AI is about to hit the trough of disillusionment.

We’ve been here before. The dot-com bubble. Social media “revolutions.” Crypto and web3. Same playbook:

• Hype attracts capital.
• Capital attracts consultants.
• Consultants attract grifters.
• Then the floor falls out.

AI isn’t special, it’s just next in line.

Signs are everywhere:
• PE firms bragging about “AI initiatives” that are nothing more than fancy chatbots stapled onto legacy processes.
• Management teams who can’t explain the ROI of their AI pilots beyond “efficiency” and “future-proofing.”
• Entire “AI advisory” practices springing up to cash the cheque before the music stops.

Most AI in the market today is theatre. It looks good in board packs and conference keynotes but does nothing for EBITDA.

And when the hype cycle snaps, the bill will come due. Hundreds of millions written off as “strategic investments.” Decks quietly shelved. CEOs moving on.

But don’t mistake the trough for the end. The 90% of noise will die, but the 10% of real, working AI products will transform business models in ways we’ve not seen since the internet.

Think:
• Call centers run entirely on AI agents, halving cost per call.
• Healthcare diagnostics with 10x the throughput of human clinicians.
• Manufacturing predictive maintenance slashing downtime and inventory waste.

The trick isn’t to avoid AI, it’s to avoid AI theatre.

Investors and operators need to ask one simple question before writing the cheque:
Show me how this makes or saves $1 today. Not in 2030. Not in theory. Today.

Otherwise, you’re just paying for the next shiny PowerPoint.

Read more