Can AI already replace the digitalization consultant?
At first glance, Artificial Intelligence appears omniscient. But is it really?
What used to be “Googled” is now simply entered into ChatGPT. More and more people rely on AI’s answers – often without questioning how they are generated.
Many have long known: The quality of the results largely depends on the right question – and the right AI.
I personally tested how helpful an AI is in selecting a new digitalization solution. The result? Both sobering – and instructive.
It became clear to me: Without an experienced consultant, it’s (still) not possible.
In the current blog post, I explain why this is the case – and what risks arise if one blindly relies on AI.
AI vs. Consultant in the Project. Is there a winner?
(drawn by ChatGPT)
Example: The question about the number of TMS solutions in the DACH region and the suitable consultant
Anyone who asks ChatGPT how many TMS (Transport Management) solutions exist in our region will be surprised. It refers to the BVL-Matchmaker, stating approximately 100 solutions.
The AI seems unaware that this platform (which, by the way, Trovarit AG and we manage) now lists > 140 solutions.
It is quite interesting, however, that other sources are not even mentioned.
This raises a question: Whether 100 or > 140, which is the right solution? The layman cannot see the wood for the trees. The same applies to finding the right consultant!
The subsequent question of which consultant would be best suited for TMS selection is also puzzling. Because the SCC-Center or Trovarit AG, on whose know-how and data this very platform is based, are not even mentioned.
The result is rather “strange” companies. For example, a manufacturer of bags and folders, a software provider from the SAP environment (certainly completely neutral), a management consultancy as an SAP partner, a consultant for the manufacturing industry, and generally the Big4 management consultancies. In short, total rubbish!
One has to wonder where the “intelligence” has gone, because the answers actually lead one astray rather than helping.
At least here it’s obvious. In other areas, however, everything sounds logical, and one relies on the answer.
With the variety of offerings, “customers often can’t see the wood for the trees”!
(drawn by ChatGPT)
Example: The question about a detailed TMS requirements specification for a tender
A selection project also includes detailed requirements. One might think that sufficient data for the requirements specification, as a basis for an RFQ, is already available on the internet.
The problem: ChatGPT doesn’t even ask what kind of company it is. It makes assumption after assumption and ultimately spits out a document that is at best suitable for an RFI.
The content feigns expertise where there is none. Anyone who blindly takes this information and sends it out will certainly not receive any usable answers.
Therefore, caution is advised here too.
Proper and good requirements specifications are actually only available from a consultant. In our example, we have a detailed requirements specification with almost 2,000 individual criteria, which can be individualized for the company.
By the way: The question about “Use Cases” also backfired considerably. Instead of a use case, there was only a series of questions, but no requirements.
Always be the first to receive the latest news, interviews, and expert articles?
When NI (Natural Intelligence) is still far superior to AI
There is a problem with AI, because AI needs data to learn. This includes data that tells the AI which data is useful for the learning process.
The examples mentioned above clearly show that an “unsupervised” AI, which simply draws from the internet, does more harm than good.
Furthermore, AI lacks one thing: LIVE experience. From projects, from implementation, from discussions, and from very personal analyses.
As long as these are not digitally recorded anywhere, but for example in the consultant’s brain (as in my case), the consultant is always and everywhere superior to AI in precisely these areas!
No, I won’t let AI tap into my brain
(drawn by ChatGPT)
And what are examples of relevant experiences?
The know-how of a long-standing consultant usually represents the decisive added value compared to pure AI support.
These cannot be found in any review portal or database on the internet from which AI can learn.
Here are some examples:
- Relevant contacts
- Knowledge of “good and bad” salespeople and consultants
- Pitfalls of marketing vs. real life – promises vs. implementation
- Quality of presentations, workshops, etc.
- Quality of contracts
- Willingness to negotiate
- Efficiency of the implementation methodology
- Handling of problems
- and much more.
As a “one-eyed man among the blind,” you won’t win a trophy!
(drawn by ChatGPT)
Successful digitalization projects are only possible with the right consultant!
If you want to know how that works and if we are a good fit, feel free to arrange a free initial consultation.
To schedule a free initial consultation, simply request an appointment here via the link. It will certainly be worthwhile!
Conclusion: Brain, heart, and intuition cannot yet be replaced by AI. True to the motto “Beware of the trap,” the consultant is currently still part of the formula for success!
P.S. Perhaps I used the wrong AI. However, ChatGPT could not, or would not, recommend a “better” one for this purpose.
Image sources: ChatGPT
