Gen-AI and the business user

The last few months have been all about AI. To be particular, Generative AI. As the name suggests, the Gen part is about generating new content - text, dialogue, code, pictures, music, what have you.

The technology is remarkable. If you have not tried it, I suggest you do. The 'aha' factor is huge.

As with anything this fundamental, there has been a bunch of discussion about how AI is going to create and destroy jobs, change the economy and so on.

Based on personal experience, I'm offering a hypothesis as to where Gen-AI is going to have the maximum impact on the average business user's life. It is, in two words, Analysis/Synthesis.

The biggest impact of GenAI in business will be in Analysis/Synthesis

Most firms I have seen have increasingly been adding jobs called 'analyst' - essentially a catch-all to describe people who work with systems, processes and data to help the business operators make decisions and manage work. This has pervaded all functions and sub-functions of the organization - every business leader needs his/her analysts to support their work. This is a natural by product of 3 forces:

1. More data - with system enablement / automation of more business processes. Data is video, audio, pictures, text AND numbers.
2. More decisions - Every firm wants to ape startups, epitomized by Amazon, in making more decisions, about more things, more frequently. Business reviews used to be monthly affairs - they're moving to a weekly cadence. Each review must result in some decisions - I'm yet to see a review where people say, "everything's ok, wait and watch, then".
3. Data based decisions - A natural corollary of 1 & 2 - decision makers want decisions to be data driven. More and more, gut feel feels like a dirty word.

Firms want to make more decisions, about more things, more frequently

As a result, there is mountains of effort going into converting mountains of data into metrics, metrics into trends, all of it into reports, getting business folks to get it and finally take decisions. A lot of this effort is going to go away.

The unique feature of gen-AI LLMs seems to be the ability to work with unstructured data to collate, clean, interpret, tabulate, compare and most of all summarize on demand. This is 90% of the effort of your typical 'analyst'. I am not talking about sophisticated modeling which remains a niche. Just plain and simple analyst grunt work.

Bottomline, while there are many other applications of gen-AI in coding, creative content creation etc. that seem to hog the limelight relatively more currently, in my mind the killer feature is analysis/synthesis, which will have the biggest immediate impact on your average business of any scale.

p.s.

1. Fully aware this is kind of tunnel vision based on what I do today :)
2. My recent experience relates to gen-AI applications in reporting on sales conversations and business process SOPs
3. I have, like many others I'm sure, played with chatGPT for a few days and gotten bored
4. Garden variety search is a more consumer oriented use case for analysis/synthesis of unstrutured data