An hour into a conversation with @karaswisher two years ago, @elonmusk told a story I still think about whenever I hear CEOs talk.

Two years earlier, Elon said he was "going to take apart a tunnel-boring machine" and "improve its efficiency between 500 and 1,000 percent."

In response, an unnamed tunneling consultant was quoted as saying, "Give me a break. You think someone can take apart a Boeing plane and put it back together, improving it by 500 percent? Elon’s got a very steep learning curve."

Which is exactly what Elon thought too. At the time, he said, "We have no idea what we’re doing. We’re going to get this machine, take it apart, figure out how to make it go much faster and still be safe. We’ll see how much progress we can make."

Two years later, Elon was telling the story to Kara:

"What I discovered in talking to the engineering heads of the various tunneling companies is that they had no idea what the true potential of tunneling was."

"The first question I’d ask was, 'Is your tunneling machine power-limited or thermally-limited?' That’s a very obvious question from a physics standpoint. Nobody knew. None of them could answer that question."

"The answer is everything’s power-limited within the framework of how much you can transport heat away from the face using some sort of liquid cooling system. But we’re crazy far away from the thermal limit. Increasing power is relatively easy, then you hit the threshold..."

"...where you’ve added so much power you’re melting the drill head. Then you have to put a lot of effort into cooling the drill head in order to not melt it – or go with advanced ceramics. But then you still have to cool the bearings and the bearing housings."

How to solve that?

"I’m used to a lot of this stuff from rocket engine design. The turbopump on the Merlin engine runs at 36,000 RPM and it’s got 10,000 horsepower. We’re essentially taking rocket technology and automotive technology and applying it to drilling technology."

"We need to massively crank up the power to the drill head, and then we need to cool it. But the first step is simply jacking up the power like crazy, and then automating the placement of the tunnel reinforcements."

Then Elon did two things other companies weren't doing. Care to guess?

"Anyway, massively cranking up the power of the drill, then making the drill battery-powered is not something the heads of tunneling companies ever considered."

"If you don’t make it battery-powered, then you’ve got to have this massive cable, and the cable has to be an 11,000-volt cable. Then you’ve got to have step-down transformers for the drill head, because the drill head’s going to operate at 480 volts or thereabouts."

Now the last problem:

"Then this is what blew my mind. The standard practice is to use a diesel locomotive to transport the tunneling segments to the drill head and to transport the dirt out. So, if you’ve got a diesel locomotive in a tunnel, you’ve got to blow clean air in."

At this point Kara casually asks Elon, "How else would you get the dirt out?"

Elon's response? "We just use an electric car. We took a Model 3 chassis and converted that into a Model 3-powered train chassis that’s not consuming oxygen and spewing noxious fumes." Problem solved.

Now every time I hear a CEO talk, I think of the question, "Is your tunneling machine power-limited or thermally-limited?"

I don't know if Elon achieved his goal; I simply love how he used rocket science, a battery, and a Model 3 to rethink how to dig a tunnel.