I guess my autism is getting the best of me. I just listened to the Andreessen interview, and I have a lot of thoughts. Incoming wall of text.

First of all, always want to say that I love @Malice as an interviewer, I truly think he is one of the best because he is actually open-minded but also is smart and has a good BS detector.

I feel like Andreessen is basically just acting as a salesman for AI proliferation. And that's fine if that's what he wants to be. I am firmly on the AI-skeptic side, so you can understand where I am coming from.

Not to overly nitpick, but right off the bad I feel he is throwing shade as those who would anthropomorphize AI, this need for us to "see ourselves in things that aren't human." Ok fine, but his descriptions seem to directly conflict with that sentiment. AI as a mirror of humanity, and a representation of everything people have ever thought or said. These are his descriptions, yet he's also implying it's over-simplifying things to humanize AI. When I run a program on my PC, I don't talk to it. I run commands. Why not use the same verbiage with ChatGPT or others? The designers themselves are humanizing the tech. And also it seems this is what the users want, so I won't just blame the tech overlords here. But to me it seems that treating AI as pseudo-human or at least pseudo-conscience seems appropriate.

As far as the comment on brainstorming, that's a fair point but it also glosses over a huge giant problem with these tools. There are different classifications of errors. When Thomas Edison is trying hundreds of different materials to get the lightbulb to work, those failed attempts are errors in one sense. When a lawyer submits legal documents that contain entirely made up cases, that is also an error in a sense. The problem is we already have people right now who are not using ChatGPT for brainstorming but are essentially taking it's results as "the answer." It is true that this is not a new problem, humans have taken short cuts for a long time. If the first google result agrees with you, in your mind that's proof you're right. That said, I do strongly believe that LLMs are making this human tendency much worse.

And that leads me right into Marc's point about shifting moral panics. This answer is very poor IMO. "There was a moral panic about TV, and then about social media, and now about AI." Ok, and? Again back to the Thomas Sowell quote. Yes it's not as if these technologies are all 100% bad, like just cracking yourself in the head with a hammer. But TV obviously has had and continues to have other negative unintended consequences. We all read less and also read less challenging books. Watching TV is far, far more passive. And IMO social media (especially doom scrolling) is this ramping up to a much more extreme degree. Scrolling is literally channel surfing and has obliterated the attention span and focus of large numbers of people. "We're afraid that people were going to become couch potatoes?" Does Marc not engage with the general public? I regularly see small kids at the grocery store sitting in the cart or at restaurants and scrolling endlessly. Almost zero physical motion observed other than that thumb moving. Does that mean we should all throw out TVs and phones in the trash? Probably not. We can exercise some self-control. But think about how quickly TikTok spread. And now ChatGTP use. This is not that different to a new drug being used far and wide before we even have a halfway clear idea of what the negative side effects are. This is not right-to-try at end of life. If I had to guess, there are probably more people using ChatGPT and the like than not now. People are encouraged to dive in, no questions asked, and just go to town. My suggestion is that we assume this will be addictive, and also have other harmful side effects (this will almost certainly be the case) and proceed with caution. Why exactly are we diving into this at max speed, spending money on data centers like there's no tomorrow? Spare me the China fear. Most of our elites want to "beat" China by emulating them. They hate that the average American normie does not want to live in a Chinese-style controlled society.

Also just to circle back for a minute and attack my own position, most of problems I see with AI proliferation are not new. They are simply more extreme versions of existing problems. I remember in high school in the early 2000s the sentiment still existed "you need to be able to do mental math, because you won't always have a calculator with you." This statement is obviously absurd now. That said, I would still say that the ability to do math mentally is a good skill to foster because it helps you think better. I have no idea how often you should bust out the calculator app vs just doing it yourself. How often should you challenge yourself? It's really up to each of us individually. Can't remember the year an album came out? Maybe try to think about it and work it out yourself for 15 seconds before asking Siri the question. Or don't. Each specific thing is not critical. It's not life or death that you can do 13 x 13 in your head, but fostering mental skills is. The only "skill" people are practicing now is learning how to prompt AI more efficiently. And guess what? By some metrics, AI is already better at prompting itself than you are, so where does that leave us?

Marc's point about AI doomers wanting their own version of totalitarianism, ok I'm sure some of these people exist. But also this point just seems insane. Switch to a different topic, vaccines. It would be like asking "ok you anti-vaxxer, so what do you want a cop in every doctor's office making sure this so-called 'poison' is not administered?" This is just an utterly ridiculous point on it's face, because the currently reality of the situation is SO FAR on the other side. Commercials, schools, doctors, hospitals, government agencies, they all push vaccines without question. And this is now, not even taking into account the extreme pressure during 2021-2023 roughly. The idea that anti-vax totalitarianism or anti-AI totalitarianism is anywhere even remotely close to a real threat is insane, and the fact that he's bringing this up is also insane. He's trying to do the "I'm a moderate centrist in the middle of these two extremes" type bit.

On the jobs point, I don't know, maybe I am too skeptical, I can allow that possibility. But his points just sound like commie gobbledygook. An Uber driver is going to transform into the owner and operator of a tour guide company because they were using ChatGPT? This sounds exactly like when people say so-and-so is going to write the great American novel as soon as we get UBI so he doesn't have to clean toilets anymore. I feel confident saying this is way too optimistic. Some people have drive and some don't. Not everyone is cut out to be an entrepreneur, I know I'm not. I am very skeptical that a chat bot is going to unleash this latent ability in large numbers of people that aren't already hustling out there. From what I see, the main thing you need to actually operate any kind of company successfully is unrelenting drive. That's not something you can just google. The vast VAST majority of people I personally see using ChatGPT RIGHT NOW are using it to essentially be lazy with their writing or thinking. "Write my resume for me." "Write this text for me." "Summarize this." I'm sorry, these people are simply going to get much dumber with prolonged AI use, and I really hope I'm wrong on that.

I'm also kind of a retard, so maybe I'm wrong on some of this. But I do have a bad feeling about it all. And thank you to the one person who actually read all this.