/sci/ - Science & Math

We do not know what the future brings?

>>2245383
way to pick and choose events so that they fit on your straight line. Development to our current state of living is far more of a clusterfuck than that graph would have you believe,

>>2245390

I read somewhere that graphs like these will tend towards forming straight lines, I don't know enough about math to understand why though. or if that's even true.

>>2245395

In a logarithmic plot a straight line is an exponential curve.

-.-

Hard Artificial Intelligence, posthumanity, all those are possible. Moore's Law won't change much except the "lower bound", that is, the minimum time in which these can be created due to their processing requirements.

Then it's a software issue: One idea would be to use supermegaclusterfuck computers to run evolutionary algorithms and prune the bad branches and get something out of evolution.

Another idea would be whole brain emulation, but that's not really AI: The AI researches would love to have you believe that scanning a connectocome and running a copy of it in a computer is AI, but that's because after 60 years of failure they are backing down from traditional brute-force AI and they have too much pride to admit they are emulating nature.

It can most certainly be done, and it will be done, but the results of rapid simulated evolution won't be what you expect them to (I mean, they'd be evolving from scratch: You could end up getting a particular type of consciousness that is good at, say, piloting spacecraft out of the atmosphere, then dives downwards and says "LOOK A NICKLE" and crashes). And WBE is NOT artificial intelligence GOD DAMMIT.

(cont.)

>>2245414
>typical sci poster.jpg

Now, a world where computers make us obsolete AND immortal is the Singularitarian view, and that's totally not a paradox at all. Not at all. Because super-posthuman intelligences that are, by definition, incomprehensible, will have human ethics, right guys? They will let us live in their GLORIOUS MATRIOSHKA BRAINS while they proceed to Dyson Sphere the universe, it are fact! I know it becuse a fat british neckbeard with a degree in CS told me so!!!!11!!!!1111!1!!1!!!one!11111!!!!11!!!eleven!!!11!!!!!1!!!eleventyone!!1!111!!11!

I've read an argument that stated that before humanity was capable of the necessary technology to create the super intelligences that are supposed to start the singularity, we would have already reached a point where all the lesses jobs in the economic process had become automatized (that takes less technology advancement than the singularity). So the technology advancements would probably be halted in behalf of the economy's survival, indefinitely delaying the singularity.

>>2245414
Oh look, my old virtual gaming console

Does our sun have a Dyson sphere around it? Do we see any evidence of artificial constructions anywhere?
No.


So either:

1) We are the first species in the Universe to achieve the singularity.

2) The singularity is impossible.

>>2245434
Hitting the singularity won't allow anyone to bypass the laws of physics. Lack of evidence of non-terrestrial super advanced technology just means that space travel is very, very hard and sapient races are very, very far apart.

bump

I've posted this in most singularity threads I've seen, but it never seems to catch on.
Singularity means a root in the denomenator. A pole.
i.e. -1/(x-2012) or some shit like that.
however, as was mentioned earlier, human development seems linear on a log scale, so our development should be more accurately modeled as:
a*b^x
for which there is never a pole, or root in the denominator.

tl;dr there is no singularity.

I am not trolling, can someone PLEASE tell me what "the singularity" is.

>>2246828
I am not trolling, can someone PLEASE tell me what "Wikipedia" is.

>>2246838
Wow, what an asshole. I'd rather someone give me a simple answer, than wade through all the bullshit on wikipedia.

So, suck my dick. (no homo)

>>2246853
Well it's a hard-to-define concept, so a written article would probably actually be easier to understand than us rambling at you.

But basically, the idea is that there will come a point when we create AI that itself can create AI which is better than itself. I'm sure you can see (or appreciate how impossible it is to see) what would happen after that.

Hey guys, what's going on in this thread?

>>2246864
Thanks

>>2246876
my face when I was both people

>>2246864
I sure as hell can.

Because if huanity is good at something, it is at shooting itself in the face.

>>2246892
It would be so fucking depressing if this actually happened after a thousand years of science fiction with this as its main theme.

OPs graph has a shitty y axis.

That is all.

>>2246917
There are many ways humanity can meet its end. Most people prefer not to think of it, but it is only a matter of time.

Pictured: Black death plague.

>>2246927
Yet another picture.

Note: It is not gone yet. It is currently treatable, but that doesn't mean it can't evolve and become something worse.

>>2246933
>>2246927
>implying bubonic plague is deadly
>implying it can't be cured by antibiotics
>implying any bacterial infection will be the doom of the human race and not viruses

>>2246927
The gangrenous finger tips are from a Buerger's disease: lack of arterial flow to the digits. Black death is in patches all over the body: postules. Quit screwing around.

>>2246927
>>2246933
what about leprosy?
oh wait, antibiotics win again!

thank you science

>>2246945
Reading comprehension? It IS deadly. Why do you think it was called "La Peste"? There were dead bodies everywhere.

>>2246952
Bubonic plague causes gangrene, besides the obvious bulbous lymphs. Read your sources again.

>>2246978
If antibiotics were always effective aids would be gone by now.

>>2246981
Lat. pestis, epidemy.

>Are there any good arguments against the singularity?
The law of diminishing returns
The effort that is needed to get to new knowledge is growing exponentially.
Look at the ever increasing costs of designing computer chips, which server as a counterweight to Moores Law

>>2246996
antibiotics are for bacteria, aids is due to a virus.

>>2246996
HIV is a virus. It is not affected by antibiotics. Fortunately, there are anti-viral drugs, and it seems to be possible to cure AIDS with a marrow transplant from a person who is almost entirely immune to HIV. Science!

>If antibiotics were always effective aids would be gone by now.
The stupid is strong in this one

>>2246996
and they don't work for the common cold either. we're working on that shit nigga.

>>2247004
Spanish word. "Peste" means stench.

>>2247009
Obviously. My point yet stands.

>>2247023
your point is that humanity could be wiped out in the future by something we don't know how to combat.

there is no point to this, this is just a statement.

we COULD get wiped out by a highly intelligent race traveling the galaxy on a quest of imperialism. There's no point in trying to argue it though.

>>2247023
http://www.etymonline.com/index.php?search=peste&searchmode=none
Apparently you're confusing with plague.

Also, there's 20-40% of spontaneous recoveries with bubonic plague.

>>2245420
Thank god for free-market capitalism.

In countries that embrace this system, artificially retarding technological advancement could never happen. Businesses would be free to do as they please.

>>2247266
Including artificially retarding technological advancements. The free market has a bad habit of wasting resources in dead-end enterprises.

Anyway, I hope the singularity does not come before some serious social changes. We already can't deal with living with other humans, we'll fail hard at dealing with AIs.

>>2245383

http://rationalwiki.org/wiki/Singularity

Glad to see people posted arguments against the singularity.

>>2247315
yeah, nasty free market, fortunately it's not even nearly as bad at planning ANY form of government intervention not to mention central "planning"

>>2247315

Wasting resources on dead-end enterprises from time to time is a natural part of free enterprise + creative destruction, you can't invest whit out risk, and sometimes when you risk, you lose.

Either way, the market will always be more efficient than central planning.

>>2247331
Yeah, nasty government, but actually, government funding is notably better than private funding for scientific progress.

>>2247350
yeah, probably because people naturally don't desire this research

call it "good for scientific progress", i call it "robin-hoodism" which is just another category of stealing

>>2247340
You can't invest without risk, but it won't stop "rational investors" from trying.
Anyway, just because it's not a (or the most) beneficial option, does not mean that it's not a (or the most) profitable one. Financial risk and general utility are different things.

>>2247368
If stealing is good, stealing is good. I don't care about morals.

>>2247368
And once again, between what people desire and what they pay for, there's a lot of stuff. The markets, for one.

>>2247381
>If stealing is good, stealing is good. I don't care about morals.
>good
>morals
>mfw

People have the right to do whatever they want with their stuff. If someone didn't work for it, it wouldn't even exist in that form. (Subtleties aside, please) If you want to use something created by others, you need to make them let you. If you take it forcibly, you're counterproductive AND immoral.

>>2247385
when we talk supply and demand, we always think of demand backed up by prior supply, not the wishful demand - that's how it works, everyone gets the same share he gave in, and the benefit is subjective only

Peak oil.
Nuff said.

>>2247402
this

we can find a solution to peak oil in time, or we can see research slowing down together with economy, until we can only rely on geniuses anymore, like we used to in the past centuries

there is no certainty of progress

>>2247391
I think property ownership is inherently immoral. What now?

>>2247412

>calls property ownership immoral
>forgets to make the distinction between private, personal and communal property
>forgets to make the distinction between possessions and capital

So you say no one should have anything?

>>2247412
nothing

why should i respect your point of view?

inb4 role reversal

>>2247421
herp derp, it should be clear that I was talking about private and personal property, you hairsplitting dickass

and yeah, keeping something to yourself when other people can put it to good use is immoral

I can tell you this much OP - In the next 10 or 20 years, technology will either come to a complete hault, or it will have an amazing boost it hasn't seen since the birth of the transistor.

Moores law is about to crash and processor speed will come to a hault, what will define the future of our technological advancement is if someone can think of something amazing that will change the world as we know it, much like the transistor did back in its day.

As for singularity, the question is wether or not any being can create a being smarter than itself? I'd say i'm somewhat skeptical of this myself.

>>2247391
People have the rights people give them. Taxes are not a more forcible way of depossession than regular capitalist employment. Saying X is is counterproductive, Y is not, and taxes are X, is in no way a valid argument.

And yes, there's no need for morals to see the benefits in anything.

>>2247399
Of course that's what we're talking about, and that's what artificially retards technological advancement.
For the second part of your sentence, it's so vague, yet so wrong...

On the one hand, Kurzweil's pretty graphs are completely arbitrary bullshit.

On the other hand, it will happen eventually, but I doubt it will progress nearly as quickly as he thinks and it sure as fuck won't be utopian.

Look at the rationalwiki link above. The arguments against the singularity come down to two main bullet points:
> It is impossible to create a machine smarter than a human.
That's just plain retarded. The human brain sucks at pretty much everything. We've got robots that can paint, compose music, intuit solutions to logic puzzles, we're even making real progress on shit like emotions. But all this work on AI is just the beginning. By 2050, we'll have enough raw procesing power to able to simulate billions of human brains, down the the last neuron, in real time, on a single supercomputer. Can you imagine having a billion Einsteins working together, 24/7?

The rest of is just a hundred ways of saying this:
> Kurzweil is a faggot, HURR DURR
Just fuck off with that shit, nobody cares.

>>2247460
I won't be satisfied with AI until it can demonstrate competence in language. Everything else follows from there.

>>2247466
Smarbot ?

>>2247436
Private property is a way to prevent violence. If you and I have different opinion on certain properties moral use, we will have a conflict. If we however accept we want to reduce the violence, we accept that each one of us can have property and he decides how it's used. Free market and capitalism is just a natural consequence of reduction of violence in society. And violence is opposite to freedom.

Thus the free state is the one that reduces state violence to the level required to prevent crime violence.

>>2247456
There are regulations in modern welfare states, but very idea of capitalist style employment requires voluntary agreement between employer and employee. The only restriction is neither can use force. Force used to restrict use of force has to be accepted if we want to reduce violence.

>>2247479
If there is a difference of opinion on how a resource should be used, it should be decided democratically, or by committee. Splitting up the resources is not always the best solution.

>>2245383

>Good arguments against singularity

Microsoft

that is all

http://en.wikipedia.org/wiki/Singularity_%28operating_system%29

ITT: Fucking NOBODY understands Moore's Law.

It was originally a simple observation about market trends. Then it got taken out of context. Then it got reified. Then the "definition" was changed numerous times as various physical limitations were met and architecture design evolved.

In real life, it's useless. Stop acting like it is some sort of intrinsic element to the industry or a phenomenon of scientific progress or some bullshit like that.

>>2247488
What am i reading.jpg

Democracy? Even if we assumed there is one correct way to use the resource, people's rationality vary and so does their susceptibility to influence. Democracy is not an answer to economical or moral problems.

Committee? Circular reasoning, because appointing it is a problem that requires decision itself.

Democracy is a compromise not perfection. Committee is pragmatism not idealism. Both of them should be avoided if practically possible.

>>2247505
How is democracy any more of a compromise than dividing it up equally? The problems with democracy that you just described are far more pronounced where private ownership determines how resources used. Big surprise, individuals and small groups of people tend to make more selfish and less rational decisions than large groups of people.

>>2247486
Too bad the definition of force you use is in practice irrelevant. And the definition of capitalism, too : I guess you don't boycott asian goods.
You can force a contract without using an official penal system. And for that matter, you can force a contract using one.

>Violence is opposite to freedom
Lolno.
And state violence is in the hands of capitalists. Why do you talk about your libertarian fantasies now ?

>>2247516
Do they? Cause all i see is people voting for politicians that promise privileges to lower class at the expense of the middle class and especially rich. Everywhere around the world.

>>2247521
because i am right

i'm outta here

>>2247523
China's dictatorship is pretty fair to the rich, as they tend to staff the politburo.

>>2247523
Yes, well, that is the most rational course of action. The rich don't deserve to have such extreme control of the country's resources. Most of them inherited their wealth anyway.

>>2247505
Then what would you suggest?

>>2247532
who's the moralfag now, commie?

>>2247535
uhhh that was the whole point dude, I was always talking about morality. see:

"Property ownership is inherently immoral"

Fucking FEDERAL RESERVE and DA JEWS

>>2247535
Yeah, there's two commies here. At least.

>>2247534
not gonna defend it, cause it's late at night here and i am really out of 4chan soon, but currently minarchistic republic with a strong libertarian constitution and actually working power division

inb4 all the kind of bullshit that could be easily proved wrong, from arguments of free market inefficiencies and drawbacks, through immoralities of capitalism and libertarianism, through "undesirable outcomes" like seemingly low research funding and "inequality" and to futuristic dystopian collectivist visions
etc. etc. freiland, keynesianism, socialism, welfare-poverty-crime, inflationiary policies, crimethink, state knows better, protection of minorities, bullshi<span class="math">t^100[/spoiler]

bye, fascists, commies, statists and in general pretentious people who cant earn their own money but "know" how to use that of others

>>2247549
who's the other one

>>2247553
>I don't need to actually support my position, just look at all these words and assume I know what I'm talking about.

>>2247554
I am.
The one who started with "If stealing is good, stealing is good. I don't care about morals."

>>2247559
me again

that was exactly the idea behind that post, good you noticed

>>2247553
Oh. That's a bit more reasonable than I expected. A bit hard to defend in an argument of practicalities, sure, but not nearly as asinine as some technocratic utopia or transhuman council. A shame you're leaving.

Anyway, I think this is all a case of measuring the drapes before you've bought the house. Let's just work towards the best future possible and hammer out the details along the way.

>>2247553
>implying that people on 4chan don't have money
But that's not even the funniest part of your post! The bit about "free market inefficiencies are bullshit and I can prove it" when you just used cop-outs like "people don't really want/deserve fundamental research because they don't buy/sell it"... that's precious.

>>2247566
>Let's just work towards the best future possible and hammer out the details along the way.
you marxist !

>>2247566
Well, i've spent a lot of time thinking about it, and a bit of time reading about it, and then again a lot of time discussing it. In terms of probablity, i would have heard 95% of initial counter-arguments to my ideology, and i would have to repeat what i have already wrote somewhere else in the internet, in many cases many more than once.

Even if i was not sleepy, i wouldn;t be very eager to discuss this stuff, especially with some commies present in the thread, ready to irritate me with their anachronic idiocy that dates back to Ancient Rome "equal land" idea, and is more or less evolved and more sophisticated form of the same bullshit.

>>2247584
dude we get it, you're a disaffected youth who has just discovered libertarianism and believe you know better than the rest of the world because your ideology is a unique snowflake

I've already met hundreds of people like you, hell, I've already BEEN you.

The novel "Accelerando" changed my views on the "Singularity". Made me realize it essentially entails the end of human civilization.

>>2247603
keep trolling, i might fall for it

>>2247604
well yeah, given that the singularity would advance our technological progress to an unrecognizable state overnight

we'd basically be handing over the keys to progress to machines

Moore's Law or any accelerating returns factor is not even necessary. Raw computing power is RIGHT NOW equivalent to the human brain, even by the most conservative estimates. check for yourself.

All that's missing is the right code/programming. How long will that take? 10 years? 50, 100? It actually doesn't matter much in my view because any one of those numbers should radically change your outlook/priories on things. I can't fathom intelligence being so hard it would take 100's of years (given a stable/non-catastrophe ridden world). In fact i'd imagine pure biology (like bodily functions) to be much more complex, given how much time evolution has spend developing the basics of life/biology, compared to how much time has been spent on actual generally intelligent beings near the level of humans.

>>2247606
He's quite right, though. That part about rome was ridiculous. Private property was invented by earlier states.
You sound like you only repeat some old slogans.

>>2247620
Actually those estimates don't make any sense because the brain operates on completely different principles than conventional computing. It's true that a conventional computer can manipulate data much faster than the brain does, but the brain has the advantage of massively parallel computing. It's like comparing one really fast computer to millions of really slow computers working in parallel.

>>2247642
Yeah, but european dream of equal land ownership is actually protocommunism. All the later peasant rebels, sects, jesuits in what became Paraguay, up to Soviet Union and it's satellites evolved from the idea that perfect world can be obtained if every family has exactly enough land to sustain itself. Paradoxically, this is more interesting concept than anything it inspired.

>>2247656
The estimates made for a human brain's computing power in operations per second include any gains made from using a parallel system.

The metric is operations per second. Operations per second in a brain is the same thing as operations per second in a computer. The two just get to the same end in different ways.

Regardless mimicking the brains parallel functionality is not impossible and is a trend currently already underway.

>>2247656
Exactly. Are you banking on accidentally creating a mind? We're no closer now to a computational theory of mind than we've ever been, and there's sign of ever getting there.

>>2247669
You're not getting it. The brain doesn't do "operations" AT ALL. (Unless you're talking about on the conscious level, in which case it takes 30 seconds to do a simple multiplication)

>>2247669
"Operations per second" is an extremely vague metric even when you're just considering computers. It depends greatly on how sophisticated each operation is. Where certain tasks for a computer might take a hundred thousand operations, for a human brain it might just take 10.

Computational power of the human brain:
http://www.foresight.org/nanodot/?p=3051 : from 10^11 to 10^16
http://www.merkle.com/brainLimits.html : cites between 10^12 and 10^14
http://movementarian.com/2006/08/18/flops-mips-watts-and-the-human-brain/ : 10^17 flops
http://www.transhumanist.com/volume1/moravec.htm : 10^16 operations per second

>>2245416
>implying I didn't make that picture first and he saved it

Most of the difficulties in AI from what I've heard are in pattern recognition, not processing power. That's what my astronomer friends are always bitching about: you can't make programs to analyze thousands of pictures of the sky because they can only detect what you tell them to and no one has been successful in creating programs which can recognize patterns in the manner of the human mind.

We're running out of resources on earth and are too stupid to escape or even to allocate those resources efficiently. We'll collapse into deprivation and warlordism at worst, or maybe if we're lucky we can see the rise of a progressive, pro-free speech socialist society, but that's unlikely.

This is the only acceptable outcome for terrans.

http://www.youtube.com/watch?v=rLMn3_SzBiU