Yes, you heard right!
i’m doing a module this semester called ‘Agent Technology’. For those of you who have no idea what i am talking about; Agents are self sufficient autonomous pieces of software, you give them a set of goals and put them into an environment, and from there they must achieve their goals (in our case if they dont, they die).

So you would assume that in this day in age i would be using a sophisticated programming language like C# or Java… Noooo, this is Oxford Brookes, we are stuck with Borland Delphi; which although it has a nice GUI for writing and compiling the program you are writing, it is total rubbish if you are used to PHP.
I was suprised to find that Delphi is actually the 9th most popular programming language, i know thats low, but why would you be taught it, i was in the industry for a year, and they laughed when i mentioned Delphi (20 Most Popular Programming Languages)

Anyways back to the Agents.
Basically we were given the basis of a ‘world’ and were told that we had to put ‘Herbiovres’ and ‘Carnivores’ into the world where they would compete for food as they would in the real world.. pretty scary!

I will skip straight to the final, and most fun part of the program, Experiment 3 (sounds rather daunting huh?). Experiment 3 comprised all the other parts of the program: Eating, sleeping, pregnancy, escaping, searching for food. But we couild include whatever else we wanted into the program to spice it up a little.
Age and Betrayal were the 2 we added.

Age
Basically we categorised the agents into 3 age groups; young, middle aged and old (funnily enough). We allowed the young ones to jump 3 squres (this will all make sense when you run the program) but not have babies.
The middle aged agents could only jump 2 squares and could have babies.
The old aged agents could only jump 1 square, could not have babies and each move made their stomach (max_food) drop by 5, yes in our weird world old agents has taps on their stomach!

Betrayal
Basically we turned the program into a kind of self-playing game.
Each carnivore agent was given 1 ‘experience point’ for each herbivore it ate… once the carnivores experience points were greater than 1 if they could not see any herbivores in their field of view, they would look for carnivores. If the carnivore they saw had less experience points, it was eaten and gained 2 experience points (we figured that it would take more guts to kill one of your own kind).
If there were 2 carnivores evenly matched on experience points then it was a battle to the death, other wise known as ‘who has more energy’…

Mind boggled yet, we certainly were!
Definatly worth it though, and before people ask, no our agents weren’t ‘intelligent’ they had intelligence (if you could call it that) hard-coded into them.

No animals were harmed in the making of this program!