Image 01

21st Century AI

A blog about achieving meaningful Artificial Intelligence

Archive for June, 2012

Cassandra

Tuesday, June 5th, 2012

Cassandra after the fall of Troy; nobody believed her.

Cassandra was a princess of Troy and was simultaneously blessed with the gift of prophecy and cursed because nobody would believe her predictions. I’m getting a real Cassandra complex.

From time to time I get a phone call from somebody at a company that has a problem. They are usually pretty vague about the specifics of their problem, but they would like me to fly out there and take a look. The first time this happened was about 12 years ago and the company was a medium-sized, public traded operation that was best known by the three initials of its corporate name. The company had been pretty successful for the last five or six years and wanted to expand into the ‘lucrative’ computer game business.

I can’t remember how much they told me about their ‘problem’ before I got there or how much I had to figure out when I was there but the bottom line was that they were coming up on a deadline and they were getting a little concerned about making it. This particular deadline involved passing a beta test for a mystery game involving the licensed likeness of Humphrey Bogart. After seeing where they were on the project, and how much they had left to do before they were going to make it to beta (which was about six weeks away) I reported to the VP in charge that I didn’t think their team had a prayer of making the deadline.

“You have to have faith in your people,” the VP insisted.

“Not if they can’t do the work,” I replied.

Long story short: the company blew the deadline, was sued into nonexistence by the game publisher who had the other end of the contract and the team that the VP had so much faith in was promptly thrown out of work. And, by the way, I never got paid for my consulting, either.

Now, let’s fast-forward to about 18 months ago. Again, I get a phone call from another company that is best known by the three initials of its corporate name and, again, they’ve got a deadline and a problem. They fly me out to where they are located, a limousine picks me up at the airport and, again, I discover what the real situation with their project is: they’ve got a quickly approaching milestone deadline and their program won’t scale and is slower than molasses in an Iowa winter.

At the end of the day I report to the project Architect that they’re not going to pass the milestone test. Their code is written in Java which is far too slow to ever accomplish what they need and that I suggest that they bring in a half a dozen veteran computer game programmers (and install Mountain Dew dispensers and put pizza delivery on speed dial) and rewrite the whole thing in C++ or C#.

The Architect thanked me for my advice and politely informed me that my suggestions were not going to be implemented, the limousine took me back to the airport, the company failed the milestone test and DARPA cancelled their $35 million contract. Again, I was not paid.

About two weeks ago I got a phone call from yet another company best known by the three initials of its corporate name asking would I please take a look at their ‘serious game’ that they were developing for a government agency and, yes, they were quickly approaching a beta test deadline. I took a look at their project and promptly wrote back to them that their scoring system wasn’t doing what they thought it was doing and that they weren’t going to pass their upcoming milestone.

A few days later I got an email thanking me for my predictions and I haven’t heard from them since.

Oh, I haven’t got paid from them, either.