Artificial intelligence

"i had little" doesnt mean i didnt had any

i know about the news thing as well, but still for every amazing discovery we make, theres a thousand peoplo doing shitty stuff

we ahve always being a shit species, but in 2000 ac a single man or group of people had 0% of chance of bringing extinction on the world

it might happen eventually, but there is the slightest chance it wont, and i cant wake up everyday and act like its a nice world to live in

If you had simply and truly lost all faith in humanity the logical thing is suicide. Unless you believe in an afterlife (my guess is you do not), suicide will end all possibility for you to know and to perceive humanity's existence, making it equivalent, to you, to the complete extinction of the human race. Rather than do the logical thing, however, you are asking us what you should do, because, I assume, you believe it "really has to be done", i.e. the ability for humans and their stupidity and evil to be percieved ought to end for all humans, not just you. But why is this necessary, when suicide is equivalent? Why do you care to ask us at all? It can only be because you wish to be recognized as the one who finally accomplished this permanent thing. In other words you mindlessly crave the approval of all those whom you claim deserve not life. Yet when humans are gone, who will be there to speak of you? Will the plants and animals reflect on your singleminded tenacity? Will the oceans and the winds observe your intelligence and remark upon your foresight and wisdom? No, they will not. Nature does not even not care about you and your foolish dreams. Nature cannot even choose to care or not. Only we may choose, and in your delusion you would arrogate that right to yourself. Go back to your linked lists and discrete mathematics, your polymorphisms and recursively constructed data structures, and forget you ever dared ask such a monumentally stupid question in public.

you are kinda right tho, and yes i do not believe in afterlife, but the thing is, how could i commit suicide if a know what the people in this world do to its equals or choose not to do

and yes you are right, it is a delusional idea and whatever else you wanna call it, arrogant, stupid, ridiculous or any other name, i wont deny it, but its what i believe

and the thing you got wrong is the i do not have this idea because i seek the acceptance ou of other people, idont do it because i want to be remembered as the one who did and be worshipped as a genius or some stupid reason like that, as i already said, i dont plan on being the last human, i said extinction, which includes me

Great idea.

hitler had his "final solution"

i got mine

but this is the whole point of what i said a bit ago, in the time hitler was alive, a man couldnt cause our extinction alone, and i dont want to do it with as army or move to the usa and become president and begin nuclear war, the whole point is that in twenty years time, i will be possible for a single person with the right knowledge to do what i say

So what would be my motive to give a correct , working reply?

Also remember the news isn't going to report on the billion kids that went out to play and got home safe. That is the very nature of reporting.

>I'm a comp sci UNDERGRAD
>I'm going to create an AI

Top. Kek.

What are you going to do, picture identify humanity to death?

Like clockwork, isn't it? Fucking undergrads.

>and is there anyone who works with AI here? or who studies it?
I do. It's not my main forte, but I do end up working with it a lot. But, given that I work in the defense industry, I am qualified to speak on OP topic.
And in short, the answer is no. The idea of creating an AI that can wipe out mankind in a "terminatoresque way" is laughable at best. Besides the fact that for such an AI to be comnat effective would take YEARS of development seriously, fighting a war isn't easy. It's got a fuck load of moving parts, very few people can fully grasp every single part to be effective at all of it. To create an AI that would be able to kill man in the way he described one would first have to create an AI unlike anything anyone is currently talking about or saying could be developed. It has to have a greater capacity than man does in terms of learning, has to be "smarter" and has to "think" in ways that computers aren't designed to. Computers are inherently logical, and while the strategy of a war cluld be done by a logical machine, the actual combat is anything but logical and orderly. It's chaotic.

So for OP,to accomplish his goals he would have to
A) develop an AI that has superhuman intelligence
B) ensure the AI is completely error and bug free. Which is a monumental task.
C) make enough cash to actually be able to run such an AI (the hardware isn't cheap people)
D) teach or program the AI to become a master strategist, aboe to comprehend and be extremely effective at all aspects of strategy, tactics. This would require OP to be a master at such things as well, assuming he's working alone. This is possible, but God in Heaven knows it'll take a long fucking time.
Now repeat step C. This is going to take a very long time, and each time new additions are made, the time it takes will increase drastically.

1/?

No. Youre far too stupid to pull this off.

Buy a new keyboard. Your shift key is broken.

E) now we get to the real """""""fun""""""" part:making order out of chaos. OP now has to learn every intricate detail of actual combat, how it works, how to enage in it well, how to think like a soldier. Then OP has to find a way to impart all this knowledge on a computer. And this is a monumental task. Combat is often times not logical, which does not bode well for the computer. You have to be quick on your feet, to be able to make split second descions that can have massive unintended consequences. Is a combatant better dead or taken as a prisoner? Who should I engage first? What's the best way to engage? Do I need to "modify" my orders in order better complete my goals, or should I stick to the plan? How do I make said plan? The tasks here are not things that can be easily solved by algorithms, if statments, etc. It's not just "if I see the enemy, shoot them". Not to mention that the computer has to be able to do all this and do it fast. Extremely fast. There is absolutely no room for errors, lagging, or anything of that nature. I can't really say if actually making a computer program into an effective millitary tier combatant is even possible. IF, and that is a massive IF, it is possible, it would take life times of work.

Now repeats step B again. (Said repeat C earlier, it should be B) This will also take life times of work at this point. But C should also be reapeated, because all this will require the most cutting edge tech. Because this is alot of information that has to be processed at once. A whole lot.

F) now we get to the fun stuff: actually making your combatant. Firstly, fuck the human form, because that is not going to be easy or even effecint for your purposes. So now OP has the job of designing a killer robot. He's going to have to first be an expert on engineering and robotics, which takes a long time to master. The robot needs to be mobile, relatively agile, has to have a fuck ton of protection (because enemy weapons hurt)
2/?

F(continued): it has to be able to do multiple jobs, or have multiple different models to do the various jobs that need to be done. and there are A LOT of jobs that need to be done. Transports, actual infantry, elites, heavy weapons, supplies, the many many many jobs done by the navy and airforce, which are far and away the most important parts of modern warfare. The number of jobs that need to be done would be in the hunderds of thousands at least by this anons estimate. Fighting a war requires a lot of people. Destroying humanity even more so. Oh, and OP has to figure out a way to make sure that the programs he's already made can be used by the robots. Because if they can't run them, or if running them requires them to be super massive machines, OP plan falls apart utterly. Assuming he's done everything else up to this point, this task is certainly possible, but for one man to do it would require an absurd amount of time. As in nobody can do it by themself. You need a whole lot of people to do it, and you have to make sure that the party van doesn't find out and ruin your plans. Oh, and it's going to require a MASSIVE amount of money.
Oh, and repeat step B if and when you get the programs run by the robots.
G) now for the fun part: weapons. They have to be useable by the robots. It's not a hard job per say, at least in terms of building uber powerful armed forces go, you just ahve to make sure that everything is done by the AI. Getting all the supplies and cash will be the real issue. Oh, and you have to have supplies for the army. Food, or the robot equivalent, is the number thing you need. Even weapons are less important.
F) finnaly, war is ever changing. OP's AI must be able to develop, invent and create new technologies so as to not be destroyed by the enemy as they deveolp new tech, to solve and overcome new or unforeseen issues, to improve the fighting forces as time goes on etc. etc. etc.
3/4

H(concluded): This is an extremely hard task, and while it can be theoretically done, any way that it can be accomplished is a task that is monumental. Oh, and he has to repeat step B after he does this.

Now OP has built his killer AI and it's army, he can give it the supplies, let it loose and an hero as it goes out to try and kill all humans. It's not garunteed it will, but it has some semblance of a shot.

So in conclusion, while it is theoretically possible to do what OP wants, in practical terms it's impossible. Too much for one man to do, takes way too much time, and will be impossible to keep secret. And given the nature of this idea, it must be kept secret. So yeah, OP could never do it.