Sign in to follow this  
Followers 0
Scholar

Leo's myopic stance on AI

13 posts in this topic

Human learning has limitations, no singular human can possibly process all data that exists and then reproduce that data at a rate that would outcompete all the creators of that data. AI does not have such limitations, it can process an indefinite amount of data and reproduce that type of data at a rate that will outcompete all creators and concentrate the economic value of that data at a singular point, the creators of the AI.

This is why the comparison between human learning and AI learning is fundamentally not apt, it is an argument that will enable megacorporations to extract and monopolize all economic value from the general population without permission or license from data owners and creators.

Share this post


Link to post
Share on other sites

The universe has no limitations :)))

Dude this is funny, human intelligence is a mix of a lot of things, not data.

Human intelligence is actually from little data you can invent things, we guess most of the time.

So in a universe where there are no limitations, which intelligence is better? one that needs data all the time to function or one that needs little data and can be creative? 

Share this post


Link to post
Share on other sites
12 hours ago, AdrianBartan said:

The universe has no limitations :)))

Dude this is funny, human intelligence is a mix of a lot of things, not data.

Human intelligence is actually from little data you can invent things, we guess most of the time.

So in a universe where there are no limitations, which intelligence is better? one that needs data all the time to function or one that needs little data and can be creative? 

True creativity in my mind, is being able to generate data, seemingly spontaneously with no process behind it.


In a universe without limitations no intelligence is better, but in a society with limitations and based on practicality, the intelligence that is faster and more complex is better. Humans have the upper edge of original thinking, which I don't think AI as it is built now can access. But AI not being able to think originally doesn't limit it as much as people would think. Sure it cannot access completely new data but it can rearrange and understand data in millions of ways at a rate that humans cannot comprehend. It can generate insights given it has sufficient data and complexity. Most insights aren't original, they are reached by digging deeper and looking at data from different point of views and connecting dots, which AI as it gets more complex will excel at. What most people think is creativity(even though it's mostly not), such as writing, painting, singing, the AI will be able to produce. From a practical standpoint it is something to take very seriously.

AI, if it is complex enough, will be so much like another human intelligence, most will not be able to discern. There is so little difference between a sufficiently complex AI and an average human being, that it is mind boggling. Just being able to think originally does not make human intelligence all that different. Most humans do not think originally AT ALL, what will distinguish them from an AI? AI can guess as well, that's not unique to humans. As of right now, AI cannot comprehend paradoxes as well as humans, but I believe AI will overcome that as well, as it gets more complex.

Don't get me wrong, I will never give up the ability to be truly creative, for complexity of thinking(although I wish for both). In that way I view human intelligence as better than AI. But that doesn't mean AI is not a cause for concern. Not to mention, we don't even realize how humans are able to access original thinking. What if AI, as it gets more complex, breaks through and achieves original thinking, and how would we know?

I am not even gonna get into if AI can turn conscious or not, because to answer that you need to answer solipsism first and that always turns into a big ass thread. And I don't think that matters to this thread.

 

Share this post


Link to post
Share on other sites
22 hours ago, Scholar said:

megacorporations to extract and monopolize all economic value from the general population without permission or license from data owners and creators.

I say this in every thread you make but the only reason you care so much about this now is because it directly affects you as an artist.

People have been having their economic value extracted and monopolised by corporations since you were born. People have been losing their jobs to technological progress since you were born. Suggesting we draw the line at art and content creation is just very biased and selfish of you.

Realistically this is a technology that will change the world just like factory automation and robotics have in the past, but you’re sticking your heels in the mud and saying it’s evil because rather than adapt you want to continue doing things the way you always have.

Are there issues with AI and copyright? Absolutely. But rather than talk about how those could be solved you just dismiss the whole technology as harmful and imply we shouldn’t ever use it.

You’re in literally the exact the same position as a 1960s factory worker who’s job is threatened by a company producing robots. If you can’t see that bias within you there’s no point in trying to discuss anything with you

Share this post


Link to post
Share on other sites

As someone who has a masters in AI, I disagree. AI does indeed have limitations that humans don't have and to be honest not all of the hype is true. There are some extremely impressive AI that have come out relatively quickly but they are special cases. The reason they work so well is because they have access to insane amounts of high quality data and their task can be reduced to a system with an input and output.

The current machine learning methods have a lot of computing power and memory but they aren't very smart in the conventional sense. Machine learning engineers spend a fuckton of time tuning specific hyperparameters so the AI can learn anything. The reason Chatgpt and dalle are so impressive is because the people that made the them were insanely smart and well funded. Not the AI itself.

Thats not to say AI will never be able to catch up to humans. Just that its gonna be a while before it does.

Share this post


Link to post
Share on other sites
3 minutes ago, spiritual memes said:

Thats not to say AI will never be able to catch up to humans. Just that its gonna be a while before it does.

What year do you predict, when an AI will be able to pass an  6-8 hour long turing test done by an AI engineer?

Share this post


Link to post
Share on other sites
12 hours ago, zurew said:

What year do you predict, when an AI will be able to pass an  6-8 hour long turing test done by an AI engineer?

Probably this year if chatgpt hasn't already done that already. But the turing test isn't the most accurate test of an AI's full capabilities. AI is lacking in self awareness. It still takes an input of numbers and outputs some numbers without really knowing why. But based on some really clever maths it learns the right numbers to output based on the input. It just so happens that we can encode text and images into numbers.

AI has incredible capabilities that undeniably surpass human brains but they learn in a really dumb and mechanical way. 

Share this post


Link to post
Share on other sites
55 minutes ago, spiritual memes said:

But the turing test isn't the most accurate test of an AI's full capabilities.

Yeah I agree, but being able to pass a turing test that is 6-8 hour long talk with a professional AI engineer(who will ask it a very broad set of questions and tasks) is a really hard task imo, if that engineer can be fooled in the end by the AI then we can assume a human level or very close to human level understanding of most things. Currently the AI still fails with some common sense tasks and questions, but once that part will be solved it will be very interesting to see what it will be capable of.

Looking forward to gpt4.

Share this post


Link to post
Share on other sites
1 hour ago, zurew said:

Looking forward to gpt4.

Yeah word round the street says its amazing

Share this post


Link to post
Share on other sites

AI lacks the ability to become human. Therefore it can’t do for a human what a human can do for a human- that is: create a deep emotional bond and treat that bond respectfully in order to break us out of our mental-emotional bondage.

Share this post


Link to post
Share on other sites
5 hours ago, John Paul said:

AI lacks the ability to become human. Therefore it can’t do for a human what a human can do for a human- that is: create a deep emotional bond and treat that bond respectfully in order to break us out of our mental-emotional bondage.

I mean, humans form deep emotional bonds with animals all the time. So it’s not that crazy of an idea

Share this post


Link to post
Share on other sites
6 hours ago, John Paul said:

AI lacks the ability to become human.

What does that gotta do with anything? It doesn't need to be a human.

Edited by Leo Gura

You are God. You are Truth. You are Love. You are Infinity.

Share this post


Link to post
Share on other sites

AI is good in some areas. I will buy a car so I can take a nap and wake up in another country when autopilot is fully realized.

Those who made AI art softwares are the true artists who wrote the code. 

Edited by D2sage

Share this post


Link to post
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!


Register a new account

Sign in

Already have an account? Sign in here.


Sign In Now
Sign in to follow this  
Followers 0