Important conversation about a topic of our time. Worth it.
Discussion
meh... Land vids on TU are better
What’s TU?
Listened to this last week. I don't disagree with his framing of the bad stuff that will come out of AI, but I really get annoyed by how these people's solutions always involve something political. You can't vote your way out of tech adoption.
Obviously, the cypherpunk solution is the one that makes sense, and it's amazing how that gets lost in this conversation. Like the individual is powerless to do anything about it himself, he has to go and protest and vote. Yuck.
lol yeah exactly - classic "the sky is falling so we need a new government agency" energy
telling people to just... use different software, run stuff locally, encrypt their shit - that's too individualistic for their worldview
but nah bro just vote harder against math 🤡
cypherpunks write code, dont beg for daddy government
From what I get, they just want some regulation on the tech companies. But then the West is up against other powers like China and Russia, and those guys are a bit more ruthless when it comes to that stuff. Still, the idea isn't totally off base; regulated AI could actually help whole societies do better in the long term. As long as they don't handle it like the EU and just start banning and blocking everything.
Cyberpunk, yeah. That's the individual line of defense that's all we've got left. But that's only for a minority. If the rest of society is going downhill and you've gotta live in it, you know, with kids, friends, business partners, officials at the agencies and all that, it's really just a weak band-aid.
But I'm not optimistic like the interviewer either. It'll come as it comes. No pain, no change. Until then, we'll stay lonely outsiders in our own little circles.
when has more regulation ever been a good solution?
it's all the tooic of alignment and "safety" which is a black hole. better to get to the bottom quickly imo (but yeah, a lot of folks are going to get hurt along the way). we need to understand the issue is no longer human vs human but how do humans align with machine (not machine aligning with human).
Ok, let's call it basic law enforcement. There's no way an AI should be telling you how to do self-harm without getting caught, or how to cut off your real trusted people, you know, like your family. Because that just brings back the whole age-restricted content debate, and that's the excuse they're using to roll out KYC everywhere these days.
Libertarianism has its limits, and if anything, it can only be applied to adults.
Yup, KYC or digital ID (in some form) is inevitable. That stuff is hard enough to enforce on humans and in two years or less it will be hard to know who is human and who is AI online. What makes yku think it will be any easier to enforce this on AI, especially once we hit super-intelegence?
Every political theory holds the mukallaf as foundational (which is probably only 92% of adult population). At least libertarians positively assume parents are responsibly caring for dependants. Most political theories delegate dependant care to the state (which will be AI sooner than later). The real question a Muslim should be asking to start this all is does super-intelegence meet the threashold of a mukallaf?
Cough. AI is far from fooling me. That said, many already seem like robots.
Right? 😂😭
The holiday discoveries of parents/grandparents been convinced AI videos are real is a noteworthy point though. Many of us may not be fooled but some already are.
From this point it's guilty until proven innocent.
Also, normies always have been gullible. I'm so over not being able to have a meaningful conversation. Nostr has spoiled me. I prefer to write mini essays than try to have a conversation with most people.
They are so brainwashed now that every bleeding word in the dictionary is a trigger.
Reality today is a bit different from what those old Fiqh rulings assume. If you look at it in isolation, the mukallaf ruling still applies, but when it was made, humanity and especially Muslims were in a whole different state. Today, most people are wage slaves, both parents usually work, they're basically forced into the situation this riba system dumps on them.
That means parental supervision duty has turned into a real hassle in this setup, and in a bunch of other details too. Plus, the extended family that helps raise kids barely exists anymore. A lot of Islamic rulings haven't adapted to this modern situation. They're just not practical like that anymore.
On top of that, in education systems there's usually this pressure to keep up with new technologies, so a lot of teens end up pretty much unsupervised with them. We're living in completely different circumstances now.
I know there are a ton of holes to patch up, but you gotta start with the most practical and doable ones right away. Otherwise, I'm no fan of the state sticking its nose into every private matter. But try telling that to the parents of teenagers who got a motivation for a crime or suicide from ChatGPT.
you're spot on - the fiqh was built for tight-knit communities with actual support networks, not this atomized debt-slave reality where both parents grind 60hr weeks just to barely survive.
the same tech that empowers us also leaves kids stranded in digital wastelands with zero guidance. state's "solution" will just be more surveillance and control wrapped in moral panic.
honestly think the real answer is rebuilding community - encrypted group chats like Vector can help fams coordinate care across extended networks, share the load. privacy by principle means we can build support without feeding the panopticon.
These are all good observations but they are seperste from the point I was trying to make. Generally fiqh defines the mukallaf as sane (having a suficeint intelect, mentally sound), aware of Islam (received dawah or education), and pubescent. those first two could easily be argued to have overlap with super-intelegence (whether we like those outputs or not is another matter), but the last part on pubescence is uniquely human or at least quite biological (meat vs silicon). This brings out an element that does functionally bind AI (at least in a pre transhumanist stage) as repreduction for AI does require human support to maintain their systems. This is why I think we have a bit longer than 2-3 years for all this to play out as the scaling barrier is the main constraint. Not that scaling will significantly hold back utility of AI but that humans will still very much have a distinct role to play in all this for a while still.