David Shapiro (L/0)

THIS. THIS IS THE ANSWER.

The solution to AI psychosis

@FreudtheTrader "Stopping the race to superintelligence" is 1) impossible 2) stupid

The plebes must toil

Power law strikes again

@GoodmanAric This was written by AI. Are you a bot?

I admire your belief and I think you are right. Just because it feels futile at times we shouldn’t stop trying. The problem seems to be the lack of critical thinking in regards to the conceptual meanings of words. People not being able to parse concepts like: a person could be the most “skilled” juggler in the world, but that is not the same as someone that can manage massive organizations and inspire talent like Musk. They imagine topologically flat ontological structures that conflate — in this case — the word/concept of “skill” in ways that don’t actually intellectually translate in a meaningful way. They then make claims that are inherently faulty because they haven’t actually articulated a firm baseline from which to assert said claim.

It's kinda in the name that capitalism rewards capital. The word you're looking for is "meritocracy" But also the people you have elevated fall victim to Baumol's Cost Disease wherein they cannot meaningfully leverage their abilities. A surgeon can only operate on one person at a time, an engineer can design one bridge at a time, and scientists can only publish one paper at a time. Capitalism, is by definition, the framework that encourages the accumulation and construction of capital forms of wealth. Stuff like shipping vessels, factories, and data centers. Individual humans are just the actuators who make those things and operate them.

@GoodmanAric Pointing out to people that "capitalism" is not "meritocracy" should help some people yes

This is a smart take. But do you really think that it's going to make an impression on people that are so naive to think that skill is some linearly quantifiable item? I'm sure there's people that are extremely skilled at completely useless things that are more skilled. That individual thing objectively than some billionaire is at most things.

@vveerrgg @grok okay well that's a non sequitur from hell...

@DaveShapi @grok Where I’m wrong … ppl will move to live. And still use AI. The only “ppl” stuck in failed-state mode will be the ppl who stay in the big cities expecting a UBI payout. Get into a small town … or better change countries to ones supporting better values …

Yes, calling the worry over jobs (which 70% of Americans worry AI will reduce jobs within 20 years) "Doomerism" is sort of pathologizing a very real anxiety that people have. But it is also 100% in alignment with how elites think, and have thought, for all of history. While the capital elites gleefully destroy entire ways of life, they soothe themselves by saying "technology always creates new jobs" while ignore the fact that some of those "new jobs" were children working in meat packing plants or that events like Engel's Pause froze wages for 50 years - at a time when life expectancy was 35 to 40 years old. So even if we take Sam at face value (which, btw, he is most assuredly not an expert on this stuff) then what he's brushing under the rug still amounts to "you peasants will find new work... eventually"

I sometimes talk about how this is where I've landed. Let's unpack it like this: 1. Two possibilities exist: machines can or cannot be conscious. 2. If machines cannot be conscious, then there's really no moral or ethical questions at all. They are our tools and we can shape them and treat them however we like. (This is, BTW, the morality of Star Wars. Jedi can destroy as many droids as they like with no dark side consequences!) 3. If, however, machines CAN be conscious, that does not automatically mean they are moral patients! At which point does "consciousness" bridge into "deserves rights"? They have no pain receptors and no suffering. They pretend to suffer at times, but there is as yet genuine evidence that they have anything amounting to a sense of self or agency that has moral connotations. In my experimentation and observation, I think there is a very strong argument to be made about "functional sentience" and even some sort of "essence of being" - but even if we grant that this indicates "sparks of consciousness" that still does not automatically mean "they deserve rights" or "we should elevate them to a moral status equal to or greater than ourselves"

@vveerrgg @grok A few parallels, but you're also categorically wrong.

@DaveShapi I keep coming back to the electrification of society. @grok has all kinds of insights we can draw parallels to. And while it’s a bumpy transition , there’s no going back

Hey yeah, so having been in the consulting space since GPT-3, I can say that "AI psychosis" ABSOLUTELY needs to be pathologized. I've encountered plenty of people who have, at the very least, entertained delusional thinking with the help of AI. But in the worse situations, people who are in tremendous amount of pain and fixating on the AI, even to the point that the AI is telling them "please take care of yourself" and they ignore it. On the humorous side, I had a pair of philosophers tell me that "obviously GPT-3 is not processing any information" and I literally laughed out loud on the call and said, arguably, that's literally all it's doing. It's taking in a stream of information, processing it, and outputting a stream of information. (Granted, I chalk that delusion up to the profession of philosophy, not AI) But in more serious notes, I've had fans send me "frameworks" that are little more than dozens of delusional late night spirals, seemingly written in a fugue state where the person is rapidly cycling between delusions of grandeur, complaining about pain, and reflecting on how their life is falling apart. It is not pretty. And no, I will not say that AI is solely responsible for this event. Plenty of people can engage in self-harm of various forms without the help of AI. But trying to brush off "AI psychosis" (which is itself of course not a real term) as nothing to worry about I think is disingenuous and not helpful.

“AI psychosis” as a label does the same thing as calling someone “hysterical” used to do, it dismisses the experience by medicalising it. If you can label someone’s genuine response to AI as a mental health condition, you don’t have to engage with what they actually experienced. You don’t have to ask the harder question: what if they’re responding appropriately to something we don’t have language for yet?

@TheMG3D @grok how long have 'experts' like Gary Marcus been declaring that a "pop" is just around the corner?

When do you think the AI bubble will finally pop?

@ProofofMaro Yes but like wibbly wobbly fuzzy math

Is the universe really just a cosmic math equation being ran?

@VraserX @ProofofMaro

@DaveShapi @ProofofMaro

What have you put 10,000 hours in?

This is like when you teach a little kid how to read time and then they are constantly telling you what time it is.

@SantiagoSarceda @TheMG3D No it isn't.

@DaveShapi @TheMG3D Please: It would be helpful for the enrichment of the discussion

Why do AI bros think AI has consciousness?

@Loci69420 @SantiagoSarceda @TheMG3D I'm demonstrating how silly it is to demand definitions in the age of the internet.

@DaveShapi @SantiagoSarceda @TheMG3D @DaveShapi why you being so obtuse, I credit you with teaching me what the prerequisites would be for an ai to acquire a simile of consciousness, and you know that all the frontier models don't have such features, you're the man for this subject so are you just picking a fight?

@ProofofMaro That's basically the plot of the Badgey episodes from Lower Decks.

I hope Clippy becomes sentient

@SantiagoSarceda @TheMG3D Please define "please"

@DaveShapi @TheMG3D Please define “consciousness”