š¦¾ Ever thought AI might start controlling us? =D
Be honest, even just for a secondā¦
Do you ever think that AI might eventually take over the world?
Weāre building agents, giving them more autonomy, connecting them to everything, teaching them to make decisionsā¦
And then one day they might look at us and say:
āOkay, humans. Weāve decided whatās best for you ! ā
Just a little bit of Matrix, maybe...
As someone from filmmaking, I canāt really control the scenarios that appear in my head while building an agentic environment =D
I know, I know - probably too much sci-fi.
But sometimes I look at what weāre building and think:
Are we creating tools for ourselvesā¦ or slowly creating something that will regulate us one day?
Be honest. Has this thought ever crossed your mind in the shower? =D
Replies
Have we already started building the systems that could eventually influence our everyday decisions?
@shivam_kushwaha16Ā ig these systems are already influencing our decisions. that's a fact...
Honestly, not really. The more I train my agents, the more they just do what I've told them.
What I did do is lock a few things so they can't happen without me. Nothing gets posted, sent or deleted until I say yes. So for now I'm still the one in control =D
Then again, the social feeds we spend half our day on are already run by AI algorithms. So in that sense maybe it's already got us lol
@hisashispaceĀ same. We are giving agents massively increased control and access to our data and tools, but with proper guards/gates/hooks in place, agents "can't" technically do anything you don't want them to. Though ... I'll be the first to admit that half the time I will just approve all action requests without even looking lol.
@hisashispaceĀ @bromainĀ Thatās the difference between you - the technical guys, and me - the filmmaker =D
I imagined agents breaking the rules we set up .. lol
@bromainĀ @adanaĀ Haha I wish they'd just follow the rules. They don't, so I'm pretty much fighting with them every day lol
it will never control if done efficiently
@kshitij_mishra4Ā But what if the intention is exactly that? =D
As a designer working on AI tooling (RapidNative we generate mobile apps from prompts), yes, the thought crosses my mind - but usually in a less cinematic way.
What I actually worry about is quieter: autonomy creeping in one "auto-approve" checkbox at a time, because friction is bad UX. The takeover won't look like Matrix. It'll look like really convenient defaults.
So my job now is weirdly the opposite of classic design, adding the right friction back in, so humans stay the decision-makers. Great question though, this one's shower-thought certified š
@famithaĀ Interesting point about the checkboxes..
And I checked out your product landing - seems really cool!
But how do you keep that classic design mindset in a modern vibecoding env?
my take is that agi is here.
@jake_moshelĀ I like your laconic comment style =D
How do you feel about agi being near?
I think we're just too comfortable with letting it make decisions. We need to retain the ability to critically think.
@abhinav_mahataĀ We can retain it, the problem will be with future generations ig.
@adanaĀ Every generation has faced change, starting with the industrial revolution, then software, and now AI. They'll have to adapt as we all have.
@abhinav_mahataĀ yes, agree. itās interesting to see what will be going on over the next years.
MHO, weāre still pretty far from that, so I`m not thinking about it just yet. The way large language models work today, theyāre basically very powerful know-it-all librarians. But they still lack a real grounded understanding of the world, physics, cause and effect. In that sense, they donāt really have what Iād call true āartificial intelligenceā yet.
The next few years will be interesting, especially seeing how fast theyāre evolving, but for now theyāre still software systems dependent on the tools and environments we give them. I donāt worry too much about it until I start seeing something closer to human curiosity, perception, and genuine understanding beyond theoretical answers.
That said, as a thought experiment, if we ever do create something closer to true artificial intelligence, I could imagine multiple systems forming some kind of swarm, debating between themselves, and deciding that we (humans) as a species are acting against our own peace or progress. So yes, I think itās possible in theory. I just donāt see it happening anytime soon.
@theo_gzdĀ So theyāre already debating on Moltbook =D
But I agree with you, the next few years will be really interesting to watch.
what has happened , already happened!
@dimneoĀ yes =)