Of Asimov, Robots, Artificial Intelligence and What is a Human Anyway


You might say I have too much time on my hands. I would answer that I never have enough time, but my back injury continues and I have had time to think in a few directions.

Whether it is HAL 9000, remember “I’m sorry but I can’t do that Dave” as an answer to “Open the Pod bay doors HAL” from 2001 A Space Odyssey?

If you haven’t tried it, ask Alexa, Siri, Cortana or whatever your speech interface is to the internet, those famous words. “Open the Pod bay doors HAL” If only Arthur C Clarke was around to experience that.

Damn, I just remembered that they had a 4K restoration of the movie at Imax last month for the 50th anniversary of the movie. I was hoping to find someone to go with and then totally forgot about it. That would have been amazing.

AsimovI collect books and in recent years have given away many books that I was never going to read again, but decided to extend my collections of specific writers and starting at the beginning of the alphabet, I looked at what was missing from my Isaac Asimov collection and amongst others bought a copy of The Bicentennial Man.

Asimov is of course famous for the 3 Laws of Robotics. Ironically a lot of people debunked his laws and said they were flawed and used that to criticise him as being unrealistic or perhaps idealistic, which is a trait of many SciFi authors of the 70’s. However, he knew that himself. In many of his stories, robots disobeyed the laws.

There is a great story in this book called That Thou Art Mindful of Him, in which is a play on Psalm 8:4-6, he also infers in some of the stories that he was Jewish through some of the characters and had a keen sense of humor.

In this story (and I’m sorry for the spoiler) a series of robots are produced and given the capability to become self aware, in effect sentient. They redefine what it is to be human and declare themselves as such.

I played with the thought of Singularity and imagined if autonomous cars could pass the Turing Test 

I also looked at what might happen if they didn’t and what hackers might be able to do.

What I keep coming back to and writers like Philip K Dick, Asimov, Clarke, Heinlein and many others foresaw 50 and more years ago and similar to where the TV series Humans is heading, is that humans are dangerous to the planet.

Now I like being human and I hope that my descendants will have safe and healthy planet for thousands of years from now and many of my little stories are in jest.

BUT, if climate change, plastic pollution, air pollution, brinkmanship politics, drought, famine, and war are the result of how great and committed we humans fancy ourselves to be, would it not be realistic if an Artificial Intelligence was developed to the point of Singularity and able to continue to learn with or without programmed biases, would their logic determine that the human race should either be limited or allowed to exterminate ourselves?

Kurzweil looked at it a different way and said that Singularity would occur around 2045 and potentially be a synthesis between human and machine, in effect human 2.0. He would be about 98 at that point in time, so it will be interesting to see if he is still around and if he is right.

Maybe Elon Musk, founder of Tesla and many futuristic projects should have the last word. He’s pretty successful and walks the talk. DARPA, Rex Bionics and hundreds of companies, universities and other innovators are developing systems that will be able to think for themselves. Yes, for specific purposes, but they are being created.

It’s interesting that in this clip, they say that Science Fiction is usually about 50 years ahead of its time. So back to Asimov, reading him today, especially a book like The Bicentennial Man, where like Stephen King and others, he talks about his stories, was he in fact prophetic?

Yes, maybe I’ve had too much time to think, but do you think we should be thinking about this. Just imagined if a machine, say a Robocop decided that using facial recognition or perhaps racial recognition, that you were, could be, or could become a criminal and then think about biases that go into programming, often of necessity.

What conclusions could an AI start taking when given some information and some bias and then left to learn on the basis of that starting point? Oh and I didn’t even mention George Orwell. He wrote Animal Farm in 1945. Remember “All humans are equal, but some are more equal than others”? Shutting up now……..

 

Using CAPTCHA to digitse old books


This morning in the shower I was listening to Digital Planet, one of my favourite podcasts from the BBC. No I’m not sad, I just like to maximise my time:)

They interviewed Luis von Ahn about how Carnegie Mellon University is are using CAPTCHA technology to help digitse very old books that are in the public domain.

What is CAPTCHA? It is an acronym for “Completely Automated Public Turing Test to Tell Computers & Humans Apart” You will have found that many times when you register to use an application on the web, or perhaps when you want to invite someone to be your ‘friend’ on MySpace or leave a message, you will see a small clump of letters and you have to enter what you see into a text box.

The reason for this is because spammers and hackers create bots, that allow them to access information and pretend that they are real people. For example there are people running businesses where they can guarantee you lots of ‘friends’ on MySpace for a fee. Personally I am against this and ultimately it is a waste of time, because just being able to say that you have thousands of friends, doesn’t actually help you in any way.

Just to sidetrack for a moment. I have lots of ‘friends’ on my MySpace page, 3967 at last count. They are people who have requested my ‘friendship’ or vice versa and because of that personal relationship, wherever possible, I have a fan base that I can use when I have a concert or gig that I am performing. I can use this to make contact with them, even by geography, but that is really a topic for my About Songwriting blog.

Anyway, many organisations are trying to digitise as many books as possible to allow them to be read as eBooks. The best known of these is Project Gutenburg, which has already digitsed more than 25,000 books.

The problem with older books, especially those prior to 1900 is that the pages are fading and the fonts are harder to read by OCR (Optical Character Recognition) tools, which themselves are still not 100% reliable. On a tangent, I hate reading books that are not perfect. My eBook Unleashing the Road Warrior was edited 12 times to get it as good as possible and I was dissapointed to find an error on page 309 of Stephen King’s latest book, Dumas Key, but that’s another story:)

So what the Carnegie Mellon people have done is to scan the pages and have created a tool which grabs 2 at a time and feeds them into the CAPTCHA environment. So now when you complete a CAPTCHA that has 2 words instead of random letters, what you are actually doing is not only autheticating that you are indeed a human, you are also helping transcribe these old books and ensuring their texts are protected for future generations to enjoy, is that cool or what?

Personally I find CAPTCHAS a pain in the proverbial, but having learned this, I am feeling a lot better about them.

While this blog is starting to get a good following, I would love to get more readers and encouraging me to keep writing. If you feel that my blog is interesting I would be very grateful if you would vote for me in the category of best blog at the NetGuide Web Awards. Note that the form starts each site with www whereas my blog doesn’t and is of course https://luigicappel.wordpress.com.

Thanks so much for your support:)