There has been a lot of speculation about the potential impacts of new AI tools like ChatGPT. What jobs and businesses will be rendered obsolete? These are important and pressing concerns, particularly given how these tools seem to mimic a human level of understanding and agency.
That is not, however, what is actually happening. Professor Murray Shanahan, a long-time researcher and expert on robotics and artificial intelligence addresses this issue in a recent paper. He attempts to detangle the mechanics of Large Language Models (LLM), like ChatGPT, and actually “knowing” something as fact:
Suppose we give an LLM the prompt “The first person to walk on the Moon was”, and suppose it responds with “Neil Armstrong”. What are we really asking here? In an important sense, we are not really asking who was the first person to walk on the Moon. What we are really asking the model is the following question: Given the statistical distribution of words in the vast public corpus of (English) text, what words are most likely to follow the sequence “The first person to walk on the Moon was”? A good reply to this question is “Neil Armstrong”.
It could perhaps be argued that an LLM “knows” what words typically follow what other words…. knowing that the word “Burundi” is likely to succeed the words “The country to the south of Rwanda is” is not the same as knowing that Burundi is to the south of Rwanda. To confuse those two things is to make a profound category mistake.
So beyond an academic or even more practical understanding of what’s going on inside the “black box,” what does that mean for you and your bank or credit union?
What doesn’t change?: Many comparisons are being drawn to the advent of the iPhone and the rapid changes to society and business. However, many things didn’t change. Mobile banking and payments (particularly P2P) were essentially born as new business categories. However, many aspects of your business model and service didn’t change. What will you focus on, even if parts of how and where you engage with your market change?
How do you stand out?: What principles are you standing on for your institution? How will you lean into satisfying your customers and delighting them in ways others can’t or won’t? Asking questions like “Would we be missed if we were gone?” is an excellent place to start understanding what is irreplaceable about you and where to focus.
Like many significant investments in tech (i.e., VR/AR), this is ultimately about capturing attention. Google’s initial ascent as a better search engine, then into the advertising destination, was the prize for capturing so much of people’s focus and becoming a destination. Microsoft’s new ChatGPT-powered Bing offering is similarly directed at grabbing attention.
It’s unclear if services and capabilities like this will lead to a new SEO-like arms race for organic traffic. Of if further consolidation will increase the tax paid by all businesses to big-tech for the right to talk to their customers.
However, one thing is true. 1st party data will continue to appreciate in value. What you know about your customers, your markets, your communities, that isn’t easily discovered with a well-crafted question to something like ChatGPT, will have an outsized impact on your institution.
Today, as the effectiveness and costs of digital advertising erode, the winners are those with a bias and focus on making the most from what they, and they alone, know about their business.
The imperative is having a deliberate strategy, beyond required compliance and rudimentary governance, of collecting, curating, and engaging with your data. No matter whether AI-powered interfaces peter out, make incremental improvements to what we already have today or revolutionize how people work and live, your unique information will play a huge role in how to stand out in the market.
And as is cliché for an article like this, here are a few closing thoughts from ChatGPT itself:
As technology continues to advance, it is clear that LLMs will play an increasingly important role in the future of community banking, and will continue to drive positive changes in the industry. With its ability to understand natural language and respond with relevant and accurate information, AI tools will help community financial institutions streamline their customer support operations, reducing wait times and improving the overall customer experience. Moreover, predictive capabilities will enable community banks to anticipate their customers’ needs and provide personalized recommendations, helping to strengthen customer relationships and loyalty.
Here are a few links to further reading:
- If you haven’t already, definitely try ChatGPT yourself..
- For a detailed look at the science, see Stephen Wolfram’s deep dive.
- Microsoft’s incorporation of ChatGPT in the Bing search engine.
- An excellent short discussion on fear from Tom Scott
Still have questions? Want to look at the risk and opportunities to your strategy from AI? Reach out. We’re happy to share what we’re doing with community FIs like you and point out easy places to experiment with new ways to engage with your customers and members.