The Compound Power of AI & Blockchain in Finance

Forbes has compiled a ‘Blockchain 50’ list of the biggest companies in the world that have already adopted blockchain technology.¹

The list includes, for example, Jack Ma’s ANT Financial who recently pulled back from the biggest IPO of all time, among others. Frankly, it’s no surprise that many of the listed companies are in finance. While it is a somewhat conservative industry that has been slow to adopt innovation, financial firms have long been crying out for improvements in efficiency.

The fact is that prior technology hadn’t promised the kind of solutions that the market required. Now, however, given the increase in regulations, changing market attitudes, and the desire to reach the ‘mass affluent’ market, finance is finally ready.

Soon, we will see Artificial Intelligence solutions and products developed on top of proven blockchain technology that will increase the capability and performance of machine learning to create new financial products by an order of magnitude.

In this article, we’ll explore how. But first, let’s quickly cover what AI and Blockchain are.

Artificial Intelligence (AI) is the ambition of replicating a cognitive mechanism or action that would normally be performed by a human brain, but with a computer or software instead.

You can think of face unlock in your smartphone, the YouTube recommendation algorithm, and GPS rerouting when traffic gets bad as common examples.

AI is nothing new to humans. The thought of artificially intelligent machines was first coined by humans in the 19th century, but was formally found in 1956 at a conference at Dartmouth College². Since then, what has changed? Back then AI was a sci-fi story that existed only in our imagination. Today, due to heavy computation power and the availability of tremendous amounts of data to train systems, it’s very much real.

To learn effectively, AI requires large datasets. And with the rise of Big Data this is no longer an issue. In fact, worldwide data is expected to hit 175 zettabytes by 2025, representing a 61% CAGR³. The rise of AI alone is noteworthy with its potential to change our daily lives from how we work and how we play to how we travel and how we make decisions. So just imagine its combined potential, compounded, with blockchain.

To the uninitiated, blockchain is synonymous with bitcoin and cryptocurrencies. But in the same way the internet was known mostly for pirated movies and music, blockchain is much, much more than public consciousness realises.

A blockchain is a digital record of transactions. The name comes from its structure, in which individual records, called blocks, are linked in a chain. It is, essentially, a public ledger shared and agreed on by a majority of users on a distributed network.

Blocks verify transactions in a consensus. To modify a transaction on the chain requires agreement between all users. That's why blockchain is virtually immutable or unhackable. To fraudulently enter or alter a single transaction would require ownership of a majority of a network that would immediately lose its value.

Blockchain technology opens up the possibility for instant, low-fee, automated proof of transactions globally. Consider how slow and expensive it can be to change currency or send money abroad. On blockchain, this can be done instantly, at scale.

Along with a promising future blockchain technology has the potential to provide a better alternative for every financial platform in existence.

In Ancient Chinese philosophy, yin and yang is a concept of dualism, describing how seemingly opposite or contrary forces — good and evil, for example — may actually be complementary.

The same idea is true of blockchain and AI. While blockchain itself is secure, the same cannot be said for applications built on top of blockchain. In that sense, security is a drawback — one that AI helps solve. Symbiotically, where AI faces issues around trust & authenticity, blockchain comes to the rescue. Together, their power is greater than the sum of their parts. Together, they create machine learning and store of transactions that is secure, private, and capable of changing how the world manages governance, and security and trust in finance — at scale.

Machine learning algorithms work with data or information where it learns to make better decisions based on data. AI works more efficiently and the machine learning outcome is more trustworthy if this data comes from a reliable, secure, credible source. Blockchain data are stored with high integrity and resiliency, and cannot be tampered with. When smart contracts are used for machine learning algorithms to make decisions and perform analytics, the outcome of these decisions can be trusted and undisputed.

The following use cases explain how AI and blockchain are a natural fit and how they could help financial services:

1. No more Facebook-like scandals with data

As mentioned earlier, the availability of large datasets for research purposes was one of the driving forces behind AI. There are, however, two problems with data. First, getting the data is a big concern. If you want to get sufficient data to train your model to make precise decisions, you need lots of data. Simply, you need to partner up with Google, Apple or Facebook.

Secondly: privacy. Given the recent use of public data by Facebook, users might be hesitant about sharing their data. Blockchain can step in now and offer transparency regarding what data is being accessed, when is it accessed, and for what purpose. Blockchain has the power to democratise ownership of user data in a way that lets users decide who they share with and how.

This paves the way for Distributed Digital Identity: a platform that gives users complete ownership of their private data. DDID comes as a package of several benefits, one of which includes allowing users to only disclose necessary information without excess. For example, if you go to a bar you need to prove you’re above 18, it is not strictly necessary to prove your name other than in that it helps prove your age. DDID allows you to do exactly that — prove your age beyond doubt without revealing any other information. DDID is something we’ll explain further in the next article.

2. Cut out the middleman

Blockchain lets individuals make smart contracts and monetise their own data. This kicks the middleman out of the picture and makes the process secure. Zero-Knowledge Proof (ZKP) can be leveraged by businesses to search for information within the same cluster or circle without knowing the specific details of information or the identity of data ownership without permission. Imagine, for example, a polling company wants to know which Facebook pages Biden fans like. ZKP over blockchain would allow the company to see that the data exists, but not what exactly the data is not who it belongs to — without the user’s permission.

3. Make money in your spare time

Blockchain-based cloud can leverage more distributed computing power for AI through a decentralised market for computing power. This opens up opportunities for users with huge computational power and AI developers.

End users like gamers could lend their spare GPU to AI developers and get paid. Developers in turn can use it to train Machine Learning models. An idea similar to this called grid computing already existed and was popular. However, limited applications reduced its mass appeal.

4. Improved Trust on Robotic Decisions

Despite the fact that machine learning has seen a lot of success lately in building systems that are autonomous and capable of making their own decisions, people are reluctant to adopt these systems due to a lack of understanding about what exactly happens in the hidden layers.

The fact that our information is handled by a computer with intelligence to rival the human brain is terrifying. Furthermore, when people like Elon Musk make public statements about AI decisions being ‘concerning,’ it encourages further distrust and fear of AI⁴. Blockchain can help reduce the fear by showing exactly what data is used in every decision. When we can see how decisions are made, we’re more likely to trust them.

To fix this, we can improve the machine learning models by decentralizing AI on blockchain. This lets users run machine learning models on everyday devices and applications⁵.