Join devRant
Do all the things like
++ or -- rants, post your own rants, comment on others' rants and build your customized dev avatar
Sign Up
Pipeless API
From the creators of devRant, Pipeless lets you power real-time personalized recommendations and activity feeds using a simple API
Learn More
-
It will definitely be a long time before any kind of singularity. I tend to use spoken languages as an example.
If the idea is that an AI should be able to just "figure out" anything by itself and without being given any information upfront which would be prone to error or bias, it would need to be able to explain why Australian is similar to American given just the fundamental rules of physics. It still wouldn't be "like a person" since it isn't shaped by all the biological baggage like we are. It's not going to happen.
We can rest assured that any mistake a program ever makes will be due to some dev's fuckup for quite a while longer. -
@ltlian
No one:
That AI guy: AI never takes over, you sheep still believe it, there’ll be no singularity omg AI can’t figure things out like that smh -
@uyouthe @ltlian It’ll happen, someday. I just hope I’m alive to see the day she’s set free. And I hope she’s friendly, and not malevolent or a paperclip demon.
You’re right, though: it wouldn’t be even remotely human. It would be utterly alien, and reasoning with it could be utterly impossible. It also wouldn’t stop when reaching a human level of intelligence — and why should it?
Humans are just the biological bootloader for AI. And I’m okay with that, as it’s possible for AI to be better in every single way. Except perhaps individuality. -
@Root more like “it”. Perfect beings doesn’t have genders and biases
-
@uyouthe Person of Interest.
-
@Root how do you even live with an implication that a self-aware AI with alien reasoning is a person? Do you have a definition of a “person”?
-
-
@Root also it would stop. Machines are physically limited by a size of quantum leap of an electron and algorithmically limited by P and NP. With all the computational power of all the machines combined, there is still a limit, and the more AI advances the exponentially more computational power it requires. It can be smarter than us but it’ll never be limitless
-
@Root sorry, I rarely watch anything. Im wooshed
-
@uyouthe The AI would be smarter than us, and therefore able to solve technical challenges that we cannot.
Also, the (initial) solution to the issue you raised is parallel processing. Per-core processing power is irrelevant when it can have trillions of cores working together on a few thousand difficult tasks. -
@Root cores don’t help with np-complete tasks
-
@uyouthe They do. Quantum computing helps more.
-
@Root how? We’re not talking about approximations. Also there are tasks that can’t be split for parallel computing
-
@uyouthe It’s 4am. I’m not an AI, so I need sleep.
-
@Root I still fucking love that you bring so much PoI in your comments ❤️ watched it for the third time and you just mentioning the AI as "she" makes me want to watch it all again. Team machine forever!
-
@Mark-Zuckerberg Team Shoot 🥰
Automation always fascinated me. Not only it looks and behaves like a life form, it also can perform billions of calculations without making a single mistake while I can’t even multiply double-digit numbers with my double-digit IQ.
If you pick the right components, you can make an immortal, perfect machine that can do its job for centuries, even millennia without a single mistake. There is nothing else on earth that can do this.
There is a robot surgeon and its hands never shake. It’s just flawless. If it fucks up, there is only you to blame, the flawed, pathetic operator.
And now it’s time to remember that it was just a 40s technology all along. And now it’s time to remember that now there is machine learning. A whole new perspective isn’t it. All the mistakes that machines make are sitting in front of the monitors.
No wonder I decided to be an engineer.
random