We closed the selling process of our house in Massachusetts last week. It was not easy, considering we were in Illinois all this time. Anyway, we were able to successfully execute the process. Like most other life experiences, there were aspects that I observed during the process that are relevant to the world of technology.

Early on, back in April this year, I decided to try selling my house through Opendoor Technologies Inc. Suppose you are wondering why is a technology company buying houses. In that case, Opendoor technologies use an “AI” algorithm, like Redfin and Zillow used to (before they closed those business divisions), for buying properties. So I decided to try them.

Their final offer for me was almost six figures less than what I eventually ended up selling my property for. Not to mention the fact that I sold my property during the off-season, and Opendoor’s offer was before we discovered a significant issue with the septic system of the property. The weird thing was that they also deducted 10% off their offer for the property to account for “costs”, despite no real-state agents involved. Obviously, I did not sell them the house, but I was intrigued. It is evident that the “AI” algorithm was broken. So, how was the company surviving?

So, I did my research. If you look at the operating income of the company, it has consistently been negative. Looking at the income statement summary, I see they keep getting an influx of non-operating income, though. Wonder where that comes from since I did not have time to review the detailed reports.

Source: AlphaQuery.com

Anyways, it was obvious from their operating income results why they were offering those crazy numbers. The algorithm they have is definitely flawed. But my curiosity did not stop there, so I went to online forums to see what people who have actually sold to them have to say.

That research made it apparent that almost every other home they bought, they offered more than a typical conventional real estate sales process would yield. Then, they would list these houses with a markup. Because the prices were above the market value, the property would sit for months and then eventually be sold for a loss.

And at the core of this very flawed business model is an “AI” algorithm. But we do not have a regulating process yet to question these algorithms. Not even in the scenario where the underlying business operations pivots around that algorithm.

If it is obvious to me, then it is very much evident to executives at these companies that the algorithm is flawed. Yet, consistent business decisions are being made, by a public company based on an “AI” algorithm. And Opendoor is not the only company like this. We are seeing a flood of “AI” enabled tech companies, with little or no control or validation of algorithms, flooding the market.

While masked behind the fancy label of “AI,” why is this practice not as shady as the one that led to the housing market collapse? If you allow unscrupulous “AI” enabled algorithms across industries, with no challenges posed even when there are tell-tell signs of failures, we are just sitting on a mountain of dynamite, waiting for it to explode? Investors and shareholders will eventually suffer as a result.

I saw an AI-related executive order this morning. While watermarks in deep fakes are essential too, there is a definite need to put together teams within entities like the SEC, that can validate the core algorithms of public companies whose entire business models pivot around one specific “AI” enabled technology. I did not see that within the context of that executive order. Hopefully, that will be covered later.

We are always in reactionary mode, enforcing regulations after something disastrous has already happened. Even the SEC was formed after the Wall Street crash of 1929. The AI crash can also be disastrous, considering that the penetration of “AI-enabled” products and services is now across industries. Nations that continue to be leaders remain so because of foresight. Undoubtedly, AI algorithms will eventually run a significant portion of businesses worldwide. And that makes validation and regulation of these AI algorithms critical.