BS: The Singularity is Near / Captcha

Subject: BS: The Singularity is Near / Captcha
From: Songwronger
Date: 17 Nov 13 - 10:59 PM

http://news.sciencemag.org/technology/2013/10/captcha-busted-ai-company-claims-have-broken-internets-favorite-protection-system

Link to a story about a computer deciphering Captchas. Those are the twisted lines of letters and numbers you have to type in at certain websites to prove you're human. Computers can't scan them properly.

Years ago I read a book called The Singularity is Near by Ray Kurzweil, and it talks about pattern recognition. That's an area where the human brain excels. Computers are faster and so on, but they can't analyze patterns the way our brains can. But science is working on that. When super fast computers with their massive storage capacity achieve human-like pattern recognition ability, we'll be at the point of creating artificial intelligence (strong AI, it's called).

And then after strong AI becomes established, we will approach what Kurzweil calls the Singularity, a giant leap forward in intelligence. Like from ape to homo sapiens. Only, the AI leap will be millions of times greater. No way to gauge it. And no way to stop it. Because computer speeds and storage sizes double each year, and component sizes and costs are cut in half each year, research and development labs cannot help but create strong AI at some point. In the name of business competition if nothing else, humans will create intelligent machines.

The new AI machines will be self replicating. They will build better versions of themselves, then so will the next generation, ad infinitum. They will also retain the fundamental feature of human intelligence, which is the need to overcome limitations. Intelligent machines will naturally expand outward, and in the process conquer the biggest limitation to physical expansion--the speed of light. Once we achieve faster than light travel, our intelligence (trillions of times greater than it is now) will spread outward until it permeates the universe.

Or so Kurzweil said. All I know is I still have trouble deciphering Captchas.

Subject: RE: BS: The Singularity is Near / Captcha
From: Rapparee
Date: 17 Nov 13 - 11:36 PM

Something like Teilhard de Chardin's Omega Point or the McLuhan's "Global Village"?

Subject: RE: BS: The Singularity is Near / Captcha
From: JohnInKansas
Date: 18 Nov 13 - 12:24 AM

Several months ago there was quite a bit of comment on the apparent ability of malware operators to decipher captchas.

So far as was indicated, no one in the anti malware business ever found a program that did a good job of it. (It is possible that computer methods were found, but were closely guarded to prevent them being known to more criminals, but ... .)

One method that was used was to grab a captcha and immediately "spam" it disguised as your own "proof of identity" for some real scam, or just some useless email that promised something marvelous. If you spammed a thousand people, and only ONE responded before the original captcha timed out, you had successfully persuaded a "human" (sub-level in intelligence, but competent enough) to read the captcha for you.

An immediate result was that most captchas now time-out, and a new one is substituted after a very much shorter time. Many more users also give you a new and different captcha after each "wrong answer," but this just gives better odds that a relatively crude OCR nethod will get ONE it can guess right at.

Some changes may have helped, although there's been little subsequent discussion of how the reliability of the things has actually been improved.

The captcha has not been considered a particularly reliable identity test for some years by most in the AV industry. It has some use in blocking bots at the point of creating a new account, but shouldn't be used once a secure password (perhaps with "challenges") has been created to identify the specific person authorized to have access to the account.

The "leap forward" to permit machines to read captchas is quite similar to the improvements in more mundane OCR programs needed to allow fewer errors in reading a wider range of fonts and/or handwritten (cursive or printed) text. There has been a lot of improvement in OCR programs recently, and some OCR programs can now be called "pretty good" but none are really error free - yet. (One recent book scan "read" a decorative flower border on a picture as Chinese text for me, and my OCR is supposed to be one of the better ones available.)

John

Subject: RE: BS: The Singularity is Near / Captcha
From: Don Firth
Date: 18 Nov 13 - 01:33 AM

I hope that no one is forgetting about Isaac Asimov's Three Laws of Robotics:

1. A robot may not injure a human being or, through inaction, allow a human being to come to harm.

2. A robot must obey the orders given to it by human beings, except where such orders would conflict with the First Law.

3. A robot must protect its own existence as long as such protection does not conflict with the First or Second Law.

Don Firth

Subject: RE: BS: The Singularity is Near / Captcha
From: Ebbie
Date: 18 Nov 13 - 02:22 AM

Do we know for sure that Asimov actually knew the rules? :)

Subject: RE: BS: The Singularity is Near / Captcha
From: GUEST,michaelr
Date: 18 Nov 13 - 02:28 AM

Were Asimov's laws incorporated into Windows? If not, we're fucked.

IIRC, "singularity" is another term for a black hole. Stephen Donaldson famously invented a "singularity grenade" which could be stuck onto an enemy spaceship to create a miniature black hole.

Anyway, this is the first thread started by Songwronger that surprised me by not being completely moronic. It could happen. But by then, planet Earth would likely be devoid of the capability to support human life.

Subject: RE: BS: The Singularity is Near / Captcha
From: Don Firth
Date: 18 Nov 13 - 02:44 AM

I think so, Ebbie, considering that he was the first one to enunciate them in his I, Robot, and a whole ream of other stories about robots, some of them having to do with how the rules played out, sometimes in surprising ways. I think Isaac Asimov was a bit of a genius.

Don Firth

Subject: RE: BS: The Singularity is Near / Captcha
From: TheSnail
Date: 18 Nov 13 - 06:15 AM

Once we achieve faster than light travel, our intelligence (trillions of times greater than it is now) will spread outward until it permeates the universe.

If this was going to happen, it would already have happened somewhere else in the uiverse and reached us by now. Perhaps it has and we are all part of a computer simulation.

Subject: RE: BS: The Singularity is Near / Captcha
From: Mr Red
Date: 18 Nov 13 - 08:38 AM

if it is the Ray Kurzweil I have seen on telly - he intends to live to see the Singularity - whenever it happens. He runs an AI company.
Curiously enough there was an article in the New Scientist recently that a Californian company reckon they had the algorithm that would do this. Their target is far more serious, a generic method of finding patterns and information on the web for health professionals.

I used to read Science Fiction stories where a whole planet was a computer (networks weren't mentioned in those days). I fancifully thought in terms of 100's of years before we saw an inkling of a hint of such. BUT it is reckoned that (pick a figure) 10% - 25% of electrical energy is burnt in computers - mostly PC farms of the like of Farcebook, Googly & Amazing. And what have we got right now? The sum total of human thought in one humungous entity - OK I forgot to factor in the mighty Google dollar, but Wiki, Google Earth, Google Voice Search etc are significant steps to a a...............
well what a Deep Thought ...... apologies to Doulas Adams

Subject: RE: BS: The Singularity is Near / Captcha
From: Jack Campin
Date: 18 Nov 13 - 11:21 AM

Once we achieve faster than light travel, our intelligence [...] will spread outward until it permeates the universe.,/b.
If this was going to happen, it would already have happened somewhere else in the universe and reached us by now.

Equipped with faster than light travel, stupidity would travel just as far and fast.

The evidence suggests it got here first.

Subject: RE: BS: The Singularity is Near / Captcha
From: Rapparee
Date: 18 Nov 13 - 12:09 PM

The Universe contains, among other things, intellitons and stupidons. Like eelctrons and positrons, one is the negative of the other. The Universe is composed of stupidons, which like water to a fish, we do not recognize because stupidons are all-pervasive. When an intelliton is formed (the exact process is still unknown) it is immediately recognized simply because it is different, like air bubbles in water. Like air bubbles, intellitons have a life span of picoseconds to centuries but they will eventually burst (or be overcome by stupidons) and stupidons move into the void. Not much energy is released by the death of an intelliton and even less by the death of a stupidon -- but stupidons seem to have a much longer life span.

Subject: RE: BS: The Singularity is Near / Captcha
From: Ebbie
Date: 18 Nov 13 - 01:41 PM

Do we know for sure that Asimov actually knew the rules? :)

I think so, Ebbie, considering that he was the first one to enunciate them in his I, Robot, and a whole ream of other stories about robots, some of them having to do with how the rules played out, sometimes in surprising ways. I think Isaac Asimov was a bit of a genius.

Don Firth

My point, Don, is that I think it just as likely that someone will eventually invent/create robots that are programmed to do everything from robbing banks to holding hostages. Not a pleasant thought.

Subject: RE: BS: The Singularity is Near / Captcha
From: Don Firth
Date: 18 Nov 13 - 02:41 PM

Unfortunate, Ebbie, but true. Look at how many scams are carried out on the internet. Unfortunately, with any advance in technology, the vermin soon come crawling out to try to take advantage of it.

But in the event of actual robots, the Three Laws of Robots would be a safeguard that should not be overlooked. They started as a science fiction concept, but should there be major developments in robots, particularly androids (robots that are indistinguishable from humans--think Commander Data on "Star Trek: The Next Generation," especially If no one knew "he" was an android), they could become a very important safeguard.

Isaac Asimov's Robot stories are a darn good read, while at the same time are an examination of both the beneficial side and the disastrous side of advances in artificial intelligence.

There is the story about when Asimov and a friend went to see "2001: A Space Odyssey." When Hal 9000, the shipboard computer started trying to wipe out the human crew, Asimov snarled to his companion, "They didn't program Hal with the Three Laws!" His companion responded, "Strike them with lightning, Isaac!"

Don Firth

Subject: RE: BS: The Singularity is Near / Captcha
From: Bill D
Date: 18 Nov 13 - 05:22 PM

Human: "Look what you've done, R. Edify Shultz! He's dead!"

R. Edify:"He was human? I thought he was just another Republican! He certainly wasn't rational when I questioned him."

Subject: RE: BS: The Singularity is Near / Captcha
From: Songwronger
Date: 18 Nov 13 - 08:17 PM

There's an article by Bill Joy called "Why the Future Doesn't Need Us." Wired magazine, still posted on their site I believe. He wrote it after meeting Kurzweil. I think Kurzweil is into the whole immortality thing. Aside from Artificial Intelligence, some people look forward to uploading their consciousness into computer systems. Kurzweil said in his book that he took lots of pills, had reduced his aging and so on. So, he wants to upload his life experience into some machine. He didn't say that people with uploaded consciousness could "live forever,," but for "as long as they wanted."

I forget what Kurzweil said about the Rules of Robotics. Both he and Joy surmised that humans (us, in our present forms) wouldn't be in any danger from smart machines. The new AI would be a million times more intelligent than us in just a few iterations, so what harm could we pose? Can't recall if it was Kurzweil or Joy who said we'd basically be kept as pets.

For some reason I just thought of They're Made Out of Meat by Terry Bisson. A short, humorous story about intelligence.

And here's the Bill Joy article
Why the Future Doesn't Need Us.

Subject: RE: BS: The Singularity is Near / Captcha
From: Mr Red
Date: 19 Nov 13 - 05:20 AM

The evidence suggests it got here first.

"You talking to me?"

Subject: RE: BS: The Singularity is Near / Captcha
From: JohnInKansas
Date: 21 Nov 13 - 12:36 AM

The tone of the thread suggests some might be interested in:

Science-fiction museum to open in Washington, D.C.
Tanya Lewis LiveScience
20 November 2013

Fans of Captain Kirk and Captain Nemo unite: A new science-fiction museum coming to Washington, D.C.

Its creators announced plans for the museum, which will cover a broad sampling of science fiction across literature, television, film, music, video games and art, on Nov. 4. They hope to open a preview space within a year, and launch the full museum in the beginning of 2017.

"There really wasn't a comprehensive science-fiction museum here in the United States or internationally," said Greg Viggiano, executive director of the new venture. "I thought, maybe somebody should do something about this," Viggiano told LiveScience.

The museum will showcase seven distinct galleries, featuring the creators, vehicles, time travel concepts, aliens, computers, robots and technology. The content will start with "Frankenstein" author Mary Shelley, considered by many the first sci-fi author, and run through to present-day material. "We're going to try to cover sci-fi in a pretty comprehensive manner," Viggiano said.

Several collectors have already offered to loan their materials to the museum. These include models of the Enterprise E used in "Star Trek: Insurrection," and Amargosa Observatory and Deep Space 9 in earlier "Star Trek" films. Another collector has offered to loan their collection of "Battlestar Galactica" filming miniatures.
The museum has forged a partnership to store hundreds of interviews with sci-fi greats, from Ray Bradbury to Arthur C. Clarke.

As for Viggiano himself, he likes the classics — Jules Verne's "Twenty Thousand Leagues Under the Sea" (and the Disney version), and "Return to the Forbidden Planet," so expect those to make an appearance.

The group has received significant interest already, Viggiano said. "The volunteer response has been overwhelming," and includes people with professional museum experience, he said.

The museum will be funded by private donors and through a crowd-funding campaign on the website IndieGoGo.com. The group had raised $23,914 on the site as of Nov. 19, and hopes to raise $160,000 by Dec. 11.

A big part of the museum's mission is to boost STEM (science, technology, engineering and mathematics) education, Viggiano said. Many of the exhibits will be interactive, geared toward both kids and kids-at-heart.

[Of course this may all be science fiction that will never happen.(?)]

John

Subject: RE: BS: The Singularity is Near / Captcha
From: GUEST,Troubadour
Date: 21 Nov 13 - 10:46 AM

The three laws are not sufficient in themselves, as Asimov himself pointed out.

Any half decent AI would work out how to circumvent them in seconds.
Whether it would choose to is another matter, but I think it likely if human orders conflicted with its greater knowledge.

It might even decide that there is no logical reason to keep us around.

Good thread SW. More of this kind of chat would be good.

Subject: RE: BS: The Singularity is Near / Captcha
From: GUEST
Date: 14 May 14 - 12:37 AM

Stephen Hawking: Dismissing artificial intelligence would be a mistake

LONDON, May 3 (UPI)

Stephen Hawking, in an article inspired by the new Johnny Depp flick Transcendence, said it would be the "worst mistake in history" to dismiss the threat of artificial intelligence.

In a paper he co-wrote with University at California, Berkeley computer-science professor Stuart Russell, and Massachusetts Institute of Technology physics professors Max Tegmark and Frank Wilczek, Hawking said cited several achievements in the field of artificial intelligence, including self-driving cars, Siri and the computer that won Jeopardy!

"Such achievements will probably pale against what the coming decades will bring," the article in Britain's Independent said.

"Success in creating AI would be the biggest event in human history," the article continued. "Unfortunately, it might also be the last, unless we learn how to avoid the risks."

The professors wrote that in the future there may be nothing to prevent machines with superhuman intelligence from self-improving, triggering a so-called "singularity."

http://www.upi.com/Science_News/Technology/2014/05/03/Stephen-Hawking-Dismissing-artificial-intelligence-would-be-a-mistake/7801

Subject: RE: BS: The Singularity is Near / Captcha
From: GUEST,michaelr
Date: 14 May 14 - 03:30 AM

Aww, nooo! It'll be bad for us? Really? The OP made it sound like such a cool thing!

In any case, there's no time for us to watch this happen. See this thread.

Subject: RE: BS: The Singularity is Near / Captcha
From: GUEST,Musket
Date: 14 May 14 - 03:47 AM

It's alright. We invented computers and the internet. We don't live in fear of them taking over.....

Mind you , the irony of reading this thread on my iPhone hasn't escaped me.

I recall reading of a debate where during questions from the floor, Asimov was questioned about the laws and said in some ways, his target audience likes the reassurance.

Arthur C Clarke quipped " I tend to write for a British audience. "

What ho!

Subject: RE: BS: The Singularity is Near / Captcha
From: GUEST,Grishka
Date: 14 May 14 - 05:35 AM

Computer software disastrously turning against mankind, including those who commissioned, programmed, and/or run it, is not SF, but reality. See the stock market.

Asimov was a romantic.

Subject: RE: BS: The Singularity is Near / Captcha
From: GUEST,Musket
Date: 14 May 14 - 05:47 AM

"Turning against mankind?" Did The Titanic turn against mankind?

GIGO is the more rational explanation. Poor algorithm development.

A singularity by the way, was predicted before computers were developed. Probability and uncertainty, if on a large enough scale, produces predictable outcomes or fuzzy logic.

See Heisenberg for details.

Subject: RE: BS: The Singularity is Near / Captcha
From: GUEST
Date: 14 May 14 - 08:08 AM

There is nothing probable or uncertain about what's being discussed here. The price/performance ratio makes intelligent machines inevitable. When they can reproduce themselves (and improve themselves with each new generation) we will be in their way. We will either be eradicated or tolerated, much like pets.

Subject: RE: BS: The Singularity is Near / Captcha
From: GUEST,#
Date: 14 May 14 - 08:19 AM

I read a tweet (connected to a news story) in which the young lady said something like 'OOOMG i just found out that the Titanic was a real event, not just a movie'.

If technology and its uses hasn't scared us yet, it should have.

Subject: RE: BS: The Singularity is Near / Captcha
From: Uncle_DaveO
Date: 15 May 14 - 05:30 PM

I believe someone above opined that Mary Shelley was the
first science-fiction author.

That honor, as far as I can see, would go to whoever, a LONG time ago, authored the Golem story.

Dave Oesterreich