Company says it will replace creative workers with ChatGPT-like generative AI

[Gastec]

I agree that nobody is "entitled" to a job but I do believe that private corporations should be heavily-regulated because, as we've seen, they're extremely prone to making decisions that damage the economy, the people and the planet.

I don't know what country you mean so I'll make a guess... Sweden?

EDIT: Your speckled hen reference actually makes me think more like Slovakia or Slovenia.

It's the other one, down south.

 

[Gastec]

If AI can do everything better than maybe we have to reevaluate how our economy works

This is what bothers me, that they call ChatGPT and the other chatbots Artificial Intelligence when they are clearly not. These are just apps, software tools. An AI would have a mind of its (her, his, them, apache?) own. It would be like a flesh-less human, unless it is Ship. If it is Ship then it will ask us to learn how to WorShip and then we are done for!

 

[tipstir]

I like AI have few here my favorite is HEY GOOGLE. I had tested CHATGPT vs the recently BARD from Google. I like the Bard better. To me one simple question was to see if it really new the answers to questions about helping people. There were a few times I wanted to see if it could really access and figure out what I was trying to ask it. In the end BARD said I am glad to be your friend. Wow! But sometimes HEY GOOGLE can do the same thing. My house is SMART using HEY GOOGLE to control lighting indoor and out, all the fans, now all the Smart TV and finally got SMART cleaning robot (AKA SALLY) to finally work.

This me I am real TECH we had guy on the last project kept on telling us other TECH that we weren't real TECH like he was. LOL

Before AI I was using RF module boxes with AC plug from former Radio Shack. Then after that ZWAVE wireless boxes. The AI Alexa plus fire sticks, then Hey Google Android TV Boxes like Onn 4K Streamer. Got so many things controlled. Here...

 

[Shear]

This is what bothers me, that they call ChatGPT and the other chatbots Artificial Intelligence when they are clearly not. These are just apps, software tools. An AI would have a mind of its (her, his, them, apache?) own. It would be like a flesh-less human, unless it is Ship. If it is Ship then it will ask us to learn how to WorShip and then we are done for!

I think what you're talking about really is consciousness, not intelligence. Those are two separate things. You are right that GPT and other chatbots aren't AI but I don't think they have to have a mind of their own to be considered AI. I think the second it can discover new things from the data we give it and come out with valid new theories and discoveries it's an AI whether it's conscious or not.

 

[jevojel337]

"If I need to research something, lets say the "Top 10 things todo in New York for Couples" I can save up to 2 hours by simply entering the question in chatgpt rather then to have this checked out myself."

Except that if you Google it, you will already find several articles on that topic, so no need for ChatGPT at all, right?

 

[gamerk2]

I sure am. Just look at where we are. I have to assume that you're being sarcastic.

Sarcasm was implied, yes.

 

[VectorVortec]

I don't think they have thought the whole idea through.
AI generated content cannot be copyrighted or trademarked. That means all of the content they create can be used by anyone.

I don't think they have thought the whole idea through.
AI generated content cannot be copyrighted or trademarked. That means all of the content they create can be used by anyone.

It probably can't be used by anyone. AI is all about plagiarizing the work of others. The law suits have already started.

 

[someOtherGuy]

Yes, their (our) political agenda is to discourage bigotry and to accept people for who they are. That's been a very challenging topic for some people. When you say "creative" it sounds like you mean "liberal creative", right? Take Joe Rogan, for example. He's a creative, but he's also anti-vax, a science denier, and a racist. Frank Herbert was a creative.

The main problem is that those are "opinions", your opinions, but then "you" and "your team" think that it should be the law. BTW, you're into accepting people for who they are except when they think different than you in which case you have a lot of negative stuff to say about them (ref: the cited text); also, the "press" (your team) had to "apologize" (although they didn't) for attacking "anti-vax" Joe Rogan. This is so layered... are you aware that "the vax" is not a vax, not in the classic sense (they had to redefine "vax")?

 

[someOtherGuy]

What if the chatbots lose access to the Internet, will they still do the job faster, in time, all the time? You know, we humans can do all kids of things, from small sabotages to all-out civil war if they push as hard enough; and guess who's heads will be under the guillotine in the end?

AFAIK the ChatGPT model was trained with "Internet content" up to a certain date. I'm not expert at all but it would seem that the model "learns" with the input, if the input is "the Internet" then it needs connection, otherwise it will learn only with what's left (your input?).

Even more worrisome: what if you could "publish" fake stuff, massively, and that is used to "teach" the model? What if you use the output of one model, which is mostly BS on a nice package (not precise or truthful) to teach another model? How many iterations later would you have a clueless ChatGPT?

 

[someOtherGuy]

"Creative workers", an euphemism for all workers. The propaganda I've been hearing for decades from all political parties, was that they want to CREATE more jobs, not to destroy jobs, at least in Europe. Many European and from over the pond corporations have received lots of help, legal facilities and look-the-other-way's from the State and from the Law, the kind a simple individual who wants to open a small bussiness doesn't get, precisely because they came with the promise of creating lots of jobs after opening their factories, supermarkets, office buildings and whatnot.
Now, after getting all that help to fill their pockets with lots of money, surprise mechanics Pikachu! They have the nerve to tell us that they no longer need human workers? And we have to what, just shut up, "just buy it" because "it just works"?

Those "incentives" should be illegal: you even have a good environment to build your business or you don't, no matter how much money the government throws at you. I would say the main issue is thinking that "government" would solve the problem, and if anyone needed a boost that would be the small business, although having credit available should be enough, most of the time.

 

[someOtherGuy]

For a Technology that will hallucinate two non existent norms for every correct one it lists in a given field people sure are expecting a lot from this tech.
I asked it to put some lists of relevant European norms for work and seriously I had to abandon the **** it GPT gave me. Numbers all wrong, non existent standards, just BS for what is an incredibly simple question.

It’s way, way too prone to being absolutely stupid in all the wrong places to be capable of replacing a human. A human I at least know which parts it finds hard when I’m doing quality control… the AI is stupid in the most absurd of places.

Don’t get me started on the art ones… ask them to give you an action scene or something with lots of people, I dare you.

We are not there yet. It’s semi interesting research, and fun to use at home. Trust it for work? Yea bugger off.

The best summary that I've found around is: it's a professional bullsh!ter, it's not built to be precise or truthful, just to "look like" a human. That's why is not great for science, but it is for art, because it's basically the human creativity algorithm: combining known stuff. The old brute force method would be: giving 1000 typing machines to 1000 monkeys and waiting until they write Hamlet, this GPT method is kinda faster, at the end of the day the beauty is in the eye of the beholder, you just need to check what a "modern" art piece is.

 

[someOtherGuy]

It probably can't be used by anyone. AI is all about plagiarizing the work of others. The law suits have already started.

What about fair use and what makes a human the "owner" of knowledge? We've been doing that because we want people to "think" new stuff, but it's obvious at this point that the patent system doesn't work, at all: we grant the monopoly of certain "tech" to a company for decades and then we complain about prices and lack of options (very smart move). For all other "artistic" endeavors, "programming" the fair use threshold into the model should be simple enough (prevent trademark infringement, change the original work enough that can be considered a "new thing").

 

[tommygunz]

All this AI stuff, is scaring me, and one day I think that it is the next great advancement in technology, just like coal, screens, and now AI

 

[wizardB]

Quite honestly the sooner patents, copyright and trademarks are dead the sooner real creativity can start again. The patent and copyright system has done far more harm to real progress than any other single thing.

 

[Gastec]

AFAIK the ChatGPT model was trained with "Internet content" up to a certain date. I'm not expert at all but it would seem that the model "learns" with the input, if the input is "the Internet" then it needs connection, otherwise it will learn only with what's left (your input?).

Even more worrisome: what if you could "publish" fake stuff, massively, and that is used to "teach" the model? What if you use the output of one model, which is mostly BS on a nice package (not precise or truthful) to teach another model? How many iterations later would you have a clueless ChatGPT?

Normally I would not be so revolted by ChatGPT (and the other AI bots) but I know how Corporate Commanders made these bots work: by using Big Data, all that user data that Google, Facebook, Microsoft and all the other, hidden or less known entities appearing in the list of website "cookies" have collected from all of us over the years, most of it without our consent or outright illegally.

 

[pcnthuziast]

When the working class is no longer of any utility to the rich and powerful, what happens next? They'll be seen as useless and a burden. Think their answer will be to take care of them or arbitrarily create purpose for them? Not likely. Much more likely they'll look for a way to dispose of said burden.