AI and ROI: An Either/Or Scenario?

I confess: I’m torn.

As the head of Syska Innovations’ newly formed AI Power Bunch, I am staunchly advocating the use of AI to service the needs of our clients in the most comprehensive and efficient way possible. As a business owner, I am less enthusiastic, because achieving an ROI on AI investments is tricky.

This means I’m at a bit of a crossroads: In my technical capacity I’m ready to spend all our cash reserves on AI research. But in my business capacity I’m reminding myself about other technologies that failed to live up to expectations.

ChatGPT springs immediately to mind. Despite all the hoopla that surrounded its official launch, most people I know would say that AI has not changed their job materially, if at all, in the last year.

We are being told chatbots will replace the majority of customer service representation. Then I read a story that a company chatbot returns a wrong answer and the company is on the hook for the incorrect information.

Remember IBM Watson? Fifteen years ago, it swaggered in, declaring, “I’ll diagnose your sniffles better than any intern!” But here’s the plot twist: Doctors shrugged and continued scribbling prescriptions. Why? Because Watson aced Jeopardy, not medical school.

But tech giants are tossing cash at AI like confetti at an Italian wedding. They’re spending billions on infrastructure. Three GPUs that fit in the palm of my hand are more expensive than the new Audi in my driveway. But where’s the ROI hiding? We’re investing like drunken sailors, hoping our GDP will moonwalk into trillions.

Business Me worries that the stock is up to new highs based on the AI market craze. Because I’m old enough to remember the dotcom bust, when companies couldn’t turn investment money into profit.

History has repeated itself numerous times as over-investment in cool tech has led to failures for companies. I remember people talking about how 5G companies were going to transform the world, and those companies were the darlings of the stock market. I remember telling a friend who was asking questions about investing what I thought, and I said that 5G was really cool and transformative, but that it wouldn’t generate a profit quickly enough to justify the valuations of these companies.

Tech Me concedes that Business Me has some valid points. After you unpack the “wow, this is so cool and amazing,” there are still roadblocks to getting anything implemented that is truly beneficial to your company.
But Tech Me is willing to compromise. That’s why Syska is focusing our efforts on RPA and AI tools that yield incremental improvements. In AEC our designs cannot be wrong; we almost have no ability to tolerate a mistake on our projects. And we’re fully auditing and cross-checking the AI used to help generate our designs.

Fortunately, I’m able to balance my enthusiasm for AI with business pragmatism. I firmly believe that AI is the future, but a lot of the talk and rhetoric needs to be supported with common sense and real-world use-cases.

And in case you were wondering, I did not write this blog with ChatGPT.

Written By Robert Ioanna