ApicalDendrite

I interviewed there a couple of years ago. I was really excited going into the interview because I'm a big fan of Atul Gawande and I thought there was huge potential for a transformative impact. But when I talked to them it was clear that they didn't have a clear vision of how specifically they were going to transform healthcare. I got the sense that the engineers were just messing around waiting for the healthcare and product people to tell them what to build. They clearly had a lot of talent and resources and I'm sad they weren't able to make it work.

I've worked with engineers at Amazon up to the senior principal level. Even if you grant that people in cybersecurity see a broader swath of the company, there's just no way that he can talk authoritatively about everything from AWS sales tactics, to hiring practices for former DoD procurement people, to real estate strategy to Amazon's own supply chain, to Jeff B.'s personal life. It would be like a mid-level employee at State Department talking about how the Trump administration views farm policy, and what the National Park Service is planning, and what types of cool things the US Mint is cooking up. That State Department guy might have heard about a lot of those things either from reading the news or from talking to buddies, but he wouldn't come within a mile of it as part of his job.

To me it feels like someone was kind of bullshitting with their friends and maybe making themselves seem more knowledgeable and important than they really are. It's totally irresponsible to put it in print.

This person is very clearly mixing things that he knows directly, things that he's read about in books (e.g. Amazon's strategy in the early days), and rumors that he's heard (e.g. Jeff Bezos' personal life). There's no way that a cybersecurity engineer at a company as sprawling as Amazon has reliable firsthand knowledge on so many different unrelated topics.

The writer is a faculty associate at the Berkman Klein Center at Harvard Law, which is an organization I have tremendous respect for, so it's really a shame that they were naive enough to publish this.

Were there any concerns about GPT-3's latency? It looks like it takes a long time for online use cases.

A friend discovered that he had a pocketknife with him just before he was about to go through security. So he went out to the airport smoker's area, dug a little hole in the ground, and buried the pocketknife. When he returned from his trip he went back and dug it up.

Nothing I saw indicated that the Echo Show was in any way influenced by Nucleus.

I suspect, but don't know directly, that this investment was more about trying to jumpstart an ecosystem of third-party Echo devices rather than hedging bets.

I worked on the Echo Show team. The product had been in development for over a year when we invested in Nucleus. I remember thinking it was very strange that Amazon was investing in that company when we were building such a similar product internally.

Fitbit users are probably more representative than Apple or Garmin users. Apple Watches are a lot more expensive, and Garmin markets more towards athletes.