AI and Blockchain

AI and Blockchain

Hyped technical buzzword plus another hyped technical buzzword in the same project is a secret sauce for any get-rich-fast scheme. This is why we talk about Artificial Intelligence (which is an over-marketed word for statistics) and blockchain (which is the over-marketed word for databases) going together.

Generally, AI and blockchain is a great combination of technologies because of two shared qualities. A lot of people know that some other people made lots of money from these technologies. At the same time very little people know how do they work and how exactly do you use them to join the ranks of that lucky minority. Thrilling, isn’t it?

The mystery aside, let’s go through some trends where the best application of blockchain and machine learning are today and speculate about the most promising startups to drive the market tomorrow.

Crypto-trading AI

Traditionally exchange trading is the most lucrative place for people who are good at seeing patterns or designing algorithms that see patterns. An algorithm that detects anything predictable in the price oscillations and transforms these predictions into corrective orders is just as good as a money printing press. The cryptocurrency market is much smaller than traditional financial or commodities markets. However, its entry barriers are much lower as well. Almost every modern exchange has a well-documented API and publicly available historical data. Most exchanges are open to trade for citizens from all over the world. The market is rapidly evolving with lots of price shifts. Therefore the crypto market is a great place for aspiring data analysis enthusiasts to field-test their first setup from a trading bot and a homebrewed machine learning algorithm.

Anonymous cryptocurrencies vs fraud-detecting AI.

Seasoned machine learning experts are widely recruited by the cryptocurrency exchanges and government agencies. The government agencies are developing AI to keep track of the anonymous cryptocurrencies. The exchanges are developing AI to reluctantly detect fraud transactions and actively keep watch on the pumps and dumps from the rival exchanges.

It is a tug of war. Anonymous cryptocurrencies are evolving to outmatch AI that tracks them. Government agencies upgrading AI and finding other smart ways to peek into transactions.

Distributed zero-knowledge machine learning.

There is a bunch of cases where there is a need to gather sensitive data to get statistics from it. However, because of the data sensitivity, the owners of the data are quite reluctant to give access to it. The best example of this would be medical data. You don’t want anyone except your doctor to have access to the information about your health problems. On the other side, you would want to have a computer diagnostic system that would show your most probable health risks. The solution to this problem is a distributed zero-knowledge machine-learning platform that would encrypt all the incoming data to everybody but the machine learning algorithm inside. The output of this distributed black box is an algorithm that makes accurate prediction while the information used to train this algorithm remains inaccessible. The process happens in a distributed environment so every party of the project has the opportunity to check the code and ensure there is no data leaks.