/sci/ - Science & Math

>turn around

what did she mean by this?

>What are we doing with all that horsepower? Do we "solve" all of physics? Cure all diseases?

there's no clear way in which increased computing power could lead to either of these things

>>7485554
>So what's the average day like in the 2040s?
Wake up. Go outside. Realize the looting and rioting still hasn't stopped. Geiger counter sill ticking away. Slink back in to bunker. Spend the rest of the day thinking about how you'll avoid the riots and radiation while looking for food tomorrow.

>>7485554
I'll tell you whats gunna happen OP, were gunna have materials so smart they'll apply a little pressure to your skin and just enough of an electrical shock to make it feel like something actually touched to hit you

We'll be in tanks with tubes down our throats having experiences with such, this will only get cheaper as the technology gets paid off

and 'muh computing power' is just such an obvious newfag post that I'm literally straining my dick muscles urges to not take you from behind and give you a proper punishment

If you include the post that broke it, you made it to 4 replies, OP.

A round of applause for this board, please.

The worst part is that I only browse this board these days, and I enjoy that.

>>7485573
Your post doesn't make any sense
Do you not believe that "computing power" is real? I'm using a computer right now, to post this, in fact. It has 8 cores and runs somewhere around 3 GHz. What bothers you about that?

>>7485580
Not even. The very first reply was a pure shitpost.

>>7485567
There's no reason to think that they won't. We can already do basic protein folding, etc, and that kind of computing could absolutely pay off in biotech.

>>7485593
how would folding proteins more quickly lead to the curing of all diseases

>>7485586
WE HAVE HAD A COMPUTING POWER THREAD LIKE LITERALLY A THOUSAND TIMES THIS YEAR

LITERALLY

>>7485606
dude
Protein folding is an example. That's what the "etc" is there for. It is short for "etcetera." "Etcetera" is kind of like "yadda yadda" or "and so on." It means "Hey I'm not going to write out every possible example of this thing that I'm talking about, so use your imagination here."

We can use computers to simulate and research biology.
What has happened to /sci/?
I'm sad.

>>7485621
Aww, I'm sorry that you feel obligated to participate in a thread that reminds you of another thread you saw recently. But thanks for stopping by, anyway.

>>7485626
it can also mean "i'm talking out of my ass and don't really know any other examples"

>>7485554
>By 2030, our home PCs would be doing a trillion instructions per second. By 2040, maybe a quadrillion or so. What are we doing with all that horsepower? Do we "solve" all of physics? Cure all diseases? Unless you're 70 years old right now, this is probably within your lifetime. So what's the average day like in the 2040s?
Quantum mechanical limitations on circuit design will prevent this.

>>7486504
Note that the size limits only apply if your chip has to be two-dimensional. If your chip is three-dimensional, you can make *area* density arbitrarily small by increasing the height and ultimately building your components on a per-volume basis. (This also helps with the delay issue, because you don't have to make your wires as long if they can go over instead of around components and your components can be packed more densely in terms of three instead of two dimensions)

This does run into heat dissipation problems, but those are much harder to put solid physical constraints on compared to "look, you obviously can't scale your transistors past atom-sized, and even if you could you couldn't scale the wires that small, so there's a hard density limit"

>>7486510
Ok, so show me one that does that.

Probably simulating complex neural networks. This seems like the way AI is going.

>>7486504
NOTHING CAN STOP THE SPANI...I MEAN SINGULARITY INQUISITION!

>>7485561
"She" is good at tucking.

Can we treat computational power the same way we do population dynamics? If we do that, we will see that simplifying our model gives us an exponential growth, but as our model gets more accurate, the growth becomes less accentuated and eventually halts, for a period of time, before progressing again.
(1) dP/dt = kP(t)
(2) dP/dt = rN(1-N/k)

Can we treat computational power the same way we do population dynamics? If we do that, we will see that simplifying our model gives us an exponential growth, but as our model gets more accurate, the growth becomes less accentuated and eventually halts, for a period of time, before progressing again.
(1) dP/dt = kP(t)
(2) dP/dt = nP(1-P/k)

>>7486518
>one fucking team creates a code that meshes animal picture fractals with ordinary pictures
>FUCKING NEURAL NETWORK MAN

I hate you

>>7485554
> Do we "solve" all of physics

Lemme stop you right there. No. That's not how computing power fucking words, for the nth time fuck these goddamn threads. Is Ray Kurzweil sitting in a fucking Chinese AI factory overseeing a slow and steady raid on /sci/? Fuck him, and fuck these ideas.

Computing power doesn't magically translate to processes. You can have a program that doesn't do jack shit and still eats up all of your computing power like it was fucking running water.

>>7485554
what film is OP's gif?

also I can see more technological advances bringing forth the same but not computer power = curing ailments/disease

>>7487196
Do you see that little blue arrow next to your post number?
Ever wondered what it does?

>>7486504
>ignoring the possibility of commercially available photonic computing

>>7485561
>Every now and then
>I get a little bit lonely
>And you're never coming round

>>7487192
Exactly, we'd probably end up using the majority of it on the transhumanist equivalent of shitposting on 4chan.
Like the deathsports in Prime Intellect lol.

>>7486515
what is that a pic of?

I'm going to ignore all of the shitposting going on here and reply directly to OP so if my post goes against something you, the reader, has already posted, it's because you presented it without much backup, so throw a tantrum. I don't give a shit.

Assuming computing power continues its growth like OP suggests, and computing power can be used for beneficial things rather than browsing the hip social network of the time, I see NN and AI being the real beneficial part of computing power. It can be assumed that there will be a developer somewhere, sometime between now and OP's 2040 that will write a network abstract enough to learn like a human but with this exponentiated computing power. And assuming this is done right, they'll be thinking up the concepts needed to "solve" physics or whatever it is you want them to think about

>>7487202
I know what it does, it summons snacks back from the Marianas trench to permaban you for stupidity

>>7487340
Alleged quantum computer prototype from D-Wave Systems

https://www.quantamagazine.org/20130821-the-proof-in-the-quantum-pudding/

Going to be interesting to find out if it's for real or not.

>>7487356
Google is supposedly using it in machine learning algorithms.

>>7487366
Eventually, Siri will develop just enough intelligence to tell OP to stop asking stupid fucking questions.

>>7485826
Yes you're right computers are not used to study biology
What's the abbreviation for "you're too stupid to live and you may forget to breathe soon" because I need to use it in a sentence right now

>>7486504
Yeah, the "paradigm" just shifts whenever progress hits an apparent wall. You can't run a vacuum tube at 4 GHz, but somehow we still managed to get there.

>>7485561
Isn't this a /tv/ meme? Get the hell out of here you lowlife!

>>7487192
You have a semi-valid point, but you're ignoring the obvious fact there is a lot of research we can do with more computing power. There are scientists (and yes, physicists) who want to study and simulate all kinds of things, but can't because they need a million or a billion or a trillion times more computing power.

So more computing power does translate to scientific advance. It isn't a 1-to-1 causal relationship... scientific discoveries don't just magically pop up out of the circuit board because it's running faster. But you're oversimplifying.

Probally, the same we do these days: dank memes, shitposting, playing a random shooter at 300 fps and masturbating.

>>7487539
Progress is like fluid filling a maze, you can measure the distance from the entry point to the current position, or measure the distance along the path.

Certainly the maze is the final physical limit of the maze, and the finite possibilities represent all the cooridors.

Yet, you can't use the 'as the crow flies' distance to something and say that since we reach this point, we will get to the next point because it's the same distance away.

This is the nature of conservative vs progressive arguments, and we do have to accept that the goal in the maze may not be the same location.

Attempting to predict the future is a rather pointless task.

My guess however, is that massive parallelism will become the driving force behind processing power. To get more human-like behaviour out of a computer, the architecture's going to have to become more human like.

>>7487537
i don't know why i bother with this board anymore. it's a bunch of faggot teens and neckbeards from /tg/ having a big fucking jerk off fest over how amazingly everything is going to work out for everyone becuase TECHOLOGOLGY

there's such a severe lack of critical thinking about any of this stuff and the minute you bring it up there is immediate backlash since people like you literally have nothing to look forward to other than the possibility that the future is going to be this magical place

>>7487602
its almost like everyone have an iphone in their asses

>>7487617
or a cent-iPad

>>7487585
Human like, whatever that means.

But yeah, the computers right now are becoming parallel, but they've barely scratched the surface with parallel computing.

Most programs still only utilize a single core, because there's no straight forward architecture for most programmers to easily manage a parallel processing platform.

Of course, some of the problems we deal with simply don't into parallel calculations, so a subset of problems will not be helped by having parallel processes.

UI and stuff like that can be sped up, but algorithms that are merely chains of calculations won't see dick.

So right now, any gain is being fed into multiple cores, and we've yet to deliver a full parallel system.

>>7485586
gtfo AMD fag

>>7487602
>claims board can't into critical thinking
>Can't google the words "integrative biology"

Do you think the entire science of biology stopped at the end of your highschool taxonomy class? Use your imagination to think of some things the rest of every kingdom of life might be able to do better than humans, and how we could use it

>>7488195
what does that have to do with computational power?

>>7487602
I hate /tg/ too and all the other cross boarders. They just come in here because of pop-sci and "geek culture". It wouldn't be so bad if they weren't such fucking retards.

>>7487602
I don't, either... or why you would have ever come here in the first place
maybe what you should do is start calling people faggots and neckbeards and then misspell "technology" because hey that will really make *them* look dumb and then maybe no one will notice how angry you are over absolutely nothing

Never mind the fact that the whole point of this thread on this board that you hate so much literally said:
> instead of a mysterious threshold that we cross over and suddenly everyone has magical super powers

Your contribution was to jump into that and ask a stupid question and then to become super defensive when you got exactly the answer you deserved -- and yet you think that /tg/ is what's wrong with /sci/. Well done.

>>7485554
There will be trillions of those fucking disney 3D-Movies to watch ... trillions!