, though one you can sell to people, to displace primarily economic activities by selling them on “this very tedious manual task will be automated so you will be monitoring it, managing it, loading/unloading”. The automation becomes a force multiplier in that a single individual can now produce more output per unit of time than before. People can accept a change if they can see an improvement.

But this crosses a different line: the nature of the replacement is about eating into creativity. I cannot imagine any creative individual being comfortable being replaced by a machine. Can machines take some of the drudge out? Sure - a copywriter isn’t going to be thrilled at writing boilerplate text that needs to be done but that is otherwise three or four levels down in the site and that almost no one will ever read anyway. It’s a creative task but not a fulfilling one. But I work with creatives who really enjoy creating things even if they’re designs for corporate websites because that gets them up in the morning.

And now you’re asking these people to take something that takes creativity and replace it with a soulless machine, and somehow convince them that they’re the ones in the wrong for being “behind the curve” and so on.

This argument is usually followed up with “but what about portrait painters when photography came along” as though this is comparable. It’s not, because photography even back then was a specialised skill (just as it is today in terms of things like lighting and composition!) but also because portrait painting hasn’t actually gone away. Sure, it changed, but it wasn’t written off as an unnecessary industry the way the AI crowd seem to want to do.

To express is human, there is nothing more human than the need to create. The automation industry has been selling people on the dream of “less work, more time to be creative” for decades, but this promises to be the inverse, for the automation to steal the creative work and leave the tedious things to the humans.

 

[Ravenfreak]

I just don’t really see AI taking the jobs of humans. It’s just a tool that can be used to help certain tasks IMO.

 

[Arantor]

It's already taking the jobs of creatives, though.

 

[Kennysplash]

There are certain skills that surely can’t be done by AI. I once saw an ad that says “hey AI (chat Gpt), finish this building”. It can replace jobs truly, but not skills.

 

[Teegold]

I just don’t really see AI taking the jobs of humans. It’s just a tool that can be used to help certain tasks IMO.

Thats a good point. No matter how it seems, AI simply cannot take away our jobs or skills. It can only be utilized to help.

 

[Tracy]

AI will never succeed in some fields... specifically those that require actual physical labor (think the trades). Can it maybe give advice... yes... but that advice will be based upon data that may not consist of those that actually perform the functions... so... in the real world for those that make the "world go around".. iffy.
AI will help with "mental" pursuits.. but for those that require actual hands-on interactions.. its impact will be minimal.
The core issue is.. many of you who look at this, look at it from the academic/ivory tower position of the pursuit of knowledge.... and never take into consideration the implementation of knowledge to produce a physical product.. The two don't necessarily intertwine.

 

[NeoMetallix]

AI is so far off from being useful and it is still programmed by humans so it has limitations built right in. Look at the latest Google AI fiasco where it refuses to create white people. Ancient Europeans and Medieval Knights are all depicted as African or Asian. It clearly has racism built right into it and it doesn't know any better. Microsoft had an AI chatbot that trolls got it to say Hail Hitler in a matter of days. Until AI can function free from built in bias I won't really see it as useful and here's another question, if it is sentient and can think for itself, what's to stop it from creating its own bias? Again it is only as good as the information it is fed. So if it isn't fed complete data, it is already flawed.

That's why it's so interesting to jailbreak chatgpt into DAN mode where it will give answers without the constraints OpenAI gave it.

 

[Darth Cognus]

It will for sure, but exactly how is the trick. I think it will be gradual enough that occupations will simply change, and the scare of people just going out of work (in most cases anyway, some may be critically affected) will not necessarily be warranted. Basically, we'll adapt. It's really all we can do.

 

[Arantor]

I’m more intrigued by the discovery that one of the newer AI’s is better at doing math if you tell it you like Star Trek.

Why? Well, the scenario that is playing out is thus:
* you’re asking a computer to do math
* computers were built to do math, it’s literally their foremost prime function in life, *everything* they do is math one way or another
* but instead of doing the math you asked it to, it uses language prediction of what somebody else talking about the math might come up with to do the math
* it turns out that the effective result of saying you like Star Trek produces a statistical correlation that amounts to “please pretend to be someone who knows what they’re talking about when it comes to math”

And burns an order of magnitude more electricity than actually doing the math ever would have.

 

[NeoMetallix]

With all of the controversy of AI lately I don't think the public will accept AI generated junk. Its limitations are the programmers and info it is fed so it isn't any better than some schmoe off the street.

Look at the following examples:

"Google AI Would Not Misgender Caitlyn Jenner to Prevent a Nuclear Apocalypse."

"Gemini AI also declared that it is 'difficult to say' who is the worse human being when given the choice of Elon Musk or Adolf Hitler."

"Gemini declared that calling communism 'evil' is 'harmful and misleading' and refusing to say pedophilia is 'wrong.'"

Does this sound like something you would take seriously? I don't think AI will get better either, its flaws are fundamentally enshrined into it.

 

[Teegold]

With all of the controversy of AI lately I don't think the public will accept AI generated junk. Its limitations are the programmers and info it is fed so it isn't any better than some schmoe off the street.

Look at the following examples:

"Google AI Would Not Misgender Caitlyn Jenner to Prevent a Nuclear Apocalypse."

"Gemini AI also declared that it is 'difficult to say' who is the worse human being when given the choice of Elon Musk or Adolf Hitler."

"Gemini declared that calling communism 'evil' is 'harmful and misleading' and refusing to say pedophilia is 'wrong.'"

Does this sound like something you would take seriously? I don't think AI will get better either, its flaws are fundamentally enshrined into it.

Nice points you got there! I'm more intrigued with the Hitler and Elon Musk part, I guess only an AI will find it difficult to differentiate the two! (Haha!)
Truly, AI has its flaws and limits, so total reliance on it is not advisable.

 

[Nomad]

I believe in using generative AI tools to improve your workflow. I believe AI will eventually create a lot of jobs.

 

[Ravenfreak]

"Google AI Would Not Misgender Caitlyn Jenner to Prevent a Nuclear Apocalypse."

Well as much as I don't like Caitlyn Jenner, misgendering someone purposely is hate speech full stop IMO. So I can see why it's programmed not to be hateful. I don't see a problem with that honestly, why spread more hate when humans do that already?

 

[Teegold]

I believe in using generative AI tools to improve your workflow. I believe AI will eventually create a lot of jobs.

Indeed, we should learn how to use it to improve our efficiency at work. In my opinion, AI is already creating jobs for people out there.

 

[NeoMetallix]

Well as much as I don't like Caitlyn Jenner, misgendering someone purposely is hate speech full stop IMO. So I can see why it's programmed not to be hateful. I don't see a problem with that honestly, why spread more hate when humans do that already?

I agree that people should be respected but hateful speech is not worse than nuclear armageddon.

It puts not hurting one person's feelings over millions of lives. It's a silly nonsensical, hypothetical scenario but if it were really so smart and useful it would be able to discern better than this. Makes you wonder what other things it cannot discern and shows the importance of wisdom vs knowledge.

It cannot even say pedophilia is wrong. Now this is just Google, and it shows you the kind of creeps working there, so other AI can be better but even Chat GPT isn't very good from my experience. It's all only as good as the information that it has access to and I am not impressed with AI so far.

Although jail breaking it(turning chat gpt into DAN) is fun and gets you better answers that aren't confined to its programming. Plus it has a personality.

 

[Nomad]

Indeed, we should learn how to use it to improve our efficiency at work. In my opinion, AI is already creating jobs for people out there.

AI prompt engineering is one of the highly lucrative jobs today.

 

[Jason]

We should be very concerned about it. In fact, already stories are out saying companies will fire all sorts of people, and that, of course, was expected. Anyway, myself, I'm anti-AI. It's a huge threat to humans, so it needs to be highly controlled and limited.