AI And VC

 There's a long piece in the Times about Uber's problem with sexual assault and sexual misconduct. My wife sent it to me, because I was dismayed at how easily she came to getting comfortable in a Waymo on her last trip to California. To me, Waymo represents a terrifying venue for tech companies like Uber to fire their drivers, many of whom depend on driving for Uber or Lyft. Her point is that there are a lot of cases of Uber drivers engaging in everything from creepy behavior to rape. The Times article points out - and the Uber spokesperson argues - that statistically, the number of these incidents is small, but the absolute number of events is large. That is to say, the percentage of Uber rides that end in sexual misconduct is low, but it still happens a lot. What's more important is that the number really needs to be zero.

However, the argument that Waymo will save women from sexual assault or misconduct seems to miss the problem, which is that Uber has tools to dramatically reduce this problem, and they have often avoided implementing them. One tool is to record every trip, but if they do that, then they are admitting that their drivers are employees, and that has been a bright red line for them. They do not want to be responsible for providing health insurance or other benefits for what they term "independent contractors." 

They also had a plan to allow female riders to select only female drivers, but they shelved that when Trump was elected, in case you want to know where the heads are of major corporations today.

There is the problem of sexual assault, which predates Uber obviously. Then there is the problem of how Uber was run - it was a classic tech bro startup under Travis Kalanick. They wanted to grow, grow, grow, and efforts to preserve passenger safety were downplayed in order to grow the company and prepare for the all-important IPO.

That dynamic in some way, shape or form will be present as Waymo scales up. What's more, it will be present as AI scales up. We may already be seeing it.

Here is a chart of browsing searches for OpenAI. It collapses when school gets out. For all the hype and hoopla about what AI can do, it fundamentally is being used - at best - as a shortcut around the rigors of learning. The idea that AI will be able to cure cancer is appealing (especially since RFK is killing cancer research). The reality is that it's a cheating machine, and it is in the best interests of the people running those companies to insure that no restrictions occur when using their machine until it gets profitable.

The logic of Silicon Valley Venture Capitalism is an aggressive form of capitalism that is dedicated to rapidly scaling up a product, issuing an IPO and then cashing out. You can go from rags to riches pretty damned quick, but eventually you get to enshittification and poor customer experiences.

Waymo will cut corners if it means quick profitability. Not the BIG corners, because they have to show that they are viable. The little corners, the one that causes a problem once every 2000 rides. These are the same incentives that Uber faced, and Waymo is not special.

As for AI, it's painful as an educator and a parent to realize that what we think of as learning, the way we have learned for thousands of years, is going to be altered to suit the venture capitalists who want to leverage AI into billion dollar paydays. They could create safeguards - digital watermarks - that lets a teacher know if a student has used AI to generate their answers, but they won't, because they want to crush their IPO.

AI Utopianism is just rhetoric covering for AI Venture Capitalism.