The Singularity

Will IBM's Watson beat the world's two greatest Jeopardy champions at their own game again today?   If yesterday's match is any indication, it will.  Not only did Watson win, I think it would be a fair statement to say it creamed the humans.  After witnessing Watson's unmatched buzzer speed and incredible knowledge-base, I must say I am more than impressed.*

Appropriately, the cover of last weeks Time Magazine had the following headline:  "2045 The Year Man Becomes Immortal."  Inside, I found an interesting article about a topic that has been on the tip of conspiracy theorists tongues for a couple decades, and has lately garnered the attention of the intellectual elites of the world as well.  If you haven't guessed already, I am talking about the Singularity movement.

Though my regular readers may find it hard to belief, I have yet to completely form my opinion on this topic.  It's not that I'm not leaning towards buying into the theory.  There are parts that intrigue me, and seem not only plausible, but probable.  Specifically, I agree with the exponential direction our technological growth is taking.  However, I am not certain that I understand the conclusion Singulatarians (is that a word?) imagine.  In fact, the diversity of ideas they seem to offer on just where this exponential growth will take us seems like proof that they don't really know themselves. (Which is odd for a movement that defines itself by a term implying a single, unifying moment of time somewhere in the future.)

So, in an effort to catalog what I learn about this interesting topic, I have decided to share a couple of the points I highlighted as I read the article from Time.   (All quotes to follow are from pages 46-49 of February 21, 2011 issue of Time Magazine.)

As I mentioned in the paragraphs above, I have no problem buying into the idea that "technological progress happens exponentially, not linearly."  This seems like common sense to me.  Furthermore, I love the "interdisciplinary nature of Singularity theory."  As a reader and adorer of Ken Wilbur's interdisciplinary approach to history and the humanities,** I find it refreshing that technological gurus are willing to step outside of their own fields.

I also heard about the age-defying potential of telomerase before, on one of those news shows like 60 minutes:  "They (researchers at Harvard) administered telomerase to a group of mice suffering from age-related degeneration.  The damage went away.  The mice didn't just get better; they got younger."    Which seems to prove prove Aubrey de Grey's point quite well:  "'The whole of medicine consists of messing about with what looks pretty inevitable until you figure out how to make it not inevitable.'"

The fact that our ability to make machines smarter at an extraordinary rate isn't debatable.  However, predicting the end results is all purely speculative.  Instead of death, for instance, will people choose to have their brains transformed into data, and scanned into a computer?  If that happens will the consciousness inside of the machine be the same as the person who once inhabited a body?  And what happens if that data is deleted, or worse, corrupted?  An eternity as bad code seems worse than death to me.  Talk about a great idea for a short story...

What happens when computers crack the human genome to the point that they can manipulate it?  Will computers create humans with tails and horns?  Or will we try to force them to make us immortal in our fleshy shells?  Imagining the social implications of this discovery is enough to give me a headache!

Whatever the answers, I agree completely with the author of the article, Lev Grossman:  "Nothing gets old as fast as the future."

*I find it interesting that it took me until my third edit to realize I was referring to Watson as "he" and not "it."  Apparently it's human characteristics have my brained convinced computers have already developed to be like us.

**Ken Wilbur postulates that all knowledge develops exponentially, coincidentally.
___
2/17
Computer finishes off human opponents on 'Jeopardy!'  (CNN, by Jason Hanna, 2/17/2011)

My husband and I DVR'd all three episodes of Jeopardy featuring Watson.  I must admit that I was impressed by the computer's performance.  Probably the biggest factor contributing to it's success was the speed it knew whether or not it had reached the threshold to buzz in.  Almost before Alex had finished reading the question, Watson's screen was lit up.  The human contestants were often furiously trying to press their buttons too, but to no avail.

However, it was obvious that it would still need some tweaking.  For example, Watson does not hear or see.  Therefore, if another contestant had already given an incorrect answer, there was no way for it to know.  Also,  if innuendo's were involved in the clue, Watson sometimes gave hysterical replies.

I hope that IBM continues to fidget with Watson, and calls for a rematch in the future.  Or, perhaps a better idea would be to have them try to develop another computer designed to beat Watson!
___
2/20
The Greatest Massacre in Irish and British History:  The Truth Comes Out  (Arch News)

As an example of how computers can be used to do the work of humans at incredible rates, I found the above linked story about deciphering depositions from the 1600's in Ireland especially intriguing.

"The team behind the study had unprecedented access to IBM’s LanguageWare© technology research team in its Dublin Software Lab and developed a unique suite of software solutions including a program called Wordsmith to analyse the Depositions and to cross-correlate an array of features of the text – a process which would be too complicated and potentially take a lifetime for a scholar to undertake manually."

As a lover of history, I find this use of super-computing especially amazing.  Especially given the implications of this particular story on the present state of affairs in Northern Ireland.

"The Depositions by Henry Jones in early 1642 of attacks on Protestants in the early months of the rebellion, the only problem was it was hersay and fed the rabid fear that divided the population along religious lines.  These division were to last and multiply with tragic consequences to the present day and formed the foundation for the Northern Ireland troubles."

I  wonder if knowing one's cultural stance toward another social group was based on lies would be enough to reverse the deep-seated hatred of the Irish people.
___
3/10
Inventor, futurist predicts dawn of total artificial intelligence  (Rocco Castoro via CNN)


When I read the opening paragraph of the article linked above, I felt like I do at the beginning of a scary movie.

"In the year 2050, if Ray Kurzweil is right, nanoscopic robots will be zooming throughout our capillaries, transforming us into nonbiological humans. We will be able to absorb and retain the entirety of the universe's knowledge, eat as much as we want without gaining weight, shape-shift into just about any physical form imaginable, live free from disease and die at the time of our choosing."


But, as I listened to Mr. Kurzweil talk in the attached video I found myself not only buying into his theory, but excited about my prospects for a brighter future.  Let me backtrack a moment, though, and share a couple of the cool examples of exponential technological advances we can see today. 


The most compelling example Mr. Kurzweil gave, in my opinion, was how computers have shrunk over the years.  At one time, a single machine took up an entire room.  Today, computers fit in our pockets in the form of cellular phones.  Why, then, would it be implausible to assume that a decade from now, they will shrink even further, to the point of being able to navigate our blood vessels? 


Less than ten years ago, Kurzeil points out, we lived without search engines.  Tomorrow, thanks to nanotechnology, we may be able to open e-mails with French Toast attached from Paris. 


One of the places the rapid rate of advancement is most evident is in gaming.  I can attest to this personally, and think it would be so cool if, by 2030, our nervous systems do all of the work of controlling the avatars.  My disabilities will no longer hamper my playing.  The winner will truly be determined by whose neurons fired quickest, not whose fingers reacted first!  Breaking the limitations of biology is probably the most promising prospect of the Singularity theory in my opinion. 


When the interview turned to the topic of Mr. Kurzweil's own daily regiment of supplements, the "scam" light bulb immediately started flashing in my head.  Then, the talk of crossing various bridges brought the nasty term "cult" to the forefront of my mind.  But these derogatory terms were quickly forgotten as I heard him do what no scam artists or cult leader do:  admit the possibility of terrible side effects.  The example he gave was  terrorists using the advances in medical science to create super-strains of the flu.


By the end of the movie, I couldn't help but be inspired.  If, as Timothy Taylor predicted, it is technology that made us human, coming together with our creator could truly be a singular experience. 
___
5/6
What will happen to us?  The Boston Globe, by Graeme Wood, 5/1/2011

The above linked article rocks.  It isn't about the Singularity specifically, but rather a great number of possible options for humanity's future.  The Singularity is included, but is far from the most outlandish proposition.  But the point of the article is not necessarily to list outcomes, but rather to prove predicting the future is important for the acting in the present. It is worth the read.
___
5/23/2011
'Jeopardy!'-winning computer delving into medicine  (MSNBC, by Jim Fitzgerald, 5/21/2011)

Would you trust a robot at your emergency room bedside after suffering a heart attack?   Would you question it's diagnosis?  Would it's computerized analysis of your symptoms leave you feeling as safe, or perhaps safer, than a living doctor?


This scenario may become common in the next decade, and Watson is charting the course.  "Watson is being fed a diet of medical textbooks and journals and taking training questions in plain language from medical students."  Eventually, he will spit out diagnosis just like he did the answers on the famous game show.  And he will even include a percentage representing his certitude. 

But what about those really bizarre cases, where the symptoms don't seem to match up with any known disease?  Perhaps the next challenge for Watson is to catch up on as many episodes of House as possible.
___
9/8/2011
Software tricks people into thinking it is human  (New Scientist, by Jacob Aron, 9/6/2011)

The Turing Test is simple.  Participants must engage in a computerized conversation, and analyze the responses to determine if the conversant on the other end is human or mechanical.  The results are tabulated into percentages.  Cleverbot recently scored nearly as high as real humans in the largest test of its kind ever conducted.  

But what does this mean?  In my opinion, it is not so much Cleverbot's ability to retrieve information from it's database, so much as it's ability to learn the material in the first place.  Learning involves the proper categorization of new ideas, and language is especially deceiving at times due to figures of speech, such as innuendos.  Thus, Cleverbot's ability to converse like a human impresses me.  Cleverbot quickly conducts searches and then spews out a response, similar to what we do.  The only difference, I believe, is that we do it without thinking.  It comes natural.  Cleverbot, on the other hand, needs to devote all of its power to imitate human conversation.