A real AI advisor makes you more capable of deciding for yourself; a repackager makes you dependent on them. Everything else — the jargon, the demos, the deck — is noise.
- "AI advisor" is one of the fastest-growing roles in consulting, and a good number of the people wearing the badge met ChatGPT last month. Telling the real ones apart is now a core leadership skill.
- Trust is the currency of interactions. The goal is not to find someone you can trust blindly; it is to build enough judgement that you can trust yourself.
- A real advisor helps you build the bridge between your business and the technology. A repackager sells you a feature set still hunting for a use case.
- You cannot know everything. You can learn to vet anyone.
Here is how you tell a real AI advisor from someone who met ChatGPT last month: the real one makes you more capable of deciding for yourself, and the repackager makes you more dependent on them. Everything else, the jargon, the demos, the deck, is noise. Rachel Botsman (a trust researcher at Oxford's Saïd Business School) puts it well: money is the currency of transactions; trust is the currency of interactions. And trust is exactly the thing that is hardest to assess when you do not yet speak the language.
Most people are good and doing their best. The Telex salesman in the 1980s (Telex being the typed message network that businesses used before the fax machine) was not trying to mislead anyone; he simply had not noticed the fax had already replaced his product. Most "AI experts" are the same: sincere, and selling the last paradigm. Your job is not to catch liars. It is to develop the judgement that tells you who has actually kept pace.
Why is "who do I trust" the wrong first question?
Because it outsources the one thing you cannot hand to anyone else: your own judgement. Henry Ford (the founder of Ford Motor Company) is said to have remarked that if he had asked customers what they wanted, they would have asked for faster horses, and the experts would have dutifully bred better horses. Ask the wrong expert the wrong question and you get an excellent answer to a question that no longer matters. The better question is this: am I informed enough to know what the real problem is, and to guide my team as they solve it?
Trusting advice blindly is expensive. I once ran a diagnostic on a company paying an agency AUD $100,000 a month for digital advertising, and only AUD $50,000 was reaching actual ad spend. The rest was skimmed. That is fraud, it happens, and it only happens to people who never learned how to check.
What separates a real advisor from a repackager?
| A real AI advisor | A ChatGPT repackager |
|---|---|
| Starts with your business problem | Starts with the tool's feature list |
| Explains it in plain language | Hides behind acronyms to seem indispensable |
| Names what AI can't do for you | Claims AI can do everything |
| Makes you more able to decide alone | Makes you dependent on them |
| Shows evidence and a way to measure it | Shows a demo and a vibe |
| Has a track record before 2023 | Discovered the category last quarter |
The tell is in the language. Experts who have invested years in a domain often make it sound inaccessible to validate the effort; when you hear a wall of jargon, you assume they must be right. A real advisor does the opposite: they translate. They turn "your lambda sensor has affected backflow combustion" into "the part that mixes air and fuel has failed, and here is how we test it."
Want a second opinion you can trust?
The AI Strategy Session gives you the principles to vet any advisor, and a ninety-day plan you own.
Book your Strategy SessionHow do you vet an AI advisor in practice?
- Make them define the problem before the solution. If they reach for the tool before they understand your business, that is your answer.
- Ask them to explain it to a non-technical colleague. If it does not land in plain language, they either cannot simplify it or do not understand it themselves.
- Ask what AI will not fix here. Anyone who says "everything" is selling, not advising.
- Ask how you will measure success, and by when. Real advisors tie themselves to an outcome; repackagers tie themselves to a retainer.
- Check the pre-2023 track record. The principles of transformation predate the latest model. Judgement is older than the hype.
This matters because the cost of getting it wrong is climbing. More than 80% of AI projects fail, roughly twice the failure rate of IT projects that do not involve AI, and the most common cause is not the technology: it is people misunderstanding or miscommunicating the problem the project was meant to solve (RAND, 2024). The earliest version of that mistake is usually trusting the wrong person about what to build.
A real AI advisor makes you more capable of deciding for yourself. A repackager makes you more dependent on them. That is the whole test.
What does trusting yourself actually take?
It does not mean knowing everything; no one can. It means being able to vet the experts: judging the proposed solution as carefully as you judge where your industry is heading. You still go to experts for advice. You simply stop taking the deck at face value, and start asking the questions that reveal whether someone has caught up to the current paradigm or is still selling the previous one. Once you can do that, the wall of jargon stops being intimidating, and the room full of "experts" starts earning its keep.
Frequently asked questions
Are AI consultants worth it, or are they all repackaging ChatGPT?
How do I tell a real AI advisor from someone who just discovered ChatGPT?
Should I just trust my chosen advisor?
About the author
British technology futurist, AI keynote speaker and advisor. Thirty years across enterprise technology and AI strategy, helping leaders navigate the future of work. The futurist who died.