Is AI a help or hindrance for consultants?
Initially consulting was dependent on the consultant, building a relationship, doing good work and providing good outputs. In recent times consulting companies like Deloitte have gotten in trouble with the Australian Government by providing almost completely AI generated consulting documents with false information and references. The Australian government paid half a million dollars for a study and it seems they cheated and used AI.
Now for where consulting is getting harder.
I have noticed clients now using more AI and are using it to hold us to account. Now while I do believe that we should responsibly deliver a good service, sometimes we also need to balance doing detailed documentation over delivering the outcome. For many of my clients we are trying to make it work with in their budget and make them successful. We do not charge an arm and a leg and sometime the reports are a bit more to the point and less beautified because we want to give them the meat in the sandwich that moves them forward rather than add 20% more time to beautification and fluffyness.
We are conservative on our use of AI as I am sure that Deloitte have put some government secrets into the LLM and we are aware of AI dreams telling lies. So our use of AI is purely internal and nothing that the customer has to interact with. The clients are now checking our work with AI which now means they are asking for more detailed time reports and lengthy explanations of things and using lengthy documents that they have generated with AI to communicate with us.
So their lengthy AI content plus their requirement for us to use more lengthy content means that we either need to increase our use of AI and risk the AI dreams or charge more for a human to do it. Essentially this drives up the cost of consulting and means that our AI is talking to their AI and they don’t really absorb the true meaning and value of the depth and the complexity of the information we are providing.
An example of this is asking a medical consultant how to lose weight and they will go into detail about diet and exercise and how you should eat certain foods and the do a certain amount of exercise. They can also go into more detail about that persons particular body type or modified exercises to compensate for injury as an example. There is a depth of information that is critical to understand because one good mean or one work out isn’t going to fix the issue. Asking AI will give a much shorter explanation and summarize it into two dot points being diet and exercise. The critical difference being that the consultant gave something very specific and the AI gave something generic. Although you can add additional prompts, it is often the why and the empathetic conversation that the provisional is going to have with the client, for example, “don’t you want to get fit so you can spend more time playing with your kids?”.
Sometimes the full story is important to convey the importance of doing something. If a client AI summarizes your advice statement they are only going to read the dot points and miss the point of what you are trying to convey. There might be 5 possible fixes but one is more valuable to the client, is easier or is some cases is harder but significantly more impactful. Some courses of action need to be decided upon in relation to other external factors such as currency rates, the government of the day or the competition. So if boutique consultancies are expected to give Deloitte style 500+ page and half million dollar consulting outputs that this prices out the small companies that need the help, even if AI is being used.