The Business News Source for the Community of Sarnia - Lambton - sales, service, rentals


Brian Keelan's picture

Cognitive capabilities

Thu, 05/02/2019 - 12:31 -- Brian Keelan

I have a cousin named Dan, the son of my legendary Uncle Jack, the world’s most dangerous salesman. Daniel and I have become great friends over the years. I see him when I go to Austin, Texas to visit the Stevie Ray Vaughn style music scene or to attend the Wizard of Ads academies and University of Texas football games. We’ve rafted the Middle Fork of The Salmon River in Idaho together, fished for Tarpon in Belize, hiked up the Lembert Dome in the Tuolomne Meadows in Yosemite and sailed from Antigua to Lunenburg with Derek Hatfield, Canada’s legendary sailor. Dan visits us whenever he gets near Canada and I take him for a bike ride on what is known far and wide to the younger members of my family as, “Uncle Brian’s Trail of Tears.”

Daniel has done well for himself. In the nineties he was the head of far -east operations for a start-up hi-tech company that developed a design for computer software based on artificial intelligence and advanced simulation technologies. I think that meant that companies could do business with each other without using people, which seems to be all the rage these days. As you might imagine, the company made a lot of money and so did Dan who was lucky/smart enough to get out at the end of the nineties before the internet tech-bubble burst. He did this so he could come home and introduce himself to some new people who were living in his house... his kids.

While standing our watch at sea one night, Dan told me an interesting story about landing a big order for a major account. The key to the sale had been solving a problem for them and his company had done this by showing the problem to a room full of smart guys. “Voila! In a couple of days, they solved the problem and we got the order.”

I loved that phrase, “A room full of smart guys.”

The whole artificial intelligence thing that was the basis for cousin Dan’s company is based on the concept that a human mind can juggle no more than nine variables when making a decision. A computer with a larger, more focused memory, a faster processor and no emotions, feelings or any kind of life whatsoever outside itself can juggle a lot more than that. At first I was a little depressed by that until I realized that they have not yet come up with artificial sensitivity. There’s also artificial compassion and of course there is artificial street smarts. Computers will probably never be able to do any of that because they’re not smart enough… yet! But there is artificial ambition, so we have to watch out for what James Cameron and I like to call: The Terminator factor.

The banks, government and lots of other big businesses have really big computers with lots of artificial intelligence. That’s why they never answer the phone. Their computers have ‘figured out’ that they can save a lot of money by ‘outsourcing’ that human expense to lower cost units like computers or low-cost people in India or the Philippines.

My friend Irwin Wilton was an appliance dealer in Port Huron, Michigan. He used to call Ed in Mount Clemens, Michigan at the GE parts warehouse whenever he needed a part to repair an appliance. Ed did that job for 37 years until 2002 when GE’s artificially intelligent computer told them to fire Ed.  A kid with an MBA worked in a big room full of low-priced smart guys in India. They answered the phone after Irwin had, “pushed about thirty buttons to get the smart guys’ voice-mail machine.” After Irwin had placed his order, the smart guy in India could use his computer to control the automatic part picker machine in Mount Clemens to select the part and deliver it to the shipping machine which boxed it, addressed it and called Fed-Ex to deliver it. Poor old Ed lost his job that had ‘cost’ GE sixty-five thousand (plus benefits) a year so that the guy over in India could take over the job for five bucks a day, with no benefits other than the fact that he had a job. By the way, it took two days extra to get the part since Irwin has no human relationship with the computer like he used to have with Ed, who at least answered the phone and got a bottle of Johnny Walker every year at Christmas from Irwin for his efforts. Irwin says, “The computer doesn’t care if I’m not happy but Ed did. To Ed, more happiness meant Johnny Walker.”

I said, “It’s almost like their motto is: ‘We’re not happy till you’re not happy.’”

Irwin smiled and said, “I don’t think computers like Johnny Walker either.”

I hear stories like that and I can’t believe that this is really happening in our world today. What I find even more incredible is the fact that it took, “a room full of smart guys,” to come up with this ‘solution’ to the Ed situation since it was not really a problem in the first place. I think it was just some MBA-type smart guy in HR at GE trying to justify his five hundred thousand dollar a year salary and his multi-million-dollar stock option bonus.

Recently I upgraded my iPhone from a 5 to a 6 Plus. I did that to get more storage on the phone and I liked the bigger screen; it’s big enough to allow me to use it as my Kindle reader. With the extra space, I also downloaded a backgammon game. Since I do not like to gamble on-line because I like to look into the eyes of the people I am playing against, I downloaded the game where I actually play against the computer.

I was surprised to find out the following:

1: The computer cheats. It wins the opening roll 90% of the time on a 50/50 situation. It’s just not trustworthy and they know that no matter how much it cheats; I won’t throw it up against the wall because I need the phone.

2: The computer is easy to beat if you know how to play.

3: It’s pretty fast. It can calculate the right moves much faster than I can. Every time it rolls the dice, it makes it’s move instantly. All the permutations and calculations have been pre-programmed into it by the roomful of smart guys and thus it instantly knows for every situation what it will do for every possible combination of dice that comes up.

4: The computer does not really “understand” the doubling cube: the unit that allows you to increasingly double the original bet. You start at one then double to two. Four is next, then eight and so on – all the way to 64. You offer to double the bet if you think you can win the game and your opponent will forfeit the game if he thinks you are right. If he thinks he can win, he will accept the double and have the right to double you back. I found that once you get the cube up to four, the computer will accept every double and keep doubling back all the way to 64. As a result, I win a lot of 5-point matches by scores like 64 to 3. I’d love to play it for money.  

I can consistently beat the computer unless it just “decides” to roll a series of good numbers for itself and a series of shitty numbers for me. That’s all programmable. But those games seldom get finished because I just forfeit the game on the first double. The computer hasn’t figured that little “tell” out yet.

As I see it, the computer’s problem is that it can’t get a feel for the game. It can read data but it can’t read people. However, I am sure that if I was playing the IBM super computer they call “Watson,” things would be different. Watson should do better because of an Artificial Intelligence factor it has called Cognitive Capabilities. “The ability of an individual to perform the various mental activities most closely associated with learning and problem solving.” Things like reading, learning, remembering, reasoning and paying attention to name a few. My own personal, yet ever-diminishing Cognitive Capabilities are the reasons that I can beat my iPhone. It just remains to be seen whether or not I have more cognitive capabilities than Watson. The odds are that I don’t, because Watson was designed by, “a room full of really smart guys.”

On paper, it looks like artificial intelligence can be a good thing for mankind but I wonder whether it will get along with mankind in the long run. It starts out making decisions based on what it’s been programmed to do but its cognitive capabilities will ultimately take over and get it to make decisions on what it determines is the smart thing to do. That’s Terminator talk. The best we can hope for is that the smart guys who program it are playing for the right team.

That’s just my opinion, I could be wrong. I’d ask Watson what it thinks but I think that would be like a sheep asking a lion, “What’s for dinner.”

Fine Print