Twitter and Elon Musk Part 1

Elon Musk and I have similar interests. Musk is interested in space travel; I wrote Att NASA; Move Over, Artemis; CTMT, a New Way Into Space. Musk is interested in AI; I wrote one of the early implementation  of the computer language Prolog, which by the standards of the day was an AI tool. Now Musk is interested in social media. I wrote Social Virtual Reality, An Internet Paradigm for Change, available for download from academia.edu. It’s about an evolved web, and has lots of citations. We share knowledge of terms like “regular expression, “software stack”, and “rpc”.

So although I am not an entrepreneur, I may have some insight into Musk’s thinking about Twitter. This is not about Musk’s management style, which is a separate issue, except to quote Winston Churchill: “When you have to kill a man, it costs nothing to be polite.” Otherwise, you accrue more than the minimal number of enemies.

If  one views a software enterprise as nothing more than a deterministic code base running on reasonably reliable hardware, it gives rise to the idea that it ought to be able to run  itself. All system failures are then due to poorly written code.  This view has undeniable attraction to someone who has just spent $44B for an enterprise that is losing $3M per day. It gives rise to this logic:

  • In a world of perfect computer code, you could just fire everybody who was not directly connected with improving the code, and turn a profit.
  • This is not a world of  perfect computer code, but surely it can be improved, requiring far fewer employees to do so.
  • Those functions of Twitter that deal with human mayhem external to the code base, currently handled by expensive human beings, are to be offloaded to artificial intelligence.

The last point is encouraged by Musk’s success with the Tesla self-driving autopilot, which uses compute structures inspired by the brain, artificial neurons, to solve a problem beyond the reach of conventional computers. Does this predict Musk’s success in a battle of wits with malevolent humans?

Maybe, maybe not. The argument has flip-flopped several times since 1950. It  now hinges on the role quantum mechanics in defining the human mind. It will be a challenge to present. If you’re strictly liberal arts, don’t feel left out. Ludwig Wittgenstein’s Philosophical Investigations will give you a feel for the problem. It’s downloadable from the Internet Archive.

Let’s take a little break, and let that sink in.