Einstein said, “The same causes produce the same effects,” but with AI, that’s not the case.
What’s fascinating about AI is that this statement is false.
Try asking CHATGPT the same question at different times, and you’ll see that you’ll never get the same answer.
For example: “What is an AI agent?”
This raises questions about how the response to a prompt is extracted and generated, according to clearly variable and random methods, without a strictly defined process, with the answer being updated after each query from other users.
Furthermore, in the context of accounting or financial processing, one might wonder whether this will impact the quality, or even the reliability, of the information produced, given that the processing of accounting and financial information can vary for the same prompt.
One might think that with financial data extractions, this possibility is reduced but not impossible, hence the importance of closely monitoring the results of AI processing.
As illustrated by the case of Starbucks, where the AI agent was unable to handle simple inventory management, one can only imagine what the production of financial statements would look like…
So, AI is not yet a panacea.
