By Pete Major, Vice President, FinTech Services


During the last six months, there’s been a lot of buzz around AI which was spurred by the release of ChatGPT. Yet AI has been around for a while; it’s been a common marketing term used to describe complex if/then programming statements for at least the last 10 years. How many of us have faced the virtual AI bot where you end up screaming at the screen, “REPRESENTATIVE!”? There definitely have been some limitations to using this technology.

I had some prior experience working with a blog service that had AI functionality built-in to generate blog posts based on some input prompts (like a landmark that you’d like to feature in your blog). The writing was decent and much faster than anything I could’ve written but it would put “facts” into the article that couldn’t be cited. And so, the Pinocchio-ness of the service made its usefulness limited.

Like some of you, I had skepticism about ChatGPT when it was released. But within a few minutes of chatting with ChatGPT, I was blown away. While it can still make up “facts” like the other AI service, it understood language in a much better way than anything I’ve seen in the past. I’ve given it tasks from generating sample travel itineraries to Japan to coming up with the top loan origination software providers for credit unions. While the results weren’t perfect, they were great starting points and certainly saved a ton of time.

AI isn’t just relegated to an improved search engine, creating blog copy, or completing research. It has been successfully used to create programming code, generate artwork, or create voiceovers. A friend used old voice recordings of me to generate new recordings of my voice reading text. Needless to say, it was weird and creepy to hear me reading something I didn’t read. Deep fakes no longer just apply to the rich and famous. Bad actors can use AI too.

MDT has been researching AI for several years. We had IBM present their Watson service to our executive team and while we didn’t think the technology was right for our clients at the time, we’ve continued to keep a pulse on the burgeoning industry. We discussed some guardrails at the MDT E3 Conference that you may consider as you investigate AI: ensure your privacy policies are updated to reflect employees using AI services and keep compliance involved, particularly with AI loan decision engines.

Whether you are excited about AI, scared about it, or both, AI is here to stay so it’s important to be educated about it. Take the first step and create a ChatGPT account or use a similar service and ask it questions (just don’t share confidential information or take the results as gospel).

At MDT, we’ll continue to research AI and the vendors that support the technology and keep you abreast of new products/services that may be useful to your credit union.