this post was submitted on 23 Nov 2023
122 points (100.0% liked)
Technology
37727 readers
637 users here now
A nice place to discuss rumors, happenings, innovations, and challenges in the technology sphere. We also welcome discussions on the intersections of technology and society. If it’s technological news or discussion of technology, it probably belongs here.
Remember the overriding ethos on Beehaw: Be(e) Nice. Each user you encounter here is a person, and should be treated with kindness (even if they’re wrong, or use a Linux distro you don’t like). Personal attacks will not be tolerated.
Subcommunities on Beehaw:
This community's icon was made by Aaron Schneider, under the CC-BY-NC-SA 4.0 license.
founded 2 years ago
MODERATORS
you are viewing a single comment's thread
view the rest of the comments
view the rest of the comments
Sorry man but control over your own image, voice etc is not dumb, why should giant corporations be allowed to replicate and use it to make money without your consent. The fact that it can be so easily automated now just makes it worse. I’m not talking about copyright though just control and privacy. With regards to DNA I’m saying the exact opposite and there we probably agree, companies should absolutely not be able to patent or copyright genes. They shouldn’t be able to use them at all without explicit consent.
I agree with you 100% but it’s hard to convert others.
Same with art - AI mimicking your art after looking at all art is no different than a human looking at all art and having a style that mimics their favorites.
Same with imitation of face and voice.
Except that it is, categorically, different. AI doesn't "learn", it builds associations between data it samples. Incorporating data from the source itself is how these algorithms work, then they reproduce these pieces with permutations applied.
LLMs are easier to explain, so I'll use one as an example. The idea is that if you use particular words in order, that exact ordering of words is given higher weight in the model by a linear association between those words following each other in sequence. When you ask an LLM to "write like Author X", it can do so (partially) by pulling the weights it generated from that authors' works.
This is fundamentally different from how our brains learn and function. We can't hold databases of billions of pieces of information in our heads and compare them all in real time. It's not really comparable at all except as an inaccurate metaphor.
Edit: Too many replies to respond to them all, but our brains don't do linear algebra on matrices with billions of elements. Our brains work in fundamentally different ways. Conflating the two is a gross oversimplification and is incorrect. That was my entire point.
You’re complaining about scale, and pretending it’s a fundamental difference. It’s not
You have a severe misunderstanding of how your own brain works and why we call them neural nets in the first place if you think otherwise.
That is also exactly how I see it. Do you think the negative view is due to some primal jealousy? I don’t know how else to describe not liking something/someone because it’s better than you.
Perhaps it is from a viewpoint of sports, where performance enhancing drugs are frowned upon.
We don’t get mad at calculators anymore, but we did at one point. There was quite a large movement to ban them in schools. Isn’t this a similar thing on the “creative” side?
I do think it’s the jealousy, the fear of being replaced, but also the pride of thinking of ourselves as somehow special and important.
We’re not.
We’re dumb fucking monkeys who learned to sometimes not be so dumb, and then a bunch of us forgot we were pretending.
The real lesson in ai is not that they’re getting super complex or sophisticated, but more us realizing the limitations of our own cognition, and hopefully finding ways to extend it.
You’re spot on about calculators. It’s really just that our schools and schoolteachers are unable to evolve, just like with the insistence that cursive is still a needed skill. Hopefully it won’t take a generation or more to update the educators mindset to taking advantage of the tools available, instead of shunning them.
You’re right. It behaves exactly like we do. And yes, it is at a much grander scale.
Is something ethically, legally, or morally wrong with a computer that does what we do, but does it better?
Because it wasn’t yours to begin with
Biometrics belong more to humanity than any individual person
Your argument would make it so even facial recognition would be illegal, because they scan and use your facial info without consent
Same with drivers license databases
You can’t say “this particular use of this existing practice bothers me, everyone else needs to change now so I feel better”
Rules on these things need to be consistent, and if they shouldn’t be allowed to use unique information that you consider yours without your consent you’ve just eliminated advertising, security checkpoints, drivers license pictures, filming cops, and a million other things both good and bad that all rely on using your likeness without your consent.
On that we completely disagree, I would argue they are some of the only things that are intrinsically yours.
So police can’t keep images of criminals faces (recreated and distributed to every cops computer in the nation through computer automation and often ai) without their consent?
And a private company can’t set that up and sell it to cops for profit?
Because I have some terribly bad news for you…