Creating Your Own Neural Network Was Never so Easy

Another great guest post by Aubrey Hansen

Training Your Own Purpose-Built Neural Network Was Never so Easy

At their best, new technologies shift horizons and save vast amounts of time which in the process works to reduce costs.  

In many cases, they remove so many steps of a process that people no longer have to think about how they use it. You definitely do not need to be a programmer to use the Internet, mobile apps or even now, design your own neural networks.

The democratising effect

Giving wider access to new innovations brings a much broader, more diverse group into the fray. In practice, what this means is the ability for people with different skill-sets to use emerging technology and potentially drive progress even further.

This is the vision of Velas, a company steeped in ideals of equal use and free data. Users can download neural network kits and run them on their own systems with a visual interface just how they would use a standard computer application.

What of quantum computing? 

Though incredible to consider just how much quantum technology is set to change the game, it may not be that far off. Earlier this year IBM unveiled their first commercially available – not to own, just for user calculations – quantum computer.

Such ultra-fast computers can train neural networks to an ideal state in next to no time at all compared to the speed of current systems. It is hard to imagine how neural nets will proliferate as a result of this, becoming ever more diverse and adapted to their individual use cases.

Neural network marketplace

The upshot of all this? An online marketplace for neural nets where users can earn monetary rewards for their contributions and buy systems from others. When applied to networks based on Velas, there is room for continued improvement to the blockchain network algorithm.

This means that users can be rewarded for how their neural networks evolve and at the same time contribute to the continual improvement of the blockchain itself. This is a process that includes each person who chooses to stake and become a block creator. In fact, every DPoS node makes an intellectual contribution to the improvement of a neural network.

For the first time, people have the chance to develop neural networks in a remarkably straightforward way. Potential implementations are limitless and only limited by the creativity of users. Being freely available to the world means that these networks can be used for everything from AI-driven music matching algorithms to natural language understanding to data analysis.  

Possibilities are endless and at the same time, these neural nets never need to stop improving upon themselves. This is an open future of artificial intelligence and machine learning which can drive a process that brings the whole world along for the ride.