0:07-1:19

Dan Cooper introduces himself, Acumen, and his guest, Roper DeGarmo. Their topic for the day: using AI in business!

1:20-5:19

Roper begins to speaks to his experience integrating AI into daily execution and processes within his company, Signature Personal Insurance, a nearly 20-year-old independent insurance brokerage. He shares that from early on he thought his industry would be deeply disrupted by machine learning, but that in reality it moved in fits and starts into the retail space and only recently became high-quality and useful. Roper wound up a bit late to the game after all! He started working with chatbots as an option for after-hours customer service, but quickly found that was not a good solution for that role. However, he found that ChatGPT gave an immediate near-perfect answer to the question, "What's the difference between reconstruction cost and market value?" and was hooked. He highly recommends getting the paid version of ChatGPT.

5:20-7:53

Roper explains that initially he was most impressed by ChatGTP's deep domain knowledge, and its ability to contextualize answers for different contexts. In terms of using it strategically in business, Roper would still like to use AI for complex client self-service, but it's not there. Currently he is researching and seeking to understand AI as a tool in a variety of different settings.

7:54-9:51

Dan asks for more detail about Roper's hopes and plans for using AI in the future. Roper asserts that AI cannot do your job for you: it will never outperform the expert. The trick is to find the things it does well or differently, to fill the gaps in ways that make things easier. For example, Roper will throw whole email chains into ChatGPT and ask for summaries, details he may have missed, etc. In the future, he hopes AI could proactively remind users of next steps or forgotten communication.

9:52-13:54

Roper shares that it's been incredible to watch his team become comfortable using AI. For example, an accountant who is not proficient is English uses it to level up her delivery of invoices to clients! He also speaks to how he earned a certificate on machine learning through MIT and used AI to help break down concepts and reinforce learning. AI applications are complex, often involving multiple domains, and mapping it all out took work, which was supported deeply by instantaneous AI information and confirmation of understanding.

13:55-16:57

Dan asks: where should a newbie start? Roper cautions against trying to use AI first in urgent, critical processes and tasks at work, and advises to instead start applying it to a fun topic or hobby, so that you can follow it down past your own knowledge and expertise at leisure and see what results without pressure. Choose a domain that you care and know about, but that isn't mission-critical! In terms of the best tools, he shares, GPT version 4 has been the top model, but Clod AI is better with large documents like contracts. Gemini was recently rolled out by Google, and it is quick and easy for restructuring and reformatting in particular.

16:58-18:33

Dan raises a common concern: security. Roper compares that concern to the fears during the initial launch of data being stored in the Cloud, and asserts that the fear of the unknown is probably less pressing than more mundane dangers like crafting more secure passwords.

18:34-21:12

Roper marvels at how unknowable the future is when it comes to AI. Change continues to happen so quickly and everything is moving so fast, and it doesn't do what we expect. There are still blue collar jobs, in spite of earlier fears that those jobs would all go to mechanical arms; and we didn't think AI could be creative, but with no fear of being wrong, it's wildly useful there! It's hard to predict what will happen next. He does wonder if any role focused on providing expert answers will move to AI, and whether the next step - the creation of workflow and process - will also move that direction eventually.

21:13-23:07

Roper advises listeners to just take some time exploring the incredibly broad and ever-changing realm of AI, which will affect every industry at some point soon. How will it affect yours? How can you start understanding and using it right now?

23:08-25:05

Dan asks how Roper is being sharpened, challenged, and inspired these days, and Roper shares about his experience using AI to create situation briefings for his Acumen Council. This led him to find out more about using AI for problem-solving in various ways and through new lenses. It was a good, sharpening experience, as was his recent certification experience.

25:06-26:14

Dan closes out asking for Roper's contact information.

Website: signaturepersonalinsurance.com

Dan shares more information about learning more about Acumen at acumenimpact.com.