Move over Digital Transformation, here comes AI Transformation

While we as individuals are enthralled with what ChatGPT and the rapid progress it is making in actually being useful, there are discussions and brainstorming sessions happening at organizations as to what AI means for them. There is an underlying FOMO (Fear Of Missing Out) that is creeping up even in company board rooms. They all know that while they did survive (some thrived) by digitalizing their businesses, they also know they were late to the game and should feel lucky they were not completely disrupted. No company wants to be late to the gate this time.

This is good and kudos to the businesses who have started looking into how AI can be useful to their business growth. Most businesses are still struggling to figure out how and where to get started. But that is fine. The only thing for company executive management and board members to ensure is that the discussion and brainstorming is not a knee jerk reaction to what is happening and is a genuine attempt to define strategy and find utility in AI.

I will state now that the title of this note is misleading and was meant to draw attention rather than suggest digital transformation is dead or useless at this point. On the contrary, the path to AI transformation goes through digital transformation. Many businesses are still in the process of executing their multi-year digital transformation strategy. They should continue with that as it is going to be the foundation for their AI transformation. At the same time, they should start to lay out a high-level strategy for the use of AI.

There are 3 ways that businesses can start thinking about how they will use AI:

1. Use AI to optimize their internal operations and make themselves more efficient and competitive.
2. Use AI to develop their products and services better and faster - think time to market a finished offering.
3. Deliver AI-enabled products and services to their customers.

I intentionally listed the 3 in that order because I recommend that businesses start with the first item first. While we need to move fast, we should note that we are still in the early stages of understanding the full breadth of what AI can do for us. There are also regulatory discussions around safety, privacy and other important matters that will ultimately shape what we can do and what we cannot. In that regard there are inherent high risks to item 3 and some risks to item 2 as well.

So, what can we do with AI to make our internal operations more efficient? It may be a bit anticlimactic to see the examples below except for the first one:

Writing - this is currently the best use case as it is already proven to be effective. We all know that getting started with writing is one of the hardest things to do whether it is for work or for personal use. Many of us find it easier to edit something already written than to start writing something from scratch. Plus, we have now already tried out the writing abilities of AI LLMs (Language Learning Models) and started feeling comfortable with it. Just think about all the writing we do at work. I am not just talking about emails. Think presentations, proposals, statements of work, contracts, reports, and possibly many more.

When I asked ChatGPT how much time could be saved, it was a bit non-committal though. Here is the response:

"Using ChatGPT can potentially save a significant amount of time in writing a 10-page report. Typically, writing a report of that length can take several hours, depending on the complexity of the topic and the writing skills of the individual. With ChatGPT, you can expedite the writing process by generating content and getting ideas faster. However, it's important to note that the extent of time saved will depend on various factors, such as how well you can articulate your thoughts and how much editing and revision the generated content requires."

Note that I am still honing my "prompt engineering" skills. So, I am sure someone else with a better prompt will get a better answer.

Process Optimization - when I bring this up, the usual response is "we are already looking at our processes and optimizing them" or "we have already reviewed our processes and eliminated inefficiencies". And my response is that the tables have somewhat turned.

Not that we want AI to tell us how we should run our business … or maybe we do. With appropriate training, AI can find inefficiencies and point them out to us. Think about a department head tasked with making their operations more efficient. Most of us know that we have biases that come in the way of doing the right things. That is why we bring in outside consultants to find inefficiencies for us. But that causes friction. We can use AI to identify inefficiencies and reduce friction between human beings.

Sales - a sales system with an AI engine can make sales processes more efficient and deliver better outcomes. Think about account assignment (which salesperson should be assigned which account), when to call, what to say at which meeting, how to price a product or a service, when to ask for signature on a deal, when to give discounts, etc. And then all the reverse actions, when not to …. We all know things like timing, affinity and several other intangibles and incalculable things are critical for sales success. AI can help with that.

There are domain specific AI models being built at this time. Very soon, AI models will be available for any domain, function or industry we can think of. Of course, it will be upon us to validate that these AI models are not fluffy and actually deliver outcomes. Most of our work will really be that: selecting the right AI. But remember, humans can also "learn" and we will be able to figure out which tools/models are spurious and don't work and which do. Thanks to human intelligence.