Future of the AI in the FinTech industry

Not so long time ago, the world was very different. Main topics were blockchain, altcoins and NFTs. I really thought that blockchain technology would find it's way through "nerdy shell" that is still wrapping it. It turns out, the only use case for the blockchain was to provide the hardest store of value humans ever invented. We are not so far away to realize the similar thing with AI, but I think it is interesting to analyze trends that had lead to the AI bubble. One of those trends is Web3 hype.

At that time, I was working as a Web3 Frontend developer, mainly connecting UI with smart contracts though popular libraries like "ethers.js". The demand for NFT marketplace smart contracts and tokenization of almost everything was at its peak. Everyone involved was making money, from investors all the way to developers, and it seemed like it would last forever. But as the old saying for stocks goes: "The higher they get, the deeper the fall", this time was no different.

I was attending a lot of FinTech conferences as well with coworkers, because it was interesting to hear what industry leaders had to say about this topic. From all corners, people were sharing their views about this new era of Smart Contract Banking, Flash Loans, 180 degress turn in the next few years, and those that are not following will be left behind. This great idea of "Code is the law" was so simple and beautiful that it bought hearts and wallets of VC investors, so it bought mine as well.

Looking back from this point in time, it seems that world has moved on and left those great ideas behind. Nothing even close to this scenario became reality. Founders just could not find problems that required putting expensive and slow blockchain layer in front of users. Users are the ones that are spoiled and expect to get everything fast and free. People usually want to have a support when something bad happens, for example, when they send money to wrong address and want their money back. All these things just were so far away from real world that it is hard to believe that so many people bought this fairytale story.

But it is different this time with AI right, right?

In my opinion, NO. And here is why.

Let's say you are a big bank, like JP Morgan. You have billions of assets under management and every decision has huge impact on the economy. I want to sell you my AI tool that is correct 99.99% of the time. That seems like a large number, so it gives me a lot of confidence to go to the board and pitch how this tool could save you work done by 100 people. Board member steps forward and says fine, lets give it 5 million loan requests from database and see the results. This would mean that with this number of loan requests there are 50,000 of them that are put in the wrong category!

First of all, bank is risk management institution at its core, so there is no room for error, especially in smaller banks. Other than that, it is still required that 100 people go through the same files and double check the results. Not to mention how expensive this operation is, there is another even more alarming issue. Someone could break inside your system by prompting AI and convincing it to share protected data with his aunt or something.

So the issues are very similar with ones in blockchain. It is too expensive and too risky. Two words banks hate the most unfortunately.

Future still looks bright

My time spent coding with AI was really interesting. If you are a senior developer, you can achieve wonders with this time machine. Most importantly, understanding every line of AI generated code is crucial for long term health of the project. Now every individual can start working on the startup that required 5 or 6 people in the past. The teams are getting smaller and more competitive. I think the same thing would happen in banking sector just as it is happening in IT right now. Smaller teams, better results.

But main question is still what has lead to AI bubble in the first place? AI was here long time ago, long before ChatGPT, but there was no cheap power to make impactful breakthrough. With rise of the Bitcoin mining, powerfull hardware was slowly accumulating in the basements of north pole, but with every Bitcoin halving, prizes were smaller and competition was fearce. Economy is really the key here. As always, just follow the money. People that were heavily invested in mining have just switched to powering AI which ensured better returns. Simple as that.

To conclude, I don't think people are going to be replaced with AI, I think it will enlarge the gap between competitive groups and ones that are not so competitive. Seniors developers will grow their salaries as their productivity grows, senior bankers will do the same. In the end, I think AI is going to be great for the economy and if you are willing to invest time and effort to learn using the AI just as you did in school, learn read and write so you can improve your knowledge by reading books. With AI it's the same, it is just a tool to help you solve problems more efficiently.