Old habits, new hype

I hate visual clutter. I’ll shove stuff under the bed just so the floor looks clean. Lately the tech press feels like that hidden mess: AI-powered headlines everywhere, none sorted.

Gartner says one in three Gen-AI pilots dies after proof-of-concept. S&P puts it closer to forty percent.

That’s not a hardware problem. It’s a thinking problem.

These AI pilots aren’t run by rookies. They’re led by people who’ve shipped real products. Yet slap “AI” on a project and we skip steps we’d never skip otherwise like user research, problem framing, and proof.

Why? Because AI feels omniscient. If the model knows everything, maybe we can skip the boring parts.

Old discipline slips, shiny-object bias slips in, and the team sprints toward a demo that solves nothing.

Old-school business analysis still works:

• Name the business pain in numbers.

• Ask what counts as success before writing code.

• Check if the data is good enough to get you there.

None of that looks like innovation, but all of it stops expensive mistakes.

Every hype cycle rewrites the same script:

1. New technology promised as salvation.

2. Leaders fear being left behind.

3. Pilots launch without clear value or exit criteria.

4. Failure blamed on the tool, not the thinking.

Mobile did it. Blockchain did it. Gen-AI is doing it now.

Different buzzwords, same shortcut.

I used to be cynical watching teams repeat the pattern. Now I feel calm. Because the fundamentals haven’t changed. Fast code doesn’t fix fuzzy thinking.

So while everyone else chases the next model size, I urge you to keep asking:

• What are we really fixing?

• How will we measure it?

• What else could we try first?

Not keynote material, but it keeps budgets intact.

If you want a deeper dive into the habits, I’ve tucked a short guide here.

Otherwise, that’s it. No pitch. No CTA. Just the reminder that old habits still win.

Until next time,
Pragati

Share