For those who want the cliff notes.

Here's his poorly though out arguments

• Evolution took a long time to make brains so teams of programmers must take a long time to make brains. (Right because those two are totally equivalent processes)
• Humans are so special and magical we can't possible be emergent complexity arising from a simple rule set.
• Inventing a new chip could take 30 years (Ai invents viable chip designs in hours)
• We still need new ideas for AGI who knows how long that will take. (Not true btw, when available computation exists there's always more than enough ideas to take near maximum advantage of that computation, and available computation grows at a predictable exponential pace. Every 20 years we have more than a million times more computation available.).

But I understand the frustration of any Ai programmer that's been in the industry for multiple decades. They are attempting an impossible problem, trying to write AGI with insufficient compute, so of course they are going to get discouraged.

Compute power is all you need to look at to make predictions. The smartest people all over the world will squeeze nearly everything possible out of that compute as it arrives.

Similar people to these geniuses also predicted Airplanes will take millions of years to develop back in 1903...

https://bigthink.com/pessimists-archive/air-space-flight-impossible/

These people are always vastly overconfident for no reason. Maybe if we had to design a computer exactly the same as a brain it would take a hundred years, but we don't, we can use vastly more power than the brain to match its computational power, we can use far larger volumes of computational area since we aren't restricted to a skull and birth canal.

We made airplanes by simply throwing more power and human engineering at the problem, I suspect we'll make AI similarly and if flight was similar AGI will happen within 2 decades, far faster than even 2 average lifespans of hunter gatherers let alone LEV humans...