[ 3 / biz / cgl / ck / diy / fa / ic / jp / lit / sci / vr / vt ] [ index / top / reports ] [ become a patron ] [ status ]

2023-11: Warosu is now out of extended maintenance.

/sci/ - Science & Math

File: 173 KB, 1600x945, FRAMED 8.png [View same] [iqdb] [saucenao] [google]

Anonymous Fri Jul 20 20:51:44 2018 No.9882329 [Reply] [Original]

How likely do you believe it is that this will prove to be correct?

>>	Anonymous Fri Jul 20 20:52:19 2018 No.9882333 >>9882329 0

>>	Anonymous Fri Jul 20 21:00:42 2018 No.9882360 You can instantly dismiss anything tech related that uses the word "singularity".

>>	Anonymous Fri Jul 20 21:28:28 2018 No.9882421 >>9882329 Techies seem to forget that natural things follow logistic curves not endless exponential growths.

>>	Anonymous Fri Jul 20 21:30:48 2018 No.9882426 >>9882329 Looks like made up shit to me.

>>	Anonymous Fri Jul 20 22:26:10 2018 No.9882509 >>9882329 ray kurzweil pls go

>>	Anonymous Fri Jul 20 22:50:48 2018 No.9882531 >>9882333 1

>>	Anonymous Sat Jul 21 02:06:35 2018 No.9882770 >>9882531 Zero to One: Notes on Startups, or How to Build the Future

>>	Anonymous Sat Jul 21 02:27:58 2018 No.9882821 >>9882770 Kek

>>	Anonymous Sat Jul 21 02:45:13 2018 No.9882850 >>9882421 >natural things

>>	Anonymous Sat Jul 21 02:46:20 2018 No.9882854 >>9882329 singularity is arbitrarily plopped in there like its somehow equal to the rest of the progression. Its made up shit.

>>	Anonymous Sat Jul 21 02:52:20 2018 No.9882867 >>9882329 >2017 Already incorrect as fuck, we are using 10-15 year old technology with minor improvements and slightly smaller semiconductor resolution... and the stagnation has been going since 2010 atleast.

>>	Anonymous Sat Jul 21 03:38:53 2018 No.9882934 Quantum computing only affects some very narrow specialized fields. It is not an do-all-end-all solution for general computing.

>>	Anonymous Sat Jul 21 03:40:38 2018 No.9882935 Terry S Davis codes in quantum computations.

>>	Anonymous Sat Jul 21 03:44:56 2018 No.9882941 >>9882934 i agree but quantum computing could lead to another technology that will make waifus real. this doesn't mean singularity will happen.

>>	Anonymous Sat Jul 21 03:59:18 2018 No.9882953 >>9882867 You're talking about architecture stagnation. They were still continuously shrinking transistors despite the fact that the architecture hasn't change.

>>	Anonymous Sat Jul 21 04:23:12 2018 No.9882980 >>9882360 /thread

>>	Anonymous Sat Jul 21 10:48:06 2018 No.9883441 >>9882329 >tfw 7nm coming next year

>>	Anonymous Sat Jul 21 13:21:38 2018 No.9883668 >>9883441 feels good man, but at 5nm we're hitting the quantum tunnelling barrier we need to go to self-organising 3D molecular chips

>>	Anonymous Sat Jul 21 16:05:56 2018 No.9884038 File: 44 KB, 500x539, serveimage.png [View same] [iqdb] [saucenao] [google] How likely do you believe it is that this will prove to be correct?

>>	Anonymous Sat Jul 21 16:14:20 2018 No.9884060 >>9884038 that hole is of today not caused my muslims, it's caused by iphones and facebook.

>>	Anonymous Sat Jul 21 17:11:55 2018 No.9884181 File: 73 KB, 620x413, singularity-682554.jpg [View same] [iqdb] [saucenao] [google] WE TRANSHUMANS NOW

>>	Anonymous Sat Jul 21 19:06:42 2018 No.9884343 no

>>	Anonymous Sat Jul 21 19:16:39 2018 No.9884363 >>9882329 show me a single chart made by a futuro"logist" that ever turned out to be correct

>>	Anonymous Sat Jul 21 19:22:22 2018 No.9884374 File: 52 KB, 2100x1499, humanaccomplishment.png [View same] [iqdb] [saucenao] [google] >>9884038 Have a better graph. http://www.unz.com/akarlin/graphing-the-dark-ages/

Anonymous Sat Jul 21 19:24:35 2018 No.9884377
File: 36 KB, 736x233, intelligencescale.jpg [View same] [iqdb] [saucenao] [google]

>>9882421
Except that an AI increasing its intelligence would make it even more efficient at increasing its intelligence, so the rate of intelligence increase will likely accelerate.

>>	Anonymous Sat Jul 21 19:25:37 2018 No.9884379 File: 99 KB, 1024x836, Intelligence2-1024x836.png [View same] [iqdb] [saucenao] [google] >>9882333 >>9882360 >>9882426 >>9882509 >>9882980 >>9884343 Not an argument

>>	Anonymous Sat Jul 21 19:48:01 2018 No.9884426 >>9884363 http://www.kurzweilai.net/images/How-My-Predictions-Are-Faring.pdf

>>	Anonymous Sat Jul 21 19:49:33 2018 No.9884428 >>9884379 Where are we on that chart now? Has AI attained ant level intelligence?

>>	Anonymous Sat Jul 21 19:57:58 2018 No.9884438 >>9884426 >make a 100 predictions >10 turn out to be varying degrees of correct >advertise the 10, stay silent about the 90 >???? >PROFIT!!!

>>	Anonymous Sat Jul 21 20:01:55 2018 No.9884444 File: 207 KB, 1068x1600, Ray-Kurzweil.jpg [View same] [iqdb] [saucenao] [google] >>9884438 Give me 10 of the 90 you're talking about.

>>	Anonymous Sat Jul 21 20:04:31 2018 No.9884449 >>9884444 checked, I sure as fuck won't research his history, but I'm pretty sure that's the mechanism, that or luck

>>	Anonymous Sat Jul 21 22:40:16 2018 No.9884708 >>9884444 singularity confirmed

>>	Anonymous Sat Jul 21 23:18:24 2018 No.9884781 >>9884449 >singularity denialists refuse to do basic research, resort to guesswork checks out

>>	Anonymous Sun Jul 22 03:02:01 2018 No.9885200 >>9884374 >significant figures born per year That's not vague at all and definitely not an asspull.

>>	Anonymous Sun Jul 22 05:29:04 2018 No.9885387 >>9882329 There will be a distinct disconnect in the price/computing of quantum. And that won't even include how quantum won't 1-1 replace classical computing. It'll be like ARM vs x86 or a CPU vs GPU. They will specialize in different things

>>	Anonymous Sun Jul 22 05:35:55 2018 No.9885396 >>9884426 he certainly is pretty eager to claim things correct like saying VR existed in 2009 when it obviously didn't by today's standards

Anonymous Sun Jul 22 05:36:03 2018 No.9885397

it is unlikely that quantum computers would be used domestically but 2026 and allow AI to become available to be used by it as quantum computing logic gates are like a formula/algorithm to be reused again and again so making a series of logic gates to be used in a way that supports AI and can be used for generic use such as a conventional computers will be very time consuming

>>	Anonymous Sun Jul 22 13:52:30 2018 No.9886127 >>9885396 He completed that document in 2010, so he couldn't really judge 2009 by 2018 standards

Anonymous Sun Jul 22 13:54:10 2018 No.9886130

>>9885397
You don't need a quantum computer in your own house or office in order to use quantum computing; you'll use it via the cloud like we use Amazon Web Services and the like. You will access it when needed for calculations more efficiently completed with quantum computing, and leave the rest to your classical computer.

>>	Anonymous Sun Jul 22 15:34:12 2018 No.9886299 Quantum computing reminds me of cold fusion technology. Always been promised. Always little experiments that show its feasibility here and there. And always 20 years away.

>>	Anonymous Sun Jul 22 16:33:40 2018 No.9886446 >>9886299 Google Bristlecone

>>	Anonymous Sun Jul 22 21:37:58 2018 No.9887003 >>9884181 Obligatory $1000 EOY

Anonymous Sun Jul 22 23:06:24 2018 No.9887109

>>9884377
the problem with that is that it assumes that superhuman AI "A" would be able to make an AI "B" smarter than it easier than we were able to make "A."
Get from A to to B may be far harder than getting to A, because all AI "A" would at best know everything that we already know and bet marginally smarter than us

>>	Anonymous Mon Jul 23 02:43:56 2018 No.9887445 The singularity is near

Anonymous Mon Jul 23 03:22:44 2018 No.9887486

>>9882329
Quantum computers are in many ways worse than digital.

You've probably seen all the proofs by excited cryptologists who show you how_certain_ algorithms perform better and indeed in an entirely different complexity class using a quantum processes.

What you probably haven't seen is all the proofs where the quantum computing is worse for the most common processes. Not just a little worse, again an entirely different complexity class.

So in the end you'll end up with some kind of hybrid quantum/digital processor.

We've pretty much reached the limit. The future now is in clustering and efficient ECC.

Anonymous Mon Jul 23 03:25:19 2018 No.9887491

>>9884377
>so the rate of intelligence increase will likely accelerate.
By what logic do you arrive at this conclusion.

Compare it to another intelligence that is capable of increasing its own intelligence; a biological brain.

Yet there are obviously limitations.

An AI will almost certainly eventually be limited by the expansion rate of its hardware so >>9882421 is correct a logistic curve is far more likely than an exponential curve.

>>	Anonymous Mon Jul 23 03:51:12 2018 No.9887506 >>9882935 Temple OS is pretty great

>>	Anonymous Mon Jul 23 13:09:31 2018 No.9887989 >>9887445 pls let it happen

>>	Anonymous Mon Jul 23 13:32:14 2018 No.9888012 >>9882329 >singularity 0

Anonymous Mon Jul 23 17:54:34 2018 No.9888384

>>9884377
The most sophisticated A.I. to date still can’t even fathom a fruit fly’s level of ability to make its own directives. There’s more to the problem of creating self-improving intelligence than just increasing processing power.

There have been no breakthroughs to suggest that we can make A.I any different than a glorified dog. Dogs and A.I. can appear to overachieve their intelligence because they’ve been specialized to mimic behavior that human’s simply consider intelligent, but they are helpless without the guidance of a human telling them what is and isn’t “intelligent”.

Anonymous Mon Jul 23 18:09:34 2018 No.9888405

>>9885200
It's actually not an awful method of quantifying human achievement - take history books and encylopedias of different nations and count how many mentions each person gets and how long the entry is. People who are mentioned in multiple nations encyclopedias and histories are 'significant' individuals, as opposed to local jamokes.

>>	Anonymous Mon Jul 23 22:01:09 2018 No.9888802 >OP thinks quantum computers will revolutionize AI

Anonymous Mon Jul 23 22:47:53 2018 No.9888870

>>9882329
>quantum
More like optical computing, which is in its infancy currently. We're talking THz scales with an order of magnitude less energy- I can see multi-exaFLOPS (even zettaFLOPS, if you're overly optimistic) human-level AI getting achieved in ~30-40 years of development and miniaturization of this tech.
Singularity in the next 100 years is a pipe dream since even a hypothetical true AGI would still need to plan, construct, test and implement new technologies and research, which takes time.

>>	Anonymous Mon Jul 23 23:05:29 2018 No.9888894 >>9884426 It's almost as if when one makes general predictions and then is allowed to self evaluate their own work it's easy to come out with lots of correct answers. I would have an infinite IQ and a 6.5 GPA If I could grade my own tests.

>>	Anonymous Mon Jul 23 23:17:19 2018 No.9888907 >>9884374 >>9884038 Dark ages aren't a real thing.

>>	Anonymous Tue Jul 24 00:00:14 2018 No.9888991 File: 783 KB, 3630x1615, 1474233509560.png [View same] [iqdb] [saucenao] [google] >>9884374 >>9884038 >pseuds not posting the complete graph never change nu-/sci/

>>	Anonymous Tue Jul 24 00:07:16 2018 No.9889003 >>9884377 Not if the next step of increasing intelligence is prohibitively more difficult or resource intense than the last.

>>	Anonymous Tue Jul 24 00:09:01 2018 No.9889010 >>9888991 What about the Old Ones

Advanced search
Text to find
Subject [?]Search by post subject. Leave empty for any.
Username [?]Search for user name. Leave empty for any user name.
Tripcode [?]Search for tripcode. Leave empty for any.
Email [?]Search by email. Leave empty for any.
Filename [?]Search by image filename. Leave empty for any.
From Date [?]Enter what date to start searching from. Format is YYYY-MM-DD
To Date [?]Enter what date to start searching until. Format is YYYY-MM-DD
Image hash
Search in	All Posts OPs Only
Deleted posts	Show all posts Show only deleted posts Only show non-deleted posts
Internal posts	Show all posts Show only internal posts Show only archived posts
Order	New posts first Old posts first
Capcode	All Posts Only by Users Only by Mods Only by Admins Only by Developers
Results	Posts Threads
Action	[ Simple ]