I would feel very unnerved if I walked into a metro or underground and it was plastered with posters saying “Stop hiring humans” (You know the ones).
I’m happy with the overall response I saw to that stunt. The CEO decided to go into truly a lot of very different subreddits to talk about it, and got some very direct responses from people. He even tried to do AMAs in a few of them, which all went as well as you’d expect.
Marketing stunt or not, I’m not here to engage with the premise of whether actively trying to replace people with AI for all jobs is ok to do or not (I really like that most knowledge in society is inside of humans, I don’t want it to change currently). I just want to know whether the thing that's responding to me in real time is or isn't human. Whether I’m typing, talking, or on video with it.
The EU AI Act agrees we want this, however you want to count that towards the cause. The US has a lot of different things bouncing around their legal system as well.
Companies that intentionally try to obscure this reality should be punished. How can we know they know how to control their AI? Even model maker companies don’t know too well.
I don’t want to briefly mention an interest in the Bible, Torah, or Kuran, only for that to change my user experience as the app pivots its personality (A more technical take on this).
Pair that with the frankly shocking number of individuals who believe the zeros and ones coming out of their screen are sentient, and the individuals are delegating increasingly large pieces of their dwindling sanity to them.
And all this before video has really taken off at all… It’s still just text and audio… And vision is our dominant sense. I’m a little worried. I use AI a lot… but it’s the most supernormal stimulus yet, isn’t it?
We’re no longer mining entropy, the entropy is mining us. I don’t think AI is sentient, but it is dangerous as a root access feedback mechanism for individuals who don’t have cognitive defenses to realize we haven’t bottled God into the silicone yet, but the silicone does have an exceptional vocabulary and knows a lot about a lot.
And you better know the difference…
(1) I very much agree that it's a good idea to have AI systems always distinguish themselves from human ones either through a watermark or just a regulated requirement for a preamble (2) I think it would be a very positive change for AI systems to be RLHF'd away from pretending to care about people and express emotions, and the only reason corps aren't doing it is profilt and (3) While the chat interface is incredibly useful because it mirrors the interface affordances of a command prompt where I re-evaluate print loop and so I don't think it should go away. It is unfortunate that most people's experience with That sort of interface is messaging actual human beings and not command lines, so they will interpret more as the former than the latter.
However, there are a few points that I think you could edit to briefly address that would make your case a lot better:
That first link, that claims that something like 40% of people think AI is sentient, is a screenshot of a Google Doc posted on a random subreddit. That doesn't seem like a very good source to me. And it, in fact, seems to be... based on completely hallucinated sources that the author of the Google Doc refuses to actually provide but provides screenshots of a chat GPT conversation and a blog post for.
Additionally, the model collapse paper works really artificially hard to achieve that model collapse whereas modern large language models are in a very substantial portion trained on synthetic data generated by earlier models.