ChatGPT

[G. Keenan]

Why can there not be a machine that responds the stimuli, solving various problems, etc having no subjective experience whatsoever?

I think there totally can be. I'd say you've just given a pretty good description of plants, for example.

This might have been one of the interviews I was remembering with Geoffrey Hinton. Good listen. How Neural Networks Revolutionized AI

[Arthur Dent]

I’m using it for work to write lengthy proposals and similar projects. Saves me a lot of time. I only make minimal revisions.

It’s only a matter of time before people on the other end use AI to interpret my lengthy proposals, to make their decisions.

It’s terrifying. Legitimately.

I guess this comes down to the question of whether writing and reading lengthy proposals is an important human activity that should be preserved for its own sake. To me, this is tedious, and if there were way to automate it while still somehow making good decisions, that would be great. On the other hand, there are other aspects of my job that I do enjoy for their own sake that could plausibly be at least dramatically reduced in human effort with stuff like this, and I would miss them and resent the lack of agency involved in such a change.

To me, this kind of thing is most usefully understood in terms of economic systems and how they control our lives, though I certainly understand that this feels pie in the sky in a way that smashing the machines seems less so. I’m not really sure.

Right now, I would not trust doomy big picture predictions or really any predictions about medium or long term implications.

[ghostrunner]

Wow

https://www.thetransmitter.org/retracti ... retracted/

An article that proposed potential benefits of private equity firms investing in autism service providers has been removed from the journal in which it was published.
The article, “Private equity investment: Friend or foe to applied behavior analysis?” was originally published in the International Electronic Journal of Elementary Education as part of a January 2023 special issue devoted to applied behavior analysis (ABA) for autism.

ABA is the most widely used therapy for autism, and companies that provide it have faced a flood of interest from private equity firms in the past decade.

The sole author of the article, Sara Gershfeld Litvak, “decided to retract the article due to her commitment to scientific integrity and ethical values,” following “a rigorous review process,” according to the undated retraction notice on page 266 of the special issue. Litvak is founder and CEO of the Behavioral Health Center of Excellence (BHCOE), a company that offers accreditation for organizations that provide ABA services, and she co-founded the Autism Investor Summit, an annual meeting focused on the business side of autism services

[sighyoung]

Wow

https://www.thetransmitter.org/retracti ... retracted/

An article that proposed potential benefits of private equity firms investing in autism service providers has been removed from the journal in which it was published.
The article, “Private equity investment: Friend or foe to applied behavior analysis?” was originally published in the International Electronic Journal of Elementary Education as part of a January 2023 special issue devoted to applied behavior analysis (ABA) for autism.

ABA is the most widely used therapy for autism, and companies that provide it have faced a flood of interest from private equity firms in the past decade.

The sole author of the article, Sara Gershfeld Litvak, “decided to retract the article due to her commitment to scientific integrity and ethical values,” following “a rigorous review process,” according to the undated retraction notice on page 266 of the special issue. Litvak is founder and CEO of the Behavioral Health Center of Excellence (BHCOE), a company that offers accreditation for organizations that provide ABA services, and she co-founded the Autism Investor Summit, an annual meeting focused on the business side of autism services

This is precisely where AI falls down: these databases don't have access to a lot of scholarship, which is often behind paywalls. You can't rely on it to know and cite current scholarship.

[Joe Shlabotnik]

I'm not calling anything sentient that you can 1. Unplug, 2. Can't nourish itself, 3. Can't reproduce.

[ghostrunner]

It does seem like we’re all jumping the gun using AI to describe this kind of software. Like how “VR” is sometimes used instead of “interactive 360 degree camera”.

[Arthur Dent]

It does seem like we’re all jumping the gun using AI to describe this kind of software. Like how “VR” is sometimes used instead of “interactive 360 degree camera”.

This was basically my previous line, but I dunno. Have you tried it recently?

Without the baggage and related motivated reasoning, it’s pretty hard to say it’s not some kind of intelligence. It’s not like a genius, and there’s plenty of quirks, and blah, blah, blah, but yeah. More than just hype for sure.

[ghostrunner]

It does seem like we’re all jumping the gun using AI to describe this kind of software. Like how “VR” is sometimes used instead of “interactive 360 degree camera”.

This was basically my previous line, but I dunno. Have you tried it recently?

Without the baggage and related motivated reasoning, it’s pretty hard to say it’s not some kind of intelligence. It’s not like a genius, and there’s plenty of quirks, and blah, blah, blah, but yeah. More than just hype for sure.

I'm no expert on what's actual intelligence or not. I know GPT passes the Turing test, or some claim it does. But in my mind it means self-awareness and the ability to create without specific references or waiting for input. Don't ask me to define specific references. From an artistic standpoint, for some reason the example I think of is Nicolas Cage in Moonstruck. It's a very weird performance, and knowing a fair amount of historic film performances I can't tell you where he might have pulled it comes from. I don't believe AI could do that for example, assuming it were technically capable of creating a performance. It might create a credible Humphrey Bogart impression, but the odds are it'll pull from something you've seen from him, rather than create a unique performance from both personal experience and decades of film history like a real actor would.

Also along the lines of what Joe said.

I certainly wouldn't consider taking art seriously if it was created without any of that. Not that it couldn't prove useful to some, but I don't think it can have cultural value. I guess sentience is what I'm looking for maybe. If it doesn't have that, it seems limited. And I'm somewhat limited by a lack of terminology to express what I mean.

Seems like the use case currently is handling non-creative busywork. I'm a little skeptical of even having it create a rough draft for a technical or analytical paper, say. Feels to me like the work involved in correcting and shaping something into final form would take longer than if you had just written something yourself. At least you know where you were coming from. It could have a chance of being poorly written, sourced or constructed, and is it worth the gamble? Would you trust it to get sources right, for example? How much time is it worth chasing those sources down vs just doing the writing?

I don't have a good answer for any of that, but it seems like the smaller the scale of what you want it to do, the better.

I'm not at all a hardcore "machines can't be intelligent beings" person, and I think the lines between AI and humans are probably thinner than people want to admit. I just know this doesn't feel like real intelligence to me (as I perceive it). Maybe that's a cop out though.

[Arthur Dent]

"Can it create great art?" seems like an unreasonably high bar for intelligence. Most people can't do that either.

I have no idea what will happen with this stuff, and the "it will become super intelligent and kill us all" prophecies seem like a huge leap. I'm just saying, it does not really make sense to say it's like various stupid hype things like the metaverse or whatever.

We can define intelligence in whatever ways we want to allow ourselves to dismiss this. That was my instinct as well. But it's a mistake. It's a huge breakthrough that displays intelligence nothing like anything that's come before.

Is that intelligence "like" human intelligence? I think probably not, but also I don't know. Getting at that type of question is largely impossible at this point and maybe forever.

[Arthur Dent]

Also, I'm curious what AI skeptic Jim thinks about this.