AI’s not getting any smarter, so how can accountants get more value?

AI has reached an intelligence plateau. 

“There’s no ‘second internet’ waiting to be discovered”, explains Diana Wolf Torres, meaning that AI has reached saturation in terms of new data it can be trained on. The internet is a finite resource, and it’s pretty much been used up.

This has interesting implications for us, the users. If we can’t expect AI to keep scaling in performance, we need to find smarter ways to work with what the platforms offer today. That’s how we’ll make gains.

We talked about how at the last AI Exchange (a meet-up of finance and accounting leaders who are adopting AI in their firms). The good news is that there are plenty of capabilities yet to be explored and the barrier to entry is low. We’re going to see more of a firm’s everyday tools connecting to large language models (LLMs) with new “agent mode” features baked in, unlocking further opportunities. And, if you’re clever with prompts and connections, firms can get a lot of value from AI with just a small time investment.