My first computer (the Acer XT) was bare bones, MS DOS, no Windows nor anything fancy. I stumbled by chance onto Microsoft Basic while snooping through the computers system files. Basic came with ~4 example programs one of which was the classic ‘Snake’ game. The XT was so limited that it would not support the graphic mode for Gorillas – another example program that came with Basic.
It was when I realised that I could change the code and that would affect how the game worked which lured me into the world of programming – I was hooked! I made my own versions of Snake that my friends and I would play during lunch at high-school. I still have an interest in game development but I what I’m working on now is neural network programming.
In simplest terms, think of hundreds of little boxes with wires connecting them in a seemingly random fashion. Now if you had a bunch of wires which allow you to send a signal into some of the boxes you would call this the neural network input and a bunch of wires running out from some of the boxes you would call these the neural network outputs. Nothing fancy.
The fancy bit is the little boxes can change which boxes they connect to in the network.
Behind the scenes there is a huge amount of maths going on. The more neurons in the network the more maths needs to be done for the neural network to ‘reconfigure itself’ in a way that yields the best result (output) for the given problem (input).
Today’s problems are all about solving simultaneous equations and neural networks lend themselves to providing a ‘good enough’ result without having to work out all the equations. For example, weather forecasting, character recognition and stock market prediction.
I’m working on a reproduction of a biological neural network also. I believe we have the technology in our desktops to replicate the 100,000 neurons required to simulate a fly however we lack the architectural understanding to put all the pieces together. Replication of an existing network is probably our best hope.
Below is a screen shot of an application i built to predict trends in the ASX. The bars in red are the inputs and the bar in grey / black is the output for the next days trade.

I’ve been working with web technologies for a eight years now. To check the latest syntax of HTML tags I use w3Schools. To see what’s up-and-coming HTML I like to read the w3c specification.
In Australia, even after the SOCOG debacle, we only enforce a level A accessibility requirement for websites at a federal government level (which is pretty poor). The double A level requirement comes in at the end of next year.
I really liked checking the accessibility of my HTML using Bobby until it was shut down by IBM in 2008. The w3 have a list of validators, to achieve triple A accessibility rating it is imperative to do more than simple code validation.