Computer named Watson challenges human intelligence
The show featured Ken Jennings, who had the most games won in succession with 74, and Brad Rutter, who was, at the time of taping, the only undefeated contestant in the shows history, in a tournament filmed and televised in mid-February.
The matches were conducted over a course of three days and the format was a two-game total point exhibition match.
So what does this all mean? Should we be nervous of Watson’s achievement, or should we embrace it?
The defeat is a massive step ahead for AI, finally showing us that computers may be able to understand the complex syntax and extra-lingual components of human language very soon.
One of the downfalls of AI is that it has great difficulty processing human verbal language very well because of the nuances and sarcasm and double-meanings that we fail to notice in everyday speech.
Watson instead had a .txt file sent to it at the very same moment that Jennings and Rutter had the questions verbally read to them by Alex Trebek, and incorporated statistical algorithms based on its complex connection of servers that have a memory capacity nearly 14,000 gigabytes of stored information about various inputted knowledge to arrive at the most probable answer.
Are humans in Jeopardy of artificial intelligence (AI) one day being able to outthink us?
I do not personally see how having advanced AI poses as an immediate threat on our humanity.
But I stick to Oberg’s dictum wih this prediction, keeping an open mind, but not so open to believe everything. Look at how fast Internet went from dial-up to becoming lightning fast and able to be handled by mobile cell phones. Just over a decade.
And when I use the term “outthink’” I should clarify that computers are simply retrieval machines full of information that we as humans put into them about our world and discoveries.
When you enter a complex mathematical problem into Microsoft’s calculator accessory, the computer isn’t thinking of the answer, at least not in the respect of how a human would.
If you want to argue that computers do think, you would have to redefine the term so it bears the clear distinction between human thought and “computer thought”.
Human thought is a precious gift. It enables us to rationalize and reason. It’s what sets us apart from the plants and animals, and for the moment, the computers.
Could computers one day reach our intellect and reach the ability to rationalize and reason about ideas and theories?
Although the evidence is scant, Thomas J. Watson himself (the very man who Watson is named after) was quoted as saying he “think(s) there is a world market for maybe five computers.”
So I hypothesize “of course” (to avoid running the risk of sounding conservative).
But will we reach this soon? It’s all relative.
One hundred years is soon, depending on what you’re talking about, and when it comes to computer AI, considering how fast we’re progressing, but keeping in mind the current limits of computer “understanding” and “thinking” I think AI has a lot of work to do as of now.
Each breakthrough for AI is a piece of the massive puzzle that is the eventual event of AI “outthinking” humans. So it’s a long journey yet.
Speculation about the limits and possibilities of robotics and AI should always be in accordance to Oberg’s dictum, and not based on current trends and limitations.
To illustrate this, how do you think Alexander Graham Bell would react if you placed an iPhone 4 in his hand?
He would probably denounce the impossibilities of wireless telephone service and you’d look like a first-class lunatic. Keep an open mind, but not so open your brains fall out.