πŸ€– ⚠️ This is what AI worries are really about ⋆ πŸš— Full self-driving cars are here ⋆ πŸš€ Attempting impossible

It's a perfect summary of why people are worried about advanced AI so much. Which robotaxi to hitch a ride in? Some things are impossible until they're not.

πŸ€– ⚠️ This is what AI worries are really about ⋆ πŸš— Full self-driving cars are here ⋆ πŸš€ Attempting impossible

Here's what the worry about AI is really about. And a reason to do things that are genuine to you, instead of whatever society wants you to do.

At this point, for many people this is interchangeable. We're getting closer to the point when you'll need to rediscover who you are.

Thing 1 - Forecasting the AI Climate: No Winter in Sight

These feel very true at the moment and it's a perfect summary of why people are worried about advanced AI so much.

summer is coming

- Sam Altman (CEO of OpenAI)
I expect we will see continual progress from here to AGI, without any period that feels like a "winter". There are just too many obvious things left to be done.

- Adam D'Angelo (CEO of Quora)

Elon replied:

There will not be a winter for AI, quite the opposite

So the general sentiment is that AI will explode in progress. Which seems plausible.

Few things are harder to predict than the ways in which someone much smarter than you might outsmart you.

That's the reason I worry about AI. Not just the problem itself, but the meta-problem that the ways the problem might play out are inherently hard to predict.

- Paul Graham (co-founder Y Combinator)

Example:

Ten-year-old about to play chess for the first time, skeptical that he'll lose to Magnus Carlsen: Β "Can you explain how he'll defeat me, when we've both got the same pieces, and I move first? Β Will he use some trick for getting all his pawns to the back row to become Queens?"

- Eliezer Yudkowsky (AI researcher)

And that's the perfect illustration of what we don't know. We're building AI in our image, but we want it to be better.

Somehow I'm (for now) less worried about AI getting out of control and destructive on its own.

I'm more worried about someone finding a way to "uncap" a powerful AI and asking it to do something stupid and horrible to see if it's gonna go through with it.

And then not being able to stop it, because it's faster than us and always steps ahead.

Thing 2 - Full self-driving cars are here

You can hitch a ride in a Cruise in San Francisco, and it's cheaper than the driver'd (driverfull?) alternatives.

Waymo website screenshot

Waymo One autonomous ride-hailing service is available to anyone in the Metro Phoenix area, onboarding San Francisco and they are preparing for Los Angeles. They claim to have the largest fully autonomous service area in the world - 180 square miles.

Lyft (partnered with Motional) completed over 100,000 self-driving rides with real people on real roads. They still have a safety person at the wheel afaik.

Baidu (Apollo Go) and Pony.ai (PonyPilot Plus) are the first robotaxi services in China that operate without safety drivers - limited to 23.1 square miles.

Zoox is a purpose-built robotaxi on open public roads with passengers. It's available in San Francisco and Las Vegas.
It doesn't even have manual controls available!

Zoox vehicle website screenshot.

Coming soon: Argo AI (through Lyft), Uber (with Toyota and Volvo)

Did I miss any?

Thing 3 - Attempting impossible

Practice even what seems impossible.

- Marcus Aurelius

If you practice jumping into orbit, there's little doubt you're not getting there.

But will it improve your jumping? No doubt about that either.

But - what if it wasn't impossible?

Some examples that broke down barriers and opened floodgates:

  • Roger Bannister's four-minute mile
  • Heavier-than-air flight
  • Harnessing nuclear energy
  • Breaking the sound barrier in an aircraft
  • Reusable rockets
  • ...

So go for it πŸ™ƒ

Cheers, Zvonimir