AI's High-Stakes Gamble On Wall Street

Welcome to this week’s Deep-fried Dive with Fry Guy! In these long-form articles, Fry Guy conducts in-depth analyses of cutting-edge artificial intelligence (AI) developments and developers. Today, Fry Guy dives into how AI is being used to predict the stock market. We hope you enjoy!

*Notice: We do not receive any monetary compensation from the people and projects we feature in the Sunday Deep-Fried Dives with Fry Guy. We explore these projects and developers solely to reveal interesting and cutting-edge AI developments and uses.*

AI is being used to predict the future.

It can predict weather far more accurately than your local news station.

It is being used to predict crimes in Argentina.

And it even has the potential to predict if you’ll get divorced from your spouse, with 95% accuracy.

It’s no surprise that AI is being used to predict many future events, from weather to breakups. But AI prediction will shine brightest on Wall Street, as models of all kinds are used to predict daily stock prices. And by doing so, it’s going to make individual investors and hedge funds rich … filthy rich.

Those who understand how AI can be used to predict the market will have insider access to the greatest financial cheat-code of all time. It will be the equivalent of owning Biff’s “Sports Almanac” book from Back to the Future. Let’s explore this reality.

A SCAM OR A SECRET?

Get-rich-quick AI videos and tutorials are littered all over the internet.

• This video claims to have used ChatGPT to make 19,527% profit in the stock market.

• This video claims it can predict the stock market with some simple computer code.

• This video claims its AI trading bot can multiply your money by 6x.

In a world where money comes at a premium (quite literally), many are buying into these “get-rich-quick” promises. People are losing trust in financial markets. Stock volatility is reaching levels we haven’t seen since the Covid-19 pandemic, and people are desperate to protect their investments at all costs. So it’s only natural for people to take a look at these get-rich-quick AI tools. AI has the solution for everything, right? So why not look to it for financial help? The problem is, in most cases, people who use these tools end up losing a huge chunk of their portfolios.

Despite the massive amount of AI money scams that proliferate the web and cause many to lose thousands, some AI stock prediction models—mostly open-source ones—are legitimate and have outperformed stock market experts over time. Models like Long Short Term Memory Networks (LSTMs), for instance, are popular and have a proven track record. LSTM model are designed to learn from years of stock price data. They then use that data to analyze certain trends or patterns to predict future stock prices.

Now, LSTMs are not guaranteed to make you money, but they do have a good record of success. A study from 2014 to 2022 followed the performance of an LSTM model that was used to predict stock prices of companies in China. The study found that the LSTM model had an average yearly gain of 31%. Although this is a far cry from the 19,527% claimed by the previously mentioned YouTube video, a 31% yearly gain is considered phenomenal in the financial world. At the very least, this study shows that LSTMs have potential to work.

In addition to LSTMs, there are other successful predictive AI models. For example, Convolution Neural Network (CNN) models are similar to LSTMs except they are trained on images instead of pure stock price data. CNN models find visual patterns in stock charts and synthesize those patterns into predictions for future stock prices. In many cases, this approach can be effective.

Along with LSTMs and CNNs, there are many other AI financial models like XGBoost or SVMs. But before you get too excited, there is some bad news. If AI prediction models like LSTMs worked 100% of the time, as intended, droves of people would be getting rich right now. But it still seems that the majority of people are living paycheck to paycheck, and the unemployment rate in the U.S. is creeping over 4.3%, which is the highest it’s been since 2021. So what’s the problem?

WHY AREN’T AI MODELS MAKING PEOPLE RICH YET?

If AI models have the potential to predict the stock market, and many models like LSTMs are outperforming industry expectations, why isn’t everyone getting rich off them?

One of the problems is that there are already a lot of people using LSTMs on the stock market, and competing LSTMs do not perform well. Sure, if you used one in a vacuum without competing investors, it might outperform most stock market analysts. But in the real world, there are thousands (if not millions) of LSTMs being used at any given second. Even billion dollar hedge funds have entire teams dedicated to developing AI stock prediction models. So the chance of one individual’s simple model (the one you downloaded from the internet) beating BlackRock’s multi-million dollar stock prediction tool is slim to none. BlackRock’s tool will instead exploit the patterns of your lowly LSTM model, and you’ll end up losing money because of it.

Although AI models like LSTMs tend to struggle in competition with other AI prediction models, there is a way to work around the competition. Instead of trying to use the same AI models that large hedge funds have used for decades—ones that will no longer make money in today’s stock market—there are innovative tools that can infiltrate industry secrets.

One tool that shows potential is called Numerai, which gamifies AI financial prediction models. It lists the best performing AI stock models at any given moment in time. Although you can’t use these models directly, you can take advantage of this platform in two ways.

One way to leverage Numerai is to create your own model—a combination of a LSTM model and CNN model, for example—upload it to Numerai, and hope it climbs to the top of their leaderboard. And if it does, Numerai will pay you directly for being one of the best models on its platform!

The second way to leverage Numerai is to invest directly with Numerai itself. That’s because this platform functions not only as an AI financial model contest, but primarily an AI hedge fund. It actually uses the data from the models in its contests for its hedge fund. So if the top model on its leaderboard made 30% returns in a year, Numerai will most likely use that model in the real world to buy and sell stocks, which could lead to 30% yearly returns for their investors. This sounds great, but once again, there’s another roadblock: Numerai’s minimum investment is $1 million. So if you’re not a trust fund baby or you didn’t sell a tech startup for $50 million dollars, you might be out of luck.

SEARCHING FOR DIAMONDS IN THE ROUGH

If you aren’t a genius computer programmer or have a million dollars lying around, there is still hope for you to leverage AI for market predictions. There is a growing number of open-source AI financial prediction tools on the web that anyone can use—tools that are available even if you’re dead broke.

One of the most popular AI finance tools is FinRobot, which allows anyone to forecast the stock market using AI. It can be downloaded on Github right now. Another great tool is Danelfin, which gives users AI-driven insights for investing. These tools can help users engage with the stock market in an informed way.

If you want an off-beat way to leverage AI for market research, there is a cutting-edge tool called Uptrends, which monitors thousands of online forums and social media sites for keywords to figure out which stocks people are talking about. Using this data, the platform ranks stocks according to their online chatter—both positive and negative—to help inform investment decisions.

AI tools like FinRobot, Danelfin, and Uptrends are just a few of the many useful financial AI resources available to the public. And over time, AI tools like this will probably become more user friendly and easier to profit from due to the competitive landscape and advancements in AI.

For those looking to use AI tools for market research and investing, it’s worth keeping in mind that getting to the punch first matters—users that take advantage of these tools first will most likely be the winners. Because as soon as a successful AI model is released, it’s almost guaranteed a new one will pop up right behind it, expose its weaknesses, and essentially render it useless.

So over the coming years, forget about asking your cousin’s financial advisor for stock advice. Instead, just fire up your laptop, click a few simple buttons, and have an AI bot make millions for you … Well, that’s the theory at least! Just don’t plan on buying that yacht anytime soon, because you could just as easily lose your entire live savings.

*Disclaimer: None of this is financial advice. It’s simply information about how AI is being used and can be used in an attempt analyze and predict market trends. Do your own research or you might become broke and homeless.*