We are generally interested in projects that spawn across multiple technology trends. We’ve previously written about combining blockchains with green energy, for example. Or take the example of the convergence between blockchain and autonomous vehicles. In our view, these are powerful trends that combine multiple technologies of the future, to gain leverage over how things unfold. Combining multiple such technologies helps accelerate the progression of both.
Today, we want to talk about another such combination – combining blockchain with AI (Artificial Intelligence) and ML (Machine Learning). This is what the project Repux is doing.
All About the Data
To truly understand the Repux project, you first need to know what it is trying to accomplish and how it fits into the existing AI frameworks. At the core of any AI system is data. Lots and lots of data. That’s how the machine learning algorithms ‘learn’ – from the data fed to them. Therefore, data becomes a valuable commodity to these AI systems.
This is why it is so hard, for example, to compete with Facebook. It is not just the traditional ‘network effects’ i.e. you are on Facebook because all your friends are on Facebook. It is even more about the amount of data that Facebook has been collecting to make its experience ‘better’, i.e. more sticky to the user. This is how the next generation of technology and internet companies are being built. Between Facebook and Google, you’re already talking about over $1 trillion of market capitalization, and that’s just scratching the surface of the AI niche.
Repux Token and Data Markets
Now that you understand how central data is to the new and emerging AI revolution, let’s look at what Repux is trying to do.
The biggest problem with AI is that it is a fairly specialized field, and requires a ton of data. Smaller firms therefore are at a considerable disadvantage, because they are small and therefore cannot collect as much data as large enterprises. This provides a huge benefit to the incumbents, since they have more data, and are getting more data every day. Therefore, in order to compete, small but nimble firms require to buy a lot of data and data sets from outside, instead of trying to build it all in-house.
That’s where Repux comes in. Repux is a marketplace for data – firms can go there and buy the data sets that they need to create their machine learning algorithms and train them appropriately. The owners of this data, who may not have the expertise in-house or may not know what to do with the data, or just want to share the data and make money off of it, are all potential data sellers.
It is also possible that the data provider sells data on the Repux platform to data scientists, who then build their machine learning models using that data, and sell the models back to the original data providers. This way, companies that are not usually heavy on the tech side and cannot afford to hire AI experts, who are usually hard to come by, can do one-off machine learning models from these experts based on their data.
Here are some examples of how Repux is being used, for example, from their own site:
Photo Credit: caseorganic