Monday, October 17, 2022

Proving the proven part

Let's say I claim to have invented an entirely new kind of artillery which can it take out protected targets by firing a projectile up in the air and then letting gravity bring it down. If you challenge my claim and we are talking about the part where the projectile goes up, the burden of proof is on me. If we are talking about the part where the projectile comes down the burden of proof is on you. I don't have to offer any evidence whatsoever to show that projectiles propelled into the air at less than escape velocity will return to earth.

Admittedly, that's a really contrived example but the basic principle does come up, often involving important questions.

Let's take masks and covid. If you accept that aerosol transmission is a significant factor in the spread of a disease, then I don't need to offer any proof that properly used face masks will reduce that transmission. It doesn't matter if there are no clinical or even observational studies to show this. Given aerosol transmission, it is almost impossible to come up with a viable scenario where wearing surgical masks will not help to some degree.

In general, productive arguments always follow the burden of proof. This applies as much to those making claims as to those challenging them.

Let's say you applied for a government contract to build a "hyperloop" (Musk's hyperloop wasn't a maglev train. It was also so unworkable that all "hyperloop" companies quietly scrapped the concept but kept the name.) Your proposal said that you could design and construct this system and build a network of underground tubes for roughly the same cost required to lay a comparable series of railroad tracks. The first part of the claim is your being able to design and build a working maglev vactrain. The second part is that you can build it at an extremely low cost per mile.

No one questions that, given sufficient time and money, you could accomplish the first part. Almost every independent expert is highly skeptical about the second part being possible. And yet, as far as I can tell (and I have been following the story closely), every highly publicized demonstration we've seen of the "hyperloop" has been focused on the first claim with nothing whatsoever focused on the second.

Which takes us to the big unveiling of Optimus.

For a while now, Elon Musk has been promising an imminent breakthrough in robotics that would revolutionize the world economy and human life as we know it.( if anyone out there thinks I am exaggerating the grandiosity of these claims, you really should get to know Musk a little better.)

This is what he delivered.

Musk promised that he was on the verge of mass producing humanoid robots that could readily step in and do a wide range of boring and dangerous jobs. This kind of off the shelf laborer would require enormous advances, solving problems that the best engineering minds in the world have been making only the slowest of incremental progress on. You'll notice that the demonstration doesn't indicate any work on those problems whatsoever.

It just shows a prototype doing things that other robots have been doing for decades. 

Sidenote: Boston Dynamics was sold not too long ago for around one billion. If Musk was serious about getting into the field, this would have been a considerably smarter purchase than Twitter.

For a deeper and far more critical overview, check out this video from long time Musk skeptic, Philip E. Mason,

No comments:

Post a Comment