AI - Hype Hope and/or Harm

Summers in Kalaburagi were brutal. Then came the first rain, and that unforgettable smell, which has a name: petrichor.

It turns out rain is the best way I’ve found to explain what happens inside AI.

Does AI just memorize the internet and look up the answer? In this episode of AI: Hype, Hope & Harm, we test that claim. You’ll see what an AI model keeps and what it throws away, why it can answer questions nobody has asked before, and why it sometimes makes things up with total confidence.

Don’t fear AI. Don’t worship it. Understand it.

What is AI's IQ? I went looking for a number. I found a pattern, and it led somewhere bigger.

In this episode: • Why IQ was designed for humans, and what happens when the test-taker isn't one • Four tests, four kinds of intelligence: can it learn, can it reason, can it do research-grade math, can it find things? • "Is that cheating?" Why tools change the score, and why real life allows them • Why "What is AI's IQ?" was never the right question • Basavanna, the Sharanas and the vachanas: what happens when a doorway widens • Why knowledge is abundant but wisdom is still ours to bring • Is this the age of abundant intelligence? Half of it.

One server rack the size of a refrigerator could soon draw as much power as 65 homes. So who pays to keep AI turning?

In Episode 3 of AI: Hype, Hope & Harm, I take on a two-part claim you've probably heard: "AI's hunger for power is out of control" and "the race is too big to stop." To explain it, I take you to a restaurant in India, where a potter and his wheel turned out to be the best explanation of AI I've ever seen.

IN THIS EPISODE
- Why "the cloud" is really steel, copper, concrete and electricity
- The AI flywheel: better AI → more use → more data centers → more power → better AI
- Germany today, Japan in 2030: data centers' growing electricity use
- Why more efficient AI still means more electricity (the Jevons Paradox)
- The two "transformer problems" AI has
- One AI request, followed all the way down to your neighborhood grid, and who ends up paying
- Hype, Hope and Harm: my verdict on the first half of the claim
- Arms-race logic: why nobody wants to be first to stop, from companies to countries
- The pot is what we make with it

Is Artificial Intelligence actually intelligent?

In this episode of AI: Hype, Hope & Harm, I explore what we really mean by intelligence — and whether today’s AI actually qualifies.

From AlphaGo’s famous Move 37 to the latest ARC-AGI-3 results, we look at learning, reasoning, adaptation, and why machine intelligence may not need to look like human intelligence at all.

The bigger question may not be whether AI thinks like us — but what happens when a different kind of intelligence begins operating alongside us.

Don’t fear AI. Don’t worship it. Understand it. And as always, stay curious, my friends.

In this episode of AI: Hype, Hope & Harm, I explore what AI is really changing: tasks, roles, skills, and the way we work.

The better question isn’t “Will AI take my job?” It’s: How do I become more valuable because AI exists?

Dont fear AI, Dont Worship it, Understand it, As always stay curious my friends