Ep. 353 Crossover: Adam Haman on Elon Musk: Hero or Villain?
Adam rejoins Bob to discuss SpaceX’s recent accomplishment, and how libertarians should assess Elon Musk.
Mentioned in the Episode and Other Links of Interest:
- The YouTube version of this interview.
- Video on economic viability of Martian colony. Bill Burr on Buzz Aldrin. The movie Capricorn One.
- BMS episode with Owen Benjamin (questioning the moon landing).
- Adam Haman substack and podcast.
- Help support the Bob Murphy Show.
I think Bob has revealed where he comes down on the question with his Elon photo selection.
shhhh
The comments here are so stupid.
Did nobody see the errors bob made?
Why …. am I the only one
You must be the light.
Haman is right about the LLMS. They just mimic the form of an answer, which makes them pretty good at art and music and subjective work; I use them all the time as something like a creative assistant for digital art. But they’re unreliable for objective answers. They make up nonsense answers: books that don’t exist, hyperlinks that don’t go anywhere, simple math and logic errors, nonsense code that doesn’t compile.
I also think the projections for the growth of AI are overblown. In a lot of ways, the LLMs are only a modest improvement over the SmarterChild chatbot from the late 90’s, which was already pretty convincing. It took the industry 30 years to make this next leap, so we shouldn’t assume even linear progress – let alone exponential progress towards some kind of event horizon. Personally, I think development’s going to hit a wall based on fundamental limitations of the LLM’s.
But I don’t want to downplay the impact. I’m sure it will be transformative as clever people find ways to leverage the existing developments in various applications for decades to come. So was the internet, but let’s not forget that just as there were people like Krugman who underestimated the impact of the internet and look silly in hindsight, there were also people who got swept away in the hype, fueling the dotcom bubble. An innovation can be real and overhyped at the same time.