19
votes
EU affirms free will in new AI regulation
Link information
This data is scraped automatically and may be incorrect.
- Title
- Artificial Intelligence Act: deal on comprehensive rules for trustworthy AI | News | European Parliament
- Published
- Sep 12 2023
- Word count
- 953 words
Mods please move this back to ~humanities, ~tech is definitely not the right place for a discussion of what "free will" means in a philosophical vs legal context which appears to be OP's desired topic of discussion
Done!
Edit: Hmm, looks like this post was moved from ~humanities to ~tech. I would say this isn't really a tech post, but more of philosophy discussion.
Under banned applications:
I have questions regarding how one argues for or defends against this rule in the new AI regulations?
If I make an AI that someone asserts caused them to "circumvent their free will" (also, what does this mean?), then can I defend against this by saying that actually there is no free will, all things have been determined, and that they were predestined to do what my AI said?
Could one challenge the notion of free will, by asserting that no thoughts are produced by the self, but that the self only examines those thoughts? There is no "I" producing a will, the will is just innate?
What if there is no "free will", that all human actions have a preceding cause, or are constrained by motivations and social environments?
Hypothetically, if we are put in the exact same situation, of the exact same environment, in an infinite amount of alternative histories, would we make the exact same choice, each time, and is then not free?
What if our will is not free, but is something that is actually a burden on us, something we cannot escape to be free from. "I can choose what I will, but not will what I will."
What if all human action is influenced by divine choice. If there is an all-knowing deity, that knows everything we will ever do, are our choices really free?
What if our brains just follow the laws of thermodynamics? So our free will is just an illusion of natural law affecting our brains. Or our brains are much more complex, and have emergent properties that allow for a kind of free will that is compatible with the laws of physics?
What if we are just brains in a vat, with our will controlled by some external electric impulses?
Maybe you're overthinking this. They were probably thinking about systems designed to objectively manipulate people, such as by identifying stimuli that are statistically likely to result in certain actions by the person in question and then using those stimuli to force an outcome desired by the developer?
They're definitely overthinking it.
There are interpretations of free will that don't attempt to reconcile conflicts of determinism or such and they're merely meant to address more surface level aspects of our behavior. That's certainly along the lines of what interpretations that this type of legislation would go with. Free will at a surface level is merely attempting to explain elements of human behavior and thought and how they intersect, going to the philosophical level is a whole different thing.
You could argue things like social media algorithms, gambling and micro transaction games are already doing this very effectively.
These things are designed to, at least partially, circumvent the forebrain, dramatically so in susceptible people. Each new study in this area seems to support the interpretation that it's a kind of 3rd party brain hacking. And all of these applications can arguably be called AI, or said to have AI components.
The question is, where do you draw the line(s)?
I didn't read beyond the article, so possibly there's more to it, but on the surface it seems to be inviting the courts to make some potentially huge decisions.
I loathe social media so I'm not particularly unbiased here. They can burn it all down for all I care! I'm sure many would disagree.
This was my immediate thought when I read the part about "free will"; hopefully the things you listed are included. They're highly manipulative on their own, and using AI would just supercharge the grip they have on us. Especially on "susceptible people", like you pointed out.
They just meant the legal concept of "free will".
Yes, the philosophical implications are very interesting, but society is built on the presumption of free will and so it'll be judged on that basis regardless of the weak philosophical underpinnings.
s/circumvent free will/manipulate human behavior/gHow does one prove manipulation of human behavior?
There is that old saying, “you can lead a horse to water, but you can't make him drink.”
But if you do get the horse to drink, how do you prove it was you?
So, I'm on vacation and have been drinking, but I think op is making a joke about how difficult this would actually be.
Op is using posix syntax for text replacement to suggest that solving this problem is as simple as replacing all the phrases on the left side of the middle / with the ones on the right.
That is extremely simple computing. This problem is not.
It's understament for just how difficult accomplishing the goals of what I see this law to be.