Hi folks! What's your take on putting algorithms to compete in order to find the best performers? Would you share your experience in different industries? I'm currently working with trading algorithms. We are building a Collective Trading Intelligence with a very ambitious goal: to outperform the trading technology created within investment companies currently dominating financial markets. The plan we are implementing involves enabling a global crowd to collectively create and evolve trading algorithms by putting them to compete to find the best specimens. The trick is that right after each competition ends the source code is opened for the community to improve and retrain the winners… and put them to compete again, unleashing a relentless evolutionary process.
Oh no! We do not expect people to bring their algorithms and make them open source... Instead, we encourage people to come together and build algorithms in a cooperative fashion, most of the times starting off with algorithms that already exist and are openly shared within the community for everyone to be able to improve on. So it's actually the other way round. People are not supposed to come in and share their algorithms. The way it works is that people come in, and we share our algos with them. It is up to them if they wish to become a valuable member of the community and start contributing knowledge to improve on the community's collective wisdom, or not. Moreover, the open source process happens only in competitions. People can still run their closed source bots and rent them to people through the marketplace. So, beyond working as a driver of evolution of the collective intelligence, competitions are a way to earn reputation. Investors would naturally want to hire the best service providers... those that are winning competitions!