Goodbye Mediocre Expert: How AI Redefines Expertise

Goodbye to the Mediocre Expert

A few weeks before I founded ReplyFabric, I had some tax questions. Like many founders, I was afraid of making mistakes during incorporation. My question was about the valuation of software and whether I could use that as part of the startup capital.

Way too early, of course. You don’t need to think about tax optimization if there’s no revenue. But I hadn't actually thought about that.

So I contacted my accountants’ office, since they also have tax specialists. We exchanged a few emails and then set up a meeting. Two people joined: a tax specialist and an accounting specialist. I knew their rates—partners billed at €170/hour, experts at €160/hour. I thought: fine, if I get real value, it’s worth it.

---

My Wake-Up Invoice

Yesterday, I received the invoice: €2,250 ex VAT. About 14 hours billed, split between the two “experts.”

Not only is that a painful number for an early-stage startup, but it also came with zero added value. I hold a Master in Applied Economics, so the basics are crystal clear to me. What I need from an expert is not what’s in textbooks, but the edge cases, the specialties, the premium knowledge that justifies the word expert.

The so-called advice I got? Standard, generic, and 100% findable online. It took me three emails to explain what I needed, and then almost half of the meeting to make my point. And then they understood. An answer? No. “We’ll need to look this up.”

I asked for three simulations: worst, normal, and best-case scenarios linked to revenue. No follow-up. No clarity. No next steps.

Stupid me — I should have known better.

---

Say Goodbye to the Average Expert

Before AI, average was good enough. If you knew the rules, the frameworks, the standards—you could charge €170/hour and people would pay.

Since AI, that world is gone. The basics? We ask ChatGPT. From an expert, we expect more: speed, precision, context, and insight. I’d happily pay €250/hour for someone who instantly understands my question and delivers sharp, actionable answers.

But I ended up paying €170/hour for two people who seemed to have switched off their brains long ago.

It’s a new era. The bar for expertise has risen, and most people don’t even realize it yet.

From now on, this will be my standard question before I engage with any expert:
“Are you using AI?”

Because if the answer is no, I can’t work with them.
I don’t want average. I want champions.

---

The AI Shift: From Juniors to Champions

This shift isn’t just anecdotal. Accordance, an AI startup featured in Forbes, showed how AI compresses career trajectories. As cofounder David Yue explained: “You give [the AI] to a junior, they become an expert. You give it to an expert, they become a champion.” AI flattens the knowledge gap. Students at universities like USF now start from day one with AI tools in their tax research classes. The new normal: accelerated expertise.

---

What Research Tells Us About High-Skill Roles

The European Journal of Management Studies recently published the study The emerging, non-routine European skills of the future by Evan Shellshear (The University of Queensland, Australia), and Kah Wee Oh (Ubidy, Australia). Three effects stand out:

• The Self-Automating Effect: Technical industries may automate themselves out of jobs by coding away their own skill needs.
• The Matthew Effect in Skills: High-skill jobs require more skills than low-skill jobs, but AI narrows the gap by automating routine parts.
• The Red Queen Effect: Professionals in high-skill roles must keep running just to stay in place. New, non-routine skills—especially in communication, management, and creativity—are added constantly.

In short: the basics are gone, and even the top of the pyramid must reskill relentlessly.

---

Lesson Learned

My €2,250 invoice was painful, but it taught me a bigger truth:

Average is over.
AI has raised the standard for what it means to be an expert.

Goodbye to the mediocre expert. The future belongs to those who can deliver true, human premium value on top of what AI already knows.