Generative AI is Automation

Generative AI (or “A.I. 2.0”, which is all the hype that started with ChatGPT at the end of 2022) exposes a concept that has existed in IT for decades: automation.

Automation is the application of technology for maximal output with minimal human input. Whenever a company requires its employees to spend days copy-pasting content from various sources, you can be sure that automation has not been understood or that the implementation has failed.

Generative AI, at first, gives the perception that automation will become easier, faster, and likely cheaper to integrate.

To get it right, it requires understanding (I am simplifying here...):

- what the current outputs are (e.g., PPTs and sales reports for executives)

- what the processes are that lead to those outputs (e.g., copy-pasting comments and numbers from emails)

- where the data is (e.g., PowerPoints sent by emails)

- what value is brought (if any) on top of the data being manipulated (e.g., sales director comments on the situation)

This is where the "Digital Transformation" process comes into play.

- First, it comes to articulate clearly what the service we want to create is: the use case (e.g., automatic generation of sales reports).

- Second, it comes to ensure the benefits are clear with an ROI that is realistic regarding the investment cost, time saved, and value created (e.g., time saved for reporting is time gained for selling).

- Third, it comes with how the implementation will be processed: introducing a new tool requires change management, which means: change in the current IT system (that includes compatibility, compliance, security assessment, etc.) AND in the way people are used to working (hours of copy-pasting are perceived by some as much more reliable and simple than anything else).

These three components are part of an implementation plan.

Generative AI as standalone components can create multiple layers of value, but if they are not part of a process, their usage will be limited and benefits will be below expectations.

Moreover,  I won't be surprised if projects triggered by the hype of Generative AI end up being delivered... without Generative AI (in the case I gave earlier, a properly set-up CRM is enough)!

Now Generative AI will open the door to a new kind of automation, particularly those which imply analyzing unstructured documents (such as contracts) the same way the Internet in the 2000s exposed the automation of a well-known use-case: e-commerce.

Which is why I present Generative AI to clients as one additional iteration of a perpetual digital transformation process (following mobile internet, smartphone, Internet for desktop computers, desktop computers replacing pen and paper... etc.). 

This way it helps companies to keep "control", not be distanced by the market and more importantly, reduce the risks of a failed and costly implementation!