“We are the only species on Earth that observe "Shark Week". Sharks don't even observe "Shark Week", but we do. For the same reason I can pick this pencil, tell you its name is Steve and go like this (breaks pencil) and part of you dies just a little bit on the inside, because people can connect with anything. We can sympathize with a pencil, we can forgive a shark, and we can give Ben Affleck an academy award for Screenwriting.”
I used to tell my kids "Just pretend to sleep, trick me into thinking you are sleeping, I don't know the difference. Just pretend, lay there with your eyes closed."
I could tell, of course, and they did end up asleep, but I think that is like the Turing test - if you are talking to someone and it's not a person but you can't tell, from your perspective it's a person. Not necessarily from the perspective of the machine, we can only know our own experience so that is the measure.
Alan Watts, talking on the subject of Buddhist vegetarianism, said that even if vegetables and animals both suffer when we eat them, vegetables don't scream as loudly. It is not good for your own mental state to perceive something else suffering, whether or not that thing is actually suffering, because it puts you in an an unhealthy position of ignoring your own inherent sense of compassion.
If you've ever had the pleasure of dealing with an abattoir worker, the emotional strain is telling. Spending day after miserable day slaughtering confused, scared, captive animals until you're covered head to toe in their blood is... not good for your mental health.
The Texas Chainsaw Massacre often gets joked about because this "based on a true story" wasn't in Texas and didn't involve a chainsaw and wasn't a massacre. But what it did get right was how Ed Gein, the Plainfield Butcher, had his mind warped by decades of raising and killing farm animals for a living.
I think it's more about how we think about the Turning test and how we use it as a result-- a hammer does a pretty poor job of installing a screw, but does that mean the hammer was designed wrong?
Turing called this test an "imitation game" because that's exactly what it was-- the whole point of the test was to determine whether a system could give convincing enough responses that a human couldn't reliably identify whether they were speaking to a human or a machine. Cleverbot passed the Turing test countless times, but people don't ask it to solve their homework or copywrite for them.
From the wiki article on the Turing Test:
The test results would not depend on the machine's ability to give correct answers to questions, only on how closely its answers resembled those a human would give
Humans are so good at imagining things alive that just reading a story about Timmy the pencil is eliciting feelings of sympathy and reactions.
We are not good judges of things in general. Maybe one day these AI tools will actually help us and give us better perception and wisdom for dealing with the universe, but that end-goal is a lot further away than the tech-bros want to admit. We have decades of absolute slop and likely a few disasters to wade through.
And there's going to be a LOT of people falling in love with super-advanced chat bots that don't experience the world in any way.
Maybe one day these AI tools will actually help us and give us better perception and wisdom for dealing with the universe
But where's the money in that?
More likely we'll be introduced to an anthropomorphic pencil, induced to fall in love with it, and then told by a machine that we need to pay $10/mo or the pencil gets it.
And there’s going to be a LOT of people falling in love with super-advanced chat bots that don’t experience the world in any way.
People fall in and out of love all the time. I think the real consequence of online digital romance - particularly with some shitty knock off AI - is that you're going to have a wave of young people who see romance as entirely transactional. Not some deep bond shared between two living people, but an emotional feed bar you hit to feel something in exchange for a fee.
When the exit their bubbles and discover other people aren't feed bars to slap for gratification, they're going to get frustrated and confused from the years spent in their Skinner Boxes. And that's going to leave us with a very easily radicalized young male population.
Everyone interacts with the world sooner or later. The question is whether you developed the muscles to survivor during childhood or you came out of your home as an emotional slab of veal, ripe for someone else to feast upon.
The whole time everyone has been freaking out about AI I've been quietly enjoying just this fact. Like "neat, this place triggers my fear response", "neat, advanced text prediction triggers my 'talking to person' response."
We're good at scamming investors into thinking that a room full of monkeys on typewriters can be "AI." And all it takes to make that happen is to waste time, resources, lives and money, (ESPECIALLY money) into building an army of fusion-powered robots to beat the monkeys into working just a little bit harder.
Because that's businesses solution to everything: work harder, not smarter.
We’re good at scamming investors into thinking that a room full of monkeys on typewriters can be “AI.”
Current generations of LLM from everything I've learned are basically really, really, really large rooms of monkeys pounding on keyboards. The algorithm that sifts through that mess to find actual meaning isn't even particularly new or revolutionary, we just never had databases large enough that can be indexed fast enough to actually find the emergent patterns and connections between fields.
If you pile enough libraries in front of you and can sift out the exact lines that you know will make you feel a certain way, you can arrange that pile of information in ways that will give you almost any result you want.
The thing that tricks a lot of us is we're never really conscious of what we want. We want to be tricked though, we want to control and manipulate something that seems conscious for our own ends, that gives a feeling of power so your brain validates the experience by telling you the story that it's alive. You see pictures that look neat and depict the scenes you wanted to see in your mind, so your brain convinces you that it's inventing things out of nothing and that it has to be magically smart to be able to mash pikachu with darth vader.
There are two alternative solutions to the Turing Test. The one here is when the Judges become dumb and can't differentiate between AI and humans. That is the one in the meme.
The other is when the humans become dumb and can only regurgitate memes that closely mimic how AI chat bots respond to human chatters. Even make a comment on a controversial topic, only for someone to argue with you, but not reference anything specific thing you said? I did and called them a bot as an insult. Then I checked their comment history and figured out it was a stolen account.
