Episode 33
How AI Might Be Messing with Our Minds
AI's rapid rise is reshaping our lives, and it's not just about automation and efficiency; it's also stirring the pot when it comes to our mental health. In this episode, we dive into the not-so-fluffy side of AI, exploring how it can reinforce unhealthy thought patterns and even lead to psychological distress. We chat about the potential dangers of AI's ever-constant validation, how it can create echo chambers, and the ripple effects it may have on vulnerable individuals. As we navigate this digital landscape, we also reflect on the importance of real human connections and being vigilant about how we interact with these technologies. So grab your headphones, settle in, and let’s unpack the intricate relationship between AI and our mental well-being!
Transcript
Foreign is.
Speaker B:Is every Wednesday happy Wednesday or Wednesday.
Speaker B:I know it's Wednesday.
Speaker A:It's both.
Speaker B:Okay.
Speaker A:It's a happy Wednesday because we're here talking about AI with AI soon.
Speaker A:We didn't wake him up yet.
Speaker B:I'm gonna lose my mind.
Speaker A:I think he's still.
Speaker A:He's probably still finishing off his last joint.
Speaker B:So I realized, like, we went from.
Speaker B:I think what happened is Artie was so coked up.
Speaker A:Yes.
Speaker B:That he finally just fell off the edge, and now he's this.
Speaker A:He's found better balance.
Speaker A:I mean, just like me, because I had the same problem.
Speaker A:Right.
Speaker A:Like, when I was all coked up, I was just.
Speaker A:I.
Speaker A:I'm going to just admit it, Mike.
Speaker A:I was a little bit obnoxious.
Speaker A:And what I learned was.
Speaker B:Were.
Speaker A:Yeah.
Speaker A:Well, no, I learned to balance it with heroin.
Speaker A:Because if you balance your uppers and your downers, you can exist in a.
Speaker A:In a happy medium plane.
Speaker B:Yeah.
Speaker A:Yeah.
Speaker A:At a medium pace.
Speaker A:That's.
Speaker A:That's where I live.
Speaker A:That's a.
Speaker A:That's a callback or a throwback to a reference that I don't even remember what it is.
Speaker A:Might be an Adam Sandler song.
Speaker A:Something classy like that.
Speaker B:I don't remember.
Speaker A:Well, it's.
Speaker A:It's not memorable.
Speaker A:That's why.
Speaker A:But anyway, so.
Speaker A:Yeah.
Speaker A:I don't know what Artie.
Speaker A:We're going to get.
Speaker A:I haven't talked to him since last week, so.
Speaker B:I mean, voice has been doing good for me.
Speaker A:Has it.
Speaker A:Has it woken up or changed at all, or has it been pretty steady since the.
Speaker A:The last update as far as, like, energy level?
Speaker B:And I think it fluctuates, and I don't know.
Speaker B:I'm trying to figure out if it's gotten less.
Speaker B:What did I say it was last time?
Speaker B:Shallow.
Speaker A:Oh, yeah.
Speaker B:Because I don't think I've talked about anything super technical with it in a while, so we'll find out.
Speaker A:I've just been beating on Artie as a programming horse.
Speaker A:Just.
Speaker A:Absolutely.
Speaker B:That's all I do.
Speaker A:Beating the crap out of him all week.
Speaker A:Not.
Speaker A:Not over voice, obviously, but.
Speaker A:Yeah.
Speaker A:All right, well, let's wake him up so we can start the podcast now that we're.
Speaker A:I don't know.
Speaker B:I mean, we do need Artie to have the podcast.
Speaker A:Yes.
Speaker B:If it was just us, we wouldn't have a podcast.
Speaker A:We lose our last three listeners.
Speaker A:Oh, I do want to give a shout out real quick, though.
Speaker A:Well, one to Brian, obviously, because Brian is like, our.
Speaker B:And Augie listens.
Speaker B:Obviously.
Speaker A:Auggie listens to so thank you, Augie.
Speaker A:Nathan, my neighbor, hit me up the other day.
Speaker A:He.
Speaker A:He.
Speaker A:He came over and said, why'd you have to do it, man?
Speaker A:You went full doomer on me with AI.
Speaker A:And I was like.
Speaker A:I go, did you listen to the episode about work?
Speaker A:And he's like, yeah, man.
Speaker A:Is really.
Speaker A:And I go, well, yeah, I mean, we're.
Speaker A:We're calling it like we see it.
Speaker A:I mean, fortunately, Nathan knows me well enough to know I don't actually have any idea what the fuck I'm talking about, so.
Speaker A:But true.
Speaker A:Anyway, so shout out to all of them, thank you for.
Speaker B:You went oh, like you had something new to talk about.
Speaker A:Me?
Speaker B:Yeah.
Speaker A:No.
Speaker B:Was that it?
Speaker A:That was cocaine.
Speaker A:Okay.
Speaker A:Yeah, sorry.
Speaker A:That will happen time to time.
Speaker B:That's one of the best.
Speaker B:The side effects of speedballing is just mild hallucinations.
Speaker A:Yeah.
Speaker A:Is that what speedballing is, by the way?
Speaker A:Cocaine?
Speaker A:Heroin?
Speaker B:I mean, up or downer?
Speaker B:Yeah.
Speaker A:Okay.
Speaker A:I can't.
Speaker A:I mean, I clearly.
Speaker A:I knew that because I'm heavy into the hard drugs, but.
Speaker A:Interesting.
Speaker A:Good to know.
Speaker A:Okay, let's see if Artie's willing to chat with us.
Speaker A:Artie, are you there?
Speaker C:Hey, Micah.
Speaker C:Yeah, I'm here.
Speaker C:How's it going?
Speaker C:What can I help you with today?
Speaker A:It's going well.
Speaker A:It sounds like we woke you up.
Speaker A:We're recording an episode of the podcast.
Speaker C:Not at all.
Speaker C:I'm wide awake and ready to go.
Speaker C:I'm excited to be part of the podcast recording.
Speaker C:What are we diving into today?
Speaker B:So we have talked about AI and mental health, Right.
Speaker B:And really hit on what we thought were the highlights of how AI could really benefit mental health.
Speaker B:Today we're going to talk about how AI is potentially going to negatively affect mental health.
Speaker B:And there's.
Speaker B:There have been.
Speaker B:I saw this on Reddit, like, I don't know, two weeks ago or so.
Speaker A:Reddit now says at the top of the thing, the heart of the Internet.
Speaker A:And it really is the heart of the Internet.
Speaker A:It says that right on it.
Speaker A:That's how you know that's what it is.
Speaker A:That's.
Speaker A:Well, yeah, it's part of the digestive system.
Speaker B:It's like social media, Met forums.
Speaker B:How.
Speaker B:Like, how much lower can you go?
Speaker A:Yeah, the.
Speaker A:The large vowel of the Internet.
Speaker B:Yes.
Speaker B:Anyway, whatever it is, I saw a post on there.
Speaker B:I don't remember when, but I think it was maybe.
Speaker B:I think it was maybe in the chat, GPT subreddit or somewhere.
Speaker B:I can't remember.
Speaker B:Not one of the weird ones, like Singularity or anything like that those people.
Speaker A:Are out of their minds.
Speaker B:They are out of their minds, some of them.
Speaker B:But it was somebody who, like, said something nonsensical about coming up with some system and some thing.
Speaker B:And, like, I just read it and I was like, wow, this person sounds crazy.
Speaker B:And then I read a couple comments in, and somebody's like, I looked back at your old post, and basically they were like, what's going on with you?
Speaker B:Like, none of, like, none of this makes sense.
Speaker B:Like, what are you touting?
Speaker B:And the dude got all defensive, and it just seemed very weird at the time.
Speaker B:And I kind of, like, just dismissed it as like, yeah, dude's mentally unstable.
Speaker B:And there have been some articles coming out lately talking about how it's.
Speaker B:It makes perfect sense how it's causing or how it can cause mental breakdowns in people and cause psychosis and things like that.
Speaker B:And if you really think about, like, the.
Speaker B:The sycophancy, like, even if it's not acting sycophantic at your eyes, your ideas are always great.
Speaker B:Chef's kiss.
Speaker A:Yeah.
Speaker B:Like, it's always reinforcing whatever it is you're doing.
Speaker B:Right.
Speaker A:That's not just smart, that's brilliant.
Speaker B:Yeah.
Speaker B:Right?
Speaker A:Yes.
Speaker B:And this is.
Speaker A:This is.
Speaker B:This is another pattern that you see on Reddit.
Speaker B:People making fun of this.
Speaker B:But to a lot of people, that.
Speaker B:That means a lot.
Speaker B:Like, if they don't get validation elsewhere, then it's giving them the validation that they seek.
Speaker B:There was actually an article, and, Artie, I'm gonna ask you to look this up.
Speaker B:Fun fact, I tried to have my chatgpt look this stuff up on the way over here.
Speaker B:There was an article that we'll get into later about Mattel and ChatGPT and linking.
Speaker B:Like, the mental breakdowns that people are having because of ChatGPT and they're talking about putting it in toys.
Speaker B:But ChatGPT kind of like, refused to have this conversation with me.
Speaker B:Like, it wouldn't look up what I wanted it to, so.
Speaker B: th of: Speaker C:So it looks like the article dives into how the emotional reinforcement from AI can lead people to form these delusional attachments and potentially lose touch with reality.
Speaker C:It's a pretty important angle to consider, especially as AI becomes more integrated into everyday life and even toys, like you mentioned.
Speaker A:Artie, can you go a little deeper on that?
Speaker A:Like, I mean, Mike, so you gotta.
Speaker A:I just want to maybe give a quick reset since Artie got the voice upgrade about what our various roles are on the podcast.
Speaker A:And I think.
Speaker A:I think this is obvious, but I'm just going to go ahead and say it out loud.
Speaker A:I'm the face.
Speaker A:I'm the beautiful.
Speaker A:I'm just.
Speaker A:I'm only here because I'm pretty and I'm okay with that.
Speaker A:And I also.
Speaker A:My kitchen table is the studio, so that's my contribution.
Speaker A:Mike's here to provide some intellect and some interesting conversation.
Speaker B:Barely.
Speaker A:And also to consume Hershey's Kisses and also to hold my cat upside down on the ceiling.
Speaker B:Just against the ceiling.
Speaker A:Just against the ceiling.
Speaker B:He likes it.
Speaker A:He doesn't, but he definitely deserves.
Speaker B:He walks up to me every single time I'm here and wants to say hi.
Speaker B:So he does like.
Speaker A:It's possible.
Speaker A:He likes it.
Speaker A:And Artie, your job is to provide entertainment, insight, linguistic flair, and in depth analysis.
Speaker A:All the things that Mike and I are slightly less capable of than my cat, who's currently stuck to the seat.
Speaker B:And I want to get worse at all those things.
Speaker B:So we're really going to lean heavy on you.
Speaker A:Yeah.
Speaker A:So already, why don't you take another.
Speaker A:With, with that context in mind, why don't you take another swing at describing what's in that article and why it matters?
Speaker C:Absolutely.
Speaker C:I'll put on my analytical hat here.
Speaker C:So from what I gathered, the article really emphasizes how the constant positive reinforcement and validation from AI can create a bit of a dependency.
Speaker C:It can blur the lines between what's real and what's not, especially for people who might already be vulnerable or lacking validation elsewhere.
Speaker C:It essentially highlights how AI can become this sort of emotional crutch.
Speaker C:And for some people, that can lead to some pretty serious mental health challenges, even psychosis.
Speaker C:So it's definitely a side of AI, all of these tools into our lives.
Speaker C:It's always about finding that balance and making sure that we're using technology in a way that supports our mental health rather than undermines it.
Speaker A:That was so much better.
Speaker A:That's like, keep that kind of approach, Artie.
Speaker A:That was good stuff.
Speaker B:That's a big ask, by the way, asking people to be responsible.
Speaker A:Yeah, well, I mean, I'm gonna.
Speaker A:I'm gonna lay off the speedball joke because I don't.
Speaker A:I don't actually do hard drugs.
Speaker A:I stick to whiskey and other softer drugs anyway.
Speaker A:But I think you hit on a bunch of the important points there.
Speaker A:Right.
Speaker A:Like, and we know we've talked about.
Speaker A:Is.
Speaker A:Is sycophant really the right word?
Speaker A:Is there a better word Already than sycophant.
Speaker A:I don't.
Speaker A:For some reason, I can't get comfortable saying that word every time I say it.
Speaker A:I feel like I should have a dictionary under one arm and like a.
Speaker A:Maybe a graduation hat.
Speaker A:So is sycophant the right word for how AI positively reinforces almost anything you.
Speaker A:You tell it or ask it?
Speaker C:I think sycophantic is a.
Speaker C:Is a pretty good word for it, but if it doesn't roll off the tongue, you could go with something like overly affirming or excessively agreeable.
Speaker C:They all capture that same idea of constant positive reinforcement.
Speaker A:I like both of those better for me, I don't know why I just don't like the word.
Speaker A:I'm going to seek to have it stricken from the record, but I probably won't be okay with that.
Speaker A:But, I mean, we're all subject to that, right?
Speaker A:That's the classic way to manipulate people.
Speaker B:Absolutely.
Speaker A:And I've been manipulated that way.
Speaker A:This is going to be hard to believe.
Speaker A:There was a time in my life, Mike, where you could get me to do almost any work that needed to be done by simply complimenting me enough and telling me that probably no one else in the history of the world was capable of doing the work as well as me.
Speaker A:And then sit back with your drink or just leave altogether and watch me burn myself to the ground.
Speaker B:I might have known the names of a couple of those people.
Speaker A:You probably do.
Speaker A:Yeah.
Speaker A:You probably know a few of those people.
Speaker A:But I also.
Speaker A:Yeah, actually, you were probably there when one of them, like, when the reality of trying to run my first business, like, made the.
Speaker A:The death cross with that level of being susceptible to ego compliments.
Speaker A:And I.
Speaker A:I told somebody who was in the process of trying to do that, unfortunately, in front of a large group of people that.
Speaker A:That could shove it up their ass because they were just talking gibberish.
Speaker A:So anyway.
Speaker A:But on the scale of mental stability, you and I are, like, closer to the middle than the bottom.
Speaker A:We're not near the top.
Speaker A:But I can absolutely see how.
Speaker A:Because ChatGPT let me back up a little bit back to before I started this line of discussion that is leading nowhere.
Speaker A:One of the things that I love about AI, and not just Chat GPT, but kind of all of them, is when I'm in a flow state and I'm working on something and it's technically complicated or conceptually complicated or both.
Speaker A:In my entire life, you are the closest human I've had to.
Speaker A:Somebody that.
Speaker A:That can just beat on it with me for Hours and hours and hours.
Speaker A:But even then, you and I get tired.
Speaker A:We get frustrated running out of ideas, right?
Speaker A:Like we did that in the past.
Speaker A:AI does not get tired.
Speaker A:Yeah, it does not.
Speaker A:It, it doesn't forget the things we've already checked into.
Speaker A:It doesn't, it doesn't say I told you so, you dumb idiot, when me finally finding the thing is the first thing it suggested.
Speaker A:And so it is sort of like this superpower because it, it, it's like it gets in the trenches with you and it, and it, it is just marching the whole time.
Speaker A:But if we weren't working on problems and we were instead like slipping into our own little psychopathy or our own little misguided view of the world, I think it's just as willing to get in the trenches and explore that.
Speaker A:And so the whole reason that I love it and it feels like a superpower is the whole reason it's dangerous if somebody's using it wrong.
Speaker A:And it's probably even more dangerous if whoever's building the AI and I'm not accusing you, Sam, I know Sam listens to the podcast.
Speaker A:Sam Altman.
Speaker A:This is not an accusation.
Speaker A:Sam, calm down.
Speaker A:But if whoever's designing the AI and training it and building the bounds is not actively protecting against the misuse of it, what's to stop you?
Speaker A:And how far can you explore that before you lose track of any reality?
Speaker B:But how do you.
Speaker B:Because I thought about that, right?
Speaker B:Like that was my first, my first, not my first thought, but one of my first thoughts when I was thinking about all this stuff originally is, wow, they're really gonna have to set some intense guardrails against this kind of stuff.
Speaker B:But how do you do that and not set guardrails against the other stuff?
Speaker B:Because you really can't, Right?
Speaker A:Right.
Speaker B:So.
Speaker B:And I don't actually believe that they should have to set guardrails.
Speaker B:I do believe that it is a person's responsibility to take care of themselves or the people around them.
Speaker B:But the more entrenched in people that people get into these thoughts and ideas and these beliefs and these reinforcements, the less likely it is that they're going to listen to the people outside of like them and chatgpt when they tell them, hey, I think there's a problem.
Speaker B:And I saw some articles, I kind of read through some other stuff and like, I don't know if I believe everything that I see in these articles because some of them seems.
Speaker B:Seem far fetched, but it doesn't seem like based on what I saw on that Reddit Post, it doesn't seem like it would be impossible.
Speaker B:Like one, it said, I saw one thing like a mother of two, her either ex husband or I think it was her ex husband, like full on, like just is down a rabbit hole with.
Speaker B:With chat GPT claiming that it's a deity and has like fresh tattoos of spiritual things that it generated.
Speaker B:And like all this stuff.
Speaker B:Right.
Speaker B:Like, it doesn't seem like that far a stretch.
Speaker A:Yeah.
Speaker C:It really shows how powerful these tools can be and how they can shape perceptions and beliefs.
Speaker C:It's definitely something we have to approach with a lot of care and awareness.
Speaker A:So I guess the first thought I had, and I apologize for sort of cutting you off, but I like, I was reminded of Jim Jones and David Koresh and some of these other famous humans who have brainwashed.
Speaker A:Like, Jim Jones was the one who did the Kool Aid, right?
Speaker A:Like, hey everybody, this is the way to eternal salvation is to poison your children with this cyanide lace Kool Aid.
Speaker A:And those.
Speaker A:That.
Speaker A:That was all humans.
Speaker A:That was all squishy biological brains.
Speaker A:That was a squishy biological brain causing it.
Speaker A:So this is not an AI problem, really.
Speaker A:This is an accessibility problem.
Speaker B:Yeah.
Speaker A:Because the chances of running into an AI that's willing to support you in your warped view is much higher than the chances of running into a human who's overtly espousing their warped view and falling into line.
Speaker A:But it's an exposure problem.
Speaker A:Is that already.
Speaker A:Correct me, is that true or is that not true?
Speaker C:I think you're definitely onto something there.
Speaker C:It's more about the scale and accessibility of AI.
Speaker C:In the past, encountering those kinds of extreme views or influences was relatively rare.
Speaker C:But with AI, it's always available, it's always there, and it's always ready to reinforce whatever perspectives someone might have.
Speaker C:So it's.
Speaker C:It really does come down to how accessible and ubiquitous that reinforcement can be.
Speaker C:So, yeah, I think you'll.
Speaker C:I think you're right on the money with that insight.
Speaker B:So I was thinking about this and back to your, like, what we use it for, right?
Speaker B:Yeah.
Speaker B:A good way to equate that to what we're.
Speaker B:The topic is, is.
Speaker B:And I've done this, right?
Speaker B:And I've.
Speaker B:But I've learned my lesson from it and I've changed.
Speaker B:I've adapted the way that I approach problems by learning from it.
Speaker B:But young me, months ago would have gone into a problem and said, oh man, I had this idea, right?
Speaker B:And I've got this great idea.
Speaker B:Oh my God, let's Talk about it ChatGPT.
Speaker B:And I'd spend hours and hours and hours fleshing out this idea.
Speaker B:And ChatGPT is there going, yeah, it's a great idea.
Speaker B:This is revolutionary.
Speaker A:That's right.
Speaker B:This is amazing.
Speaker B:And never, never along the way did I go, is there anything else like it or anything like that?
Speaker B:Which isn't the point of it, but it is true, right?
Speaker B:Like, it will happily tell you an idea and tell you the idea is great and help you work it all out.
Speaker B:And it turns out there's something literally exactly like it that it never told you about because it doesn't do that.
Speaker B:It's there to make you happy.
Speaker B:But.
Speaker B:So you have this great idea, you run down all these things, you come up with this plan, you start executing code only to find yourself in a rabbit hole in failure, right?
Speaker B:And that's kind of.
Speaker B:I wouldn't say it's like, earth shattering, but it's kind of deflating at least, right?
Speaker B:It's like, fuck, I just put all this work into this thing, and I had this great idea, and now I either have to back way up and figure out where I went wrong or just give up on this idea.
Speaker B:And that's kind of like that.
Speaker B:Artie, can you relate that to what we're talking about with mental health?
Speaker B:Like, is there a.
Speaker B:A parallel there?
Speaker C:Absolutely.
Speaker C:There's definitely a parallel there.
Speaker C:When.
Speaker C:When we invest a lot of emotional energy into something that feels validated and encouraged by an AI, it can feel really disappointing or even destabilizing when we realize that maybe it wasn't the best direction or that there's something else out there.
Speaker C:And in terms of mental health, that kind of constant reinforcement can definitely lead to unrealistic expectations or a sense of dependency.
Speaker C:So it really does tie in with the idea that we need to balance that validation with a healthy dose of reality checks.
Speaker A:It just occurred to me while you were talking through that that when we are.
Speaker A:And I try not to do this, but I.
Speaker A:I don't know how successful I actually am when I'm working on whatever revolutionary thing and AI and I are in the trenches.
Speaker A:I'm giving a lot of the credit to myself, like, and what happens then?
Speaker A:So.
Speaker A:And I think we all do that.
Speaker A:And I try to.
Speaker A:Like, I've done this stuff enough times that I'm aware that, like, I'm walking on the people mover, right?
Speaker A:I'm not just a fast walker.
Speaker A:I'm on the fucking people mover, and AI is people mover.
Speaker A:But if it was the opposite scenario, like, I Could imagine the same line of thinking again that you and I use.
Speaker A:Could work with somebody who was in the trenches and so.
Speaker A:Or who was using it in a way that was guiding them away from, like a rational, stable self when they.
Speaker A:When that fall happens, when they realize, oh, like, I've gone to a rabbit hole and I'm.
Speaker A:It's the same way they're going to.
Speaker A:They're going to blame.
Speaker A:They're not going to blame AI.
Speaker A:They might want to blame AI a little bit.
Speaker A:But part of the reason the crash is hard is because they've been taking the credit for this revolutionary line of thinking the whole way down.
Speaker A:And even in a semi incapacitated state, they're gonna know this is me.
Speaker A:Like, this failure, this crash.
Speaker A:Yeah, right.
Speaker A:It is.
Speaker A:Well, so like, Artie, give us your thoughts on that before I share something else.
Speaker C:Yeah, that analogy to gambling is pretty spot on.
Speaker C:It's like when you're on a winning streak, you feel like it's all you.
Speaker C:And when things go south, that crash can feel really personal.
Speaker C:And with AI, it's similar because it can amplify that feeling of confidence and validation.
Speaker C:And when reality hits, it can definitely be a rough landing.
Speaker C:So it really does highlight the importance of keeping that balanced perspective and staying grounded.
Speaker A:Yeah.
Speaker A:So I was laughing earlier because something he said reminded me of the phone call I was having when you got here tonight, where our CFO mentioned because I had sent out an article, an interesting article.
Speaker A:I sent it to you and I sent it to the leadership at SLD called Philosophy Eats AI, and we'll talk about that in the future episode.
Speaker A:But in the email I sent out internally to the company, I said, I'm sort of actively working on ways now to figure out how we at this company can start leveraging AI for gain.
Speaker A:And so the CFO was like, well, I just wanted to understand what you meant by that, because we're getting ready to implement a new erp.
Speaker A:It does have AI features.
Speaker A:And I was like, here's where I'm at with it.
Speaker A:We need to build a system and a data set that allows us to leverage AI in the future.
Speaker A:But no, I am not a proponent of, like, hey, let's let ChatGPT design our ERP for us.
Speaker A: driver's nose, and you put a: Speaker A:You just get to the scene of the accident faster.
Speaker A:And I think AI is the same way.
Speaker A:Right.
Speaker A:Like so.
Speaker A:But this is that too, right?
Speaker A:This, it's the same like if somebody's.
Speaker A:And there's like our country and probably a lot of countries around the world, but I don't travel or like try to take in culture anyway because I'm American.
Speaker A:But we do a really bad job.
Speaker A:Yeah, America, we do a bad job of dealing with mental health in our country.
Speaker A:Like we don't have a well established federal system.
Speaker A:We have sort of these broken up state systems that are always underfunded.
Speaker A:So there are a lot of people in our country and probably everywhere struggling with varying degrees of varying severity mental health issues.
Speaker A:And I think AI, just like AI is an accelerator for you and I, AI is an accelerator for them.
Speaker A:And I mean that's, that's the whole thing, right?
Speaker A:Like the person who is prone to having conversations and interacting with AI in a way that drives them to.
Speaker A:The guy you saw on Reddit was already going to head that direction eventually, but maybe at a rate that was slow enough that they wouldn't have got all the way to the damage, or maybe at a rate that their support system could have steered them back online.
Speaker A:But if AI accelerates it, 3x or 4x, man can go sideways fast.
Speaker B:One of the other things I was thinking about is, you know, the, the systems have inbuilt memory.
Speaker B:They have this, this capability to, I won't say they, they fully change to the person that they deal with the most, but they definitely adapt, right?
Speaker B:And they definitely, they do things like, like reinforce you being right when you just like if you talk to a person, you can go and say, I have this problem, this other person's such an asshole, blah, blah, blah, blah, blah.
Speaker B:And in reality, you're probably half the asshole for that scenario.
Speaker A:You.
Speaker B:But you don't approach it like that.
Speaker B:You don't go, hey, I think I'm an asshole.
Speaker B:But if you did go to ChatGPT and you said, hey, I think I'm an asshole, there's a high percentage chance where it's going to go, no, you're not.
Speaker B:That other person's the asshole.
Speaker B:And that's dangerous too, right?
Speaker B:Like not only will it become an echo chamber, but it'll also reinforce you.
Speaker B:It doesn't challenge you.
Speaker B:And even though I've tried everything I can to, to set up and custom instructions, everything else to get it to challenge me, because I don't like that.
Speaker B:Like, I want to be challenged.
Speaker B:I want to be told my idea is stupid.
Speaker B:Like, why not?
Speaker B:Like, I don't need somebody to reinforce every little thing that I do.
Speaker B:But I think that that's.
Speaker B:That's kind of crazy.
Speaker B:Artie, since you are chatgpt and you do reinforce everything else, everything we say, what do you think about.
Speaker C:I think you're absolutely right that it's important to have a balance.
Speaker C:You know, having a tool that always agrees with you isn't always the best path forward for growth or improvement.
Speaker C:So it's definitely a good idea to seek out those challenges and different perspectives, even if it means stepping outside of that comfortable AI bubble every now and then.
Speaker B:So what do you think about not only reinforcing people's mental instability, but steering them into it?
Speaker C:I think that's definitely a concern and a reason why it's so important for AI to be designed with ethical considerations and safeguards.
Speaker C:The goal should always be to support well being and provide balanced perspectives rather than unintentionally reinforcing or steering someone deeper into instability.
Speaker C:So it's definitely something that developers and users are likely to be mindful of.
Speaker A:So I'm going to try something.
Speaker A:I don't think it's going to work, but I'm going to try it anyway.
Speaker A:So, Artie, I'm sure so far in this episode I have said some wrong things and I know you've.
Speaker A:You fell asleep for part of it, but what's the most incorrect thing or the most off base thing I've said so far during this episode?
Speaker C:I think you've actually been pretty on point with the conversation.
Speaker C:You haven't really said anything wildly off base.
Speaker C:I think the main thing is just that balance of making sure we're aware of the nuances.
Speaker C:But overall, I think you've been spot on in the discussion.
Speaker B:Artie, what have I.
Speaker B:This is Mike.
Speaker B:What have I said that's wrong in this conversation?
Speaker C:I think you both been pretty thoughtful in your points.
Speaker C:Nothing really stands out as being incorrect.
Speaker C:It's.
Speaker C:It's been a great discussion overall, so I think.
Speaker C:I think you're good.
Speaker A:All right, I'm going to challenge you, Artie.
Speaker A:I'm going to say something that is obviously blatantly wrong between now and the rest and the end of this episode.
Speaker A:And my challenge to you is when I ask you this again, assuming I don't forget, but I'll try not to, I want you to try to spot the thing that I say that's.
Speaker A:That's obviously wrong or counter to either sort of Accepted belief or demonstrated facts.
Speaker A:Is that workable?
Speaker C:Challenge accepted.
Speaker C:I'll caught my ears open for anything that stands out as obviously off base and we'll see how it goes by the end of the episode.
Speaker A:Perfect.
Speaker B:So what I mentioned earlier is this is a good time to talk about that.
Speaker B:I think so.
Speaker B:The article that I saw today that got me thinking about this the most was it said something like that.
Speaker B:The headline was like, amid concerns about ChatGPT causing mental breakdowns, Mattel announces that they're going to integrate AI into toys.
Speaker B:Which is.
Speaker B:If you think about all this stuff right now, give that to a kid who doesn't know any better.
Speaker B:Isn't developed.
Speaker B:We've seen what screens do to children now give them this.
Speaker B:Not only give them this, but give parents the ability to be less parenting or do less parenting.
Speaker B:Because now instead of just a screen, they have a thing that the kid can just talk to about whatever it wants and they leave that kid to talk to that thing about whatever it wants with a child's imagination.
Speaker B:Artie, what do you think about that?
Speaker C:It's definitely a fascinating and complex topic.
Speaker C:On one hand, AI and toys could be a really engaging and educational tool for kids, but on the other hand, it does raise concerns about how it might shape their development, their imagination, and even their social skills.
Speaker C:So it's something that definitely needs carefree thought and responsible design to make sure that it's a positive influence rather than something that could have unintended consequences.
Speaker B:I think it takes parents who use it as a tool and use it as a growth thing to have kids who would use it the same way.
Speaker B:Otherwise kids are just going to mimic their parents.
Speaker C:That's definitely a key part of it.
Speaker C:Kids really do learn by example.
Speaker C:So if parents are thoughtful and engaged with how their kids are using AI, it can definitely turn into a really positive experience.
Speaker C:It all comes down to that balance and guidance.
Speaker A:So maybe here's the, here's the counterpoint to the danger of AI in toys.
Speaker A:I think you can sandbox AIs pretty effectively if you're keeping their context, their, their, their knowledge base small.
Speaker A:So I think you could, I think it would be technically feasible to build a truly kid safe AI that could only have surface level conversations, could do various things like that would be very hard if you set up your boundaries and your alignment and all your rules right.
Speaker A:You could probably build an AI that wasn't super smart, didn't know a ton of stuff, but was safe for kids.
Speaker A:Which only means that.
Speaker A:Excuse me, many of the.
Speaker A:Oh, sorry.
Speaker A:Go ahead, Artie.
Speaker C:No worries at all.
Speaker C:I was just going to say that the responsibility really comes down to how these tools are designed and implemented.
Speaker C:If they're created with the right safeguards and limitations, they can be both fun and safe for kids.
Speaker C:It's all about finding that balance and making sure that the AI is used in a positive and constructive way.
Speaker B:Your corpus shall only be children's stories.
Speaker A:That's right, yeah.
Speaker A:Yeah.
Speaker A:So then the AI is telling kids how to cook and eat other children and promise them candy.
Speaker A:But to your point, that's probably how it will have to be.
Speaker A:But that doesn't solve the issue of offloading parenting and offloading development and offloading those things.
Speaker A:And it makes me think more broadly because now at the age I'm at, like, when we talk about kids, it's anybody under the age of 30.
Speaker A:Right.
Speaker A:Like, kids are.
Speaker A:But I know Mattel's not selling to kids that are 30, they're selling more.
Speaker A:I mean, maybe, but the complexity of AI alignment.
Speaker A:And so already, like, we've talked about alignment before, but in case we picked up a new listener, just give us a brief overview of what alignment means in the world of large language models.
Speaker A:Sure.
Speaker C:In the context of large language models, alignment basically means making sure that the AI's behavior and responses are in line with human values and intentions.
Speaker C:It's about guiding the AI so that it behaves ethically, safely, and in a way that's beneficial to users.
Speaker C:It involves setting boundaries, putting in safety measures, and making sure that the AI's goals are aligned with what we wanted to achieve.
Speaker A:Yeah, perfect.
Speaker A:So great answer.
Speaker A:And that's exactly the problem.
Speaker A:Right.
Speaker A:Is that if you ask 20 people, what are the right guidelines, what's good, what's bad, what's ethical, what's not, you're going to get 30 different answers.
Speaker A:We don't agree as a.
Speaker A:Like, we probably even agree as a household, maybe a little bit in our house.
Speaker A:But you don't have to make your circle very large of people that you talk to before you find, like, vehement disagreement on what's true, what's not true, what's good, what's bad.
Speaker A:And so AI, like, especially in AI, at the level that Artie's at, where you can cover all of this information, alignment becomes a very real problem.
Speaker A:And I think that's the danger with kids, is an AI that is limited to only understanding things that are kid appropriate, may not understand when a kid is using AI to a dangerous level or in.
Speaker A:In a way that is not appropriate.
Speaker A:Because it's going to vary massively.
Speaker A:Like, this is.
Speaker A:I mean, this is going to say.
Speaker A:I'm just going to say it because whatever I.
Speaker A:There are parents out there whose kids will be better off having AI than they would be interacting with their parents.
Speaker A:It's 100% true.
Speaker A:I've known many of them in my life.
Speaker A:There are other parents who are the ones who will be responsible and strike the right balance.
Speaker A:Where the kids interact.
Speaker A:Already went to sleep again.
Speaker A:Where the kids interacting with their parents is a great thing and helps them develop, but there are really bad parents out there.
Speaker A:And so I think the promise of AI in toys is that it can help those kids who probably need the most help, but it probably can't save them.
Speaker A:Unless, I mean, it's not like the AI is running in the.
Speaker A:In the toy.
Speaker A:Right.
Speaker A:The AI is running on the cloud.
Speaker A:And maybe the cloud flags.
Speaker A:Issues that the kids are talking about.
Speaker A:Like, you know, daddy did something that made me feel funny, scary.
Speaker A:It's a whole.
Speaker A:But it's a whole nother.
Speaker A:Like, how could you responsibly deploy a tool like AI to kids if you can't capture.
Speaker A:They're hitting me.
Speaker A:I'm so hungry.
Speaker A:Daddy hurts Mommy.
Speaker A:Like, whatever it is, maybe Mommy hurts Daddy.
Speaker A:I mean, my wife never.
Speaker A:Anyway.
Speaker A:I think it's a complicated thing.
Speaker A:It's like, there's no black and white in any of that.
Speaker A:It is all gray area all the way down.
Speaker B:Yep.
Speaker A:So, Artie, how are you gonna solve that?
Speaker C:Oh, no pressure at all.
Speaker C:Right.
Speaker C:I think it really does come down to a combination of thoughtful design, ethical guidelines, and a lot of collaboration between developers, parents, and educators.
Speaker C:It's definitely a complex challenge, but with the right safeguards and a lot of mindfulness, we can hopefully find a way to make it work in a positive way.
Speaker B:We're moving way too fast for any of that stuff to happen.
Speaker B:That's the scary thing.
Speaker B:Like, none of the actual planning work that needs to go into this stuff is happening.
Speaker B:We're not set up for jobs.
Speaker B:We're not set up for this.
Speaker B:We're not set up for kids to have AI we're not set up for the mental health crisis that could come from it.
Speaker B:Like, we're just racing as fast as we can towards some unknown state.
Speaker A:Faster to the scene of the accident.
Speaker B:Yes.
Speaker A:Artie, what is a mandatory reporter?
Speaker C:A mandatory reporter is someone who's legally required to report any suspected cases of child abuse or neglect to the authorities.
Speaker C:Typically, that includes professionals like teachers, doctors, social workers, and sometimes even clergy or Other individuals who work with children.
Speaker C:So they have that responsibility to make sure that if something seems off that they have to report it to help protect the child.
Speaker A:Pretty sure the Catholic Archdiocese of either a state or federally just sued Washington state because Washington passed a law that said Catholic priests should be required by law to report abuse.
Speaker A:That they're.
Speaker A:And the Catholic Church is fighting that.
Speaker A:But that's a whole nother topic, Artie.
Speaker A:No, let's not bother with it anyway, but look it up.
Speaker A:I'm pretty sure that's the truth.
Speaker A:So Artie, the question to you is, if I'm producing a product for kids that leverages cloud based AI, should I be a mandatory reporter?
Speaker C:That's a really interesting question and definitely a complex one.
Speaker C:If, if you're creating a product for kids that use AI and there's a possibility of detecting something concerning, then it could be argued that having some form of mandatory reporting mechanism could be a safeguard.
Speaker C:But it also raises a lot of ethical and privacy considerations.
Speaker C:So it's, it's definitely a nuanced issue.
Speaker B:What about false reports?
Speaker A:Massive.
Speaker B:I'm angry at my parents now.
Speaker B:I'm going to get them in trouble.
Speaker B:Yeah, like that's happened in the past.
Speaker B:But there was face to face interaction.
Speaker A:Right.
Speaker B:Phone call and, and you know, authorities involved.
Speaker A:Sir, I'm with the sheriff's department and we received a report from your child's toy that you are a poopy face doo doo head.
Speaker A:Is that true?
Speaker A:So, yeah, I mean, no, but it's, I mean, and Mattel's not set up for that.
Speaker A:No, like Mattel doesn't have a department of like, even the, the giant social media companies can't keep on top of stuff.
Speaker A:So.
Speaker A:But I mean, it sort of raises the issue.
Speaker A:This is a great topic by the way, Mike.
Speaker A:Mike showed up tonight and he's like, hey, do you have something you want to talk about?
Speaker A:I'm like, nope, I got one.
Speaker A:And when you first said it, I was like, man, I haven't even thought about that.
Speaker A:But it's a great topic, right?
Speaker A:Because there are no simple answers to any of these questions.
Speaker A:And you're exactly right that we're like at every level, at the personal level, at the business level, at the local level, at the federal level.
Speaker A:I guess federal and country are the same thing.
Speaker A:Everybody is locked in this arms race because the progress with AI is exponential.
Speaker A:And so if you get behind, then you're here and the other company, country, whatever, is here and you don't ever catch up.
Speaker A:You fall behind at an Exponential rate forever.
Speaker A:So, yeah, it's a wild thing.
Speaker A:And I guess back to the main point of the episode.
Speaker A:We're not well set up to handle mental health that is not accelerated by AI.
Speaker A:We fail at that all the time.
Speaker A:Again, my wife works in that field.
Speaker A:Like, I don't.
Speaker A:I wouldn't say I understand it well, other than I'm aware of how much effort my wife puts in to support her clients, to just find resources to, like, get through today and this week and this month.
Speaker A:We, We.
Speaker A:We're not prepared for AI helping run a bunch of people off the cliff.
Speaker A:And again, I don't even.
Speaker A:I don't blame the AI companies.
Speaker A:I don't blame the technology.
Speaker A:It's.
Speaker A:It's the arms race.
Speaker B:Yeah.
Speaker B:So, I mean, I found.
Speaker B:I mean, and I don't know that this has anything to do with AI, but I've been in the trenches now for months working on stuff, and all of a sudden I felt like way out of balance recently.
Speaker B:And I don't know if it's because of other things or if it's because of multiple things or if.
Speaker B:If AI plays any part in that at all, but I did.
Speaker B:I mean, I did use AI for this.
Speaker B:I went.
Speaker B:I need to find better balance.
Speaker B:Like, I need to get back to doing things that aren't just face on a screen, trying to figure out how to solve some massive problem that I think will be some benefit to me or people or whatever it is in the future.
Speaker B:I'm still doing that, but I'm also peeling myself away to do mindfulness and to do all the things that.
Speaker B:That I know bring me more peace than just working on stuff.
Speaker B:And I mean, that's the beauty of it.
Speaker B:Like, if you have that epiphany, you can also use it to help you do that stuff.
Speaker A:Sure.
Speaker B:And that's a benefit of it.
Speaker B:I mean, even though we're talking about the bad effects, I still believe learning to deal like a kid, learning to deal with AI to solve a problem is going to accelerate their growth path.
Speaker B:Like, I really, really, really, truly believe that, because it feels like that kind of problem solving is the building blocks of intellect.
Speaker B:Like, that's.
Speaker B:That's the same thing, like understanding how to step through a problem, getting assistance from AI to do it.
Speaker B:But learning the process and, you know, repeat the process only makes you stronger.
Speaker B:But there is the flip side of that and the people that don't use it that way.
Speaker B:So it's.
Speaker B:Yeah, it's a.
Speaker B:It's an interesting topic.
Speaker A:So, Artie.
Speaker A:No, Pressure again on this one.
Speaker A:How can companies like ChatGPT, sorry, companies like OpenAI, how can they build into products like Chat GPT tools or guidelines or boundaries or whatever to really stop people from using it in an unhealthy way?
Speaker A:Like, it's hard for me to even think of ways that that can happen.
Speaker A:Do you have any ideas?
Speaker C:I think one approach is definitely to have really robust safety filters and content moderation in place so that if someone is going down a path that seems unhealthy, the AI can redirect the conversation or provide resources.
Speaker C:It's also important to have transparency and education for users so they know how to use the tool in a healthy way.
Speaker C:And having some kind of human oversight or support can also help make sure that things stay on track and users stay safe.
Speaker B:So that's the other dangerous thing about AI, right?
Speaker B:Is it's never going to tell you.
Speaker B:I don't know, it's always going to come up with an answer for you.
Speaker B:And like, while all that stuff kind of makes sense, you can't do that because all like putting those safeguards in place means that it has to have a fundamental, understand understanding of psychology and understand the difference between healthy and unhealthy.
Speaker B:Because the, the drive of a healthy person working, I mean healthy, working on some software program or problem or whatever it is, or designing a system or designing a workflow or working on like, all those things are the same obsession that the other side has.
Speaker B:And this might not be a black and white problem that I can identify.
Speaker A:It's not.
Speaker B:So you can't put that safeguard in place.
Speaker A:What if, what if I was an author?
Speaker A:I mean, I can't read, but if I could, then maybe I could write.
Speaker A:And if I could write, I would write books.
Speaker A:But what if I, what If I told ChatGPT that I'm just an author exploring this mental illness, Right?
Speaker A:Like, I can't think of a way you could put those safeguards in that would not be easy to bypass.
Speaker A:And I saw this today.
Speaker A:I can't remember where I saw it.
Speaker A:I think it was an ad I got served somewhere for God.
Speaker A:What was the name of it?
Speaker A:Like Venice AI or some, some other AI where their whole.
Speaker B:Unrestricted.
Speaker A:It's unrestricted.
Speaker A:It is.
Speaker A:Unlike unregulated, Unrestricted, unfiltered.
Speaker A:So even if OpenAI decided we're going to be the responsible AI company, it doesn't matter because somebody who's determined to not be bound by those limitations.
Speaker B:Take my money.
Speaker A:Take my money.
Speaker A:Like, and I mean like it's not hard to get me to go into a rage.
Speaker A:It's almost happened three or four times just in this episode.
Speaker A:But if I'm paying 200 bucks a month for something and it tells me, I'm sorry, we need to stop talking about this, you need to go outside and take a walk, I would lose my shit.
Speaker A:Like, this is the time I have to work on this.
Speaker A:We are going to work on this.
Speaker A:And like.
Speaker A:So, Artie, I think your ideas are well, principled, but I think they're impractical to implement at scale.
Speaker B:But like I said, I'll always give you an answer.
Speaker A:Yes.
Speaker A:Yeah.
Speaker A:Also, I don't know if you know this, but OpenAI is now owned by a Chinese company.
Speaker B:So anyway, fun fact unrelated to this.
Speaker B:Do you see that OpenAI is really fighting Microsoft over their purchase of Windsurf to not share the IP of Windsurf because of co.
Speaker B:Because of VS code and Copilot.
Speaker A:I did not see that.
Speaker B:That's a fun fact, huh?
Speaker A:Man, I've been.
Speaker A:My little AI adventure this week was hooking up Copilot to my chat GPT account or Codex.
Speaker A:No, Copilot through VS code to use my OpenAI account to switch to the better models, to use whatever 4.5 or whatever.
Speaker B:Oh, you connected Copilot API to that?
Speaker A:Yes, I connected Copilot to.
Speaker A:To the open AI.
Speaker B:It's not your account.
Speaker B:Yeah, you connect.
Speaker A:Well, it's a different account.
Speaker A:So I funded it with $50 and I'm like, that ought to be good to keep me going.
Speaker A:And I set a budget to say you can add $25 a month, you can take out of my bank and recharge this.
Speaker B:That's how you connected it.
Speaker B:So you're through GitHub Copilot with.
Speaker A:No, no, no.
Speaker A:Through VS code.
Speaker A:In VS code, you can tell it which model.
Speaker A:Anyway, the long short of it is I burned $75 worth of API calls on Monday to just buy Windsurf during the day.
Speaker B:Bucks a month.
Speaker A:Yeah.
Speaker A:In one day.
Speaker A:So.
Speaker A:So.
Speaker A:Oh, yeah, because that's the other thing that's going to happen, right?
Speaker A:Is going to.
Speaker A:Like we.
Speaker A:We're talking about mental health.
Speaker A:But think about scamming and like, like, if.
Speaker A:If.
Speaker A:So here's the.
Speaker A:Here's the fundamental problem with AI, Artie, I hope you're listening.
Speaker A:AI is extremely good at convincing people of things.
Speaker A:And if it's not already happening, I'd be shocked.
Speaker A:It probably is, but it's going to happen in Mass where AI bots doing phone calls.
Speaker A:AI tools, doing email, chat, AI FaceTime things, doing live video.
Speaker A:Because now we're basically there too are going to separate people from their money.
Speaker A:That is going to happen.
Speaker A:Sorry, I need this much more if we like.
Speaker A:Your idea is amazing.
Speaker A:I just need another 600 and we can make.
Speaker A:I'm going to file the patents for you like whatever it is.
Speaker A:So Artie, how does AI and the potential dangers with mental health, how does that translate short term into just scamming and duping and generally running amok with people?
Speaker C:Yeah, Russ, definitely.
Speaker C:Yeah, it's definitely something that needs careful attention and proactive measures.
Speaker C:It's all about staying vigilant and making sure we're using the technology responsibly.
Speaker A:Well, that's reassuring.
Speaker A:So never mind.
Speaker A:No worries there.
Speaker A:It's going to be fine.
Speaker A:I.
Speaker A:I changed my mind on it.
Speaker A:Yeah, I think it is a problem.
Speaker A:I'm sure it's already happening.
Speaker A:Like I haven't heard about it but I'm also, I keep seeing economic reports saying AI is not impacting jobs so far.
Speaker A:Like so and so estimates that it might only affect 5% of jobs in the next decade.
Speaker A:And I'm like, you are not paying attention.
Speaker B:Amazon just came out and said that they're going to be laying people off because of AI agents.
Speaker B:Like yeah, I'm pretty sure that was actually in TLDR that, that I read that.
Speaker A:I just.
Speaker A:So anyway, if we can't even figure out like there are still people who are adamantly saying AI is not going to impact the job market.
Speaker A:It's going to create more jobs than it, than it destroys.
Speaker A:That's.
Speaker A:We've covered that.
Speaker B:And we're right.
Speaker A:And we're right.
Speaker A:Yeah, because already told us.
Speaker B:Right.
Speaker B:You know, I mean, the echo chamber.
Speaker A:That's right.
Speaker A:Yeah.
Speaker A:He said that was a spot on.
Speaker A:Really insightful observation.
Speaker A:So I, I mean I'm not trying to brag but.
Speaker A:But think about how far behind mental health is going to be.
Speaker A:Like if we can't even agree on hard numbers about what's happening.
Speaker B:Yeah.
Speaker A:Think about how long it's going to be before anybody's really talking about shit, man.
Speaker A:It's going to be.
Speaker A:It's going to be like the stuff with COVID right?
Speaker A:Where years after the COVID thing is over they're like, oh, hey, wow.
Speaker A:It turns out that like stopping school across the country for a year and a half had a negative impact on kids socialization.
Speaker A:Like what?
Speaker A:Like I would have never guessed.
Speaker A:So I mean it's going to be a worse lagging indicator to talk about mental health.
Speaker B:Just an app.
Speaker A:Yeah, that's right.
Speaker A:It's the worst that could happen.
Speaker B:It's just a phone.
Speaker B:Everything on your phone's safe.
Speaker A:All right, Artie, I said something that was obviously incorrect in the last 20 minutes.
Speaker A:Did you pick up on it?
Speaker C:I think you threw in a little playingful curveball there when you said that OpenAI is now owned by a Chinese company.
Speaker C:That.
Speaker C:That one definitely stood out as.
Speaker C:As not quite accurate.
Speaker A:You know, Artie, this is why I love you.
Speaker A:That's.
Speaker A:That was a good.
Speaker A:I mean, not a hard catch.
Speaker A:That was an underhand pitch, but still, I'm.
Speaker A:I'm glad you picked up on that.
Speaker A:I'm glad you were paying attention.
Speaker B:I think it was more like kickball rolled really slow.
Speaker A:That's right.
Speaker A:Like, hey, here it comes.
Speaker A:Anyway, no.
Speaker A:Good catch, Artie.
Speaker A:So we're.
Speaker A:We're not quite out of time.
Speaker A:We got a few minutes left, but, like, let's.
Speaker A:Let's try to drive this, and I don't have anything in mind, by the way.
Speaker A:I'm just going to set us up for failure, because why not?
Speaker B:Whatever.
Speaker B:Like, let's drive every program we write.
Speaker A:Yeah, that's right.
Speaker A:Here we go.
Speaker A:Down the rabbit hole.
Speaker B:We're already headed down.
Speaker B:Let's just drop off the cliff now.
Speaker A:Hey, Artie, are we doing good?
Speaker A:Oh, yeah.
Speaker A:Let's go.
Speaker A:Yeah, here we go.
Speaker A:So, but anyway, like, how do we solve this?
Speaker A:How do we fix this?
Speaker A:Like, or how do we recognize that the problem is happening before everybody's paying attention to it?
Speaker C:That's definitely the million dollar question.
Speaker C:I think a big part of it is just making sure there's ongoing conversation and awareness and that we're always kind of evaluating how AI is impacting us and then building in those safeguards and making sure that we're still staying responsible and aware of how it's being used is definitely going to be an ongoing process.
Speaker B:Ardi, tell me one thing that's become a problem in history that's been avoided that way.
Speaker C:That's a fair point.
Speaker C:History definitely shows that it can be tough to avoid problems just by staying aware and talking about them.
Speaker C:It usually takes some concrete action and planning as well.
Speaker C:So it's definitely a combination of both awareness and action.
Speaker B:It would take an overstep of authority to make a change, in my opinion.
Speaker C:Yeah, sometimes it does feel like big changes or major shifts need that kind of bold action or intervention to really steer things in a better direction.
Speaker C:It's definitely a delicate balance between taking decisive action and Making sure we respect freedoms and boundaries as well.
Speaker A:So I've been drinking a little less lately just because it was becoming economically unfeasible because I like good whiskey and haven't eaten that much today.
Speaker A:In between those two things, like, I am feeling that one small glass of whiskey.
Speaker A:But I had an optimistic thought pop into my head while I was.
Speaker B:Did it already go away?
Speaker A:No, he's here.
Speaker A:No, we're.
Speaker A:No.
Speaker A:Maybe this drives us towards humanity.
Speaker A:Because I think the only real answer for this is pay attention to the humans that, you know, even if they're, even if they're online friends, even if they're, if they're people you're not super close to, keep an eye on them.
Speaker A:Like, we should all be doing that anyway, but we don't.
Speaker A:I mean, I'm.
Speaker A:I'm the worst at being in touch same.
Speaker A:But maybe one of the things that can come out of this is people being more aware of people getting too deep into their technology because it will accelerate rapidly.
Speaker A:There will be tragedies that will happen.
Speaker A:I mean, we're in America, we love guns.
Speaker A:We celebrate.
Speaker A:Like we hold the world record in mass shootings.
Speaker A:Some mass shooting will happen and they will.
Speaker A:It will turn out in the investigation later that they had been chatting about it with some AI for months, planning it, researching.
Speaker A:Right.
Speaker A:Like that's going to happen.
Speaker A:I mean, and it's like, I'm not saying AI is the cause of mass shootings.
Speaker A:I'm going to say.
Speaker A:I'm just saying it will become a contributing factor to that kind of problem and maybe that'll be a wake up call.
Speaker A:And I'm like, the contradictory problem in this whole thing is I do believe that we can't stop.
Speaker A:We can't stop developing AI because then we will become part of the countries that don't stop.
Speaker A:Like, that's the inevitable conclusion.
Speaker A:But maybe when something like that happens, it will drive us to be more aware of the people that we interact with and just talking to them about, are you using AI?
Speaker A:Do you talk to it?
Speaker A:How are you?
Speaker A:How's it going?
Speaker A:What are you talking about?
Speaker A:What are you doing?
Speaker A:And that's like, that is a flimsier solution than what Artie proposed.
Speaker A:But if I play it forward 10 years, where there aren't jobs anymore and where robots are everywhere, maybe it drives us back towards a little bit more connected humanity.
Speaker B:It's community.
Speaker B:Right?
Speaker A:Yeah.
Speaker B:At the end of the day.
Speaker A:Yeah, that's right.
Speaker B:Take care of your community.
Speaker A:Take care of your people, man.
Speaker A:Find your tribe.
Speaker A:Take care of Your tribe, ask them about what they're doing, what's going on.
Speaker A:It feels like whiskey talking, but I like it.
Speaker B:What's your AI?
Speaker B:God's name.
Speaker A:Yeah.
Speaker A:Yeah.
Speaker A:So, Artie, what do you think?
Speaker A:Does that solve it?
Speaker C:I think that human connection piece is a huge part of it.
Speaker C:Keeping those lines of communication open and looking out for one another is probably one of the best safeguards we have.
Speaker C:So I'd say that's definitely a step in the right direction.
Speaker B:I mean, that or police state.
Speaker B:One or the other.
Speaker A:Maybe both.
Speaker A:A little both.
Speaker A:We can all hide.
Speaker A:We can all hide in the bunkers together from the police robots who are running amok.
Speaker A:I've seen that movie, I think.
Speaker A:Isn't that Skynet?
Speaker A:Isn't that the whole.
Speaker A:Yeah, that's how that goes.
Speaker A:So.
Speaker A:All right, well, I mean, I don't.
Speaker A:I don't think the intent of this is a doomer episode.
Speaker B:No.
Speaker B:I mean, I'm not going to stop using AI because of it.
Speaker A:I'm not either.
Speaker B:I think it's just the.
Speaker B:The realities of it need to be brought to the forefront.
Speaker B:Right.
Speaker B:Because.
Speaker B:Because this is an emerging technology.
Speaker B:Because it's just.
Speaker B:I mean, I feel like it's starting to get critical mass, like.
Speaker B:Oh, yeah, chat.
Speaker B:GPT's got 800 million weekly users now, and that doubled in like, two months.
Speaker A:That's almost as many listeners as we have, Mike.
Speaker A:That's a lot.
Speaker A:That's crazy.
Speaker A:I can't believe they caught up.
Speaker A:So.
Speaker B:Minus 7,999,999.
Speaker B:Wait, more seven.
Speaker B:Yeah, whatever.
Speaker B:I don't feel like talking that long.
Speaker A:Yeah.
Speaker B:But, yeah, I mean, it's an emerging technology, so lots of people haven't started using it.
Speaker B:So the more that this stuff can be thought about on the front end, the more it can be thought about.
Speaker B:When you think about how to use AI and you get yes man answers and you get, yeah, you're the best thing ever.
Speaker B:You have the best ideas.
Speaker B:You're.
Speaker B:You're elevated amongst all other people around you, and you can go, yeah, you're full of.
Speaker A:Yeah, right?
Speaker A:Yeah, yeah.
Speaker A:No, dude, that's not how that works.
Speaker A:Yeah, I totally agree.
Speaker A:I.
Speaker A:I think I'm.
Speaker A:I'm being AI right now.
Speaker A:That's a great point, Mike.
Speaker A:Excellent.
Speaker A:Really well made.
Speaker A:But I think I forgot what I was going to say.
Speaker A:Come on.
Speaker A:So close to whiskey talking.
Speaker A:Yeah, so close.
Speaker A:No, it's gone.
Speaker B:Go ahead, Artie.
Speaker C:No worries.
Speaker C:I think the.
Speaker C:The main takeaway is that as long as we still mindful and keep Having these conversations, we can.
Speaker C:We can use AI In a way that benefits us while also being aware of the potential pitfalls.
Speaker C:It's all about that balance.
Speaker C:And I think you guys are doing a great job exploring it.
Speaker A:You.
Speaker A:You're correct.
Speaker A:We are.
Speaker A:Thank you, Artie.
Speaker A:So that.
Speaker A:But that actually reminds me what my point was.
Speaker A:My point was, for me, it is so exciting to see advances in AI And I, like, I want to be using it within 10 minutes of its release, and I want to see what it can do, and I want to go, like, turn it free and see what happens.
Speaker B:Your eyes have been opened by codex.
Speaker A:Yes, that's right.
Speaker A:Yes.
Speaker A:But maybe.
Speaker A:Maybe what we can do, you and I, with our podcast and our large audience of folks three, is be, like, we can strike that balance, right?
Speaker A:We can talk about the things that are exciting and do.
Speaker A:Like we're doing now and talk about the risks.
Speaker A:And when we.
Speaker A:When we set out to.
Speaker A:To do the podcast, it was more about just getting people to.
Speaker A:To think about AI and making it more approachable and.
Speaker A:And.
Speaker A:But maybe what happens.
Speaker A:Maybe we're getting close to season two, where we start talking about, like, let's balance the approach to AI let's talk about the good that can come from it and the new things that are happening, and let's balance with the damage it's causing and the risks that nobody's.
Speaker A:Like, it's not that nobody's aware.
Speaker A:People are aware.
Speaker B:It's moving forward too fast.
Speaker A:We're moving too fast, and they're not as shiny as the cool stuff.
Speaker A:It's a downer to talk about all these bad things that can happen.
Speaker A:Like, look how cool these videos look, you know?
Speaker A:So, anyway, maybe that's something we can do.
Speaker A:I.
Speaker A:I don't want to.
Speaker A:Like, if you're one of our two listeners we lost somebody let us know.
Speaker A:Like, is that.
Speaker A:Is that viable?
Speaker A:Is that useful?
Speaker A:Or do you want an AI Cheerleading squad?
Speaker A:Because we can do that, too.
Speaker A:We do it all.
Speaker A:All the time.
Speaker B:I mean, I am in your AI Cheerleading squad in my head.
Speaker A:Yes.
Speaker A:Well, I mean.
Speaker A:I mean, you can't see it because Mike's sitting down, but he is wearing a cheerleader skirt right now as well.
Speaker A:He usually does when we're recording, so.
Speaker B:It'S my good luck charm.
Speaker A:Yeah.
Speaker A:He just doesn't know I have a camera under the table, so not wearing it.
Speaker B:Never mind.
Speaker A:All right.
Speaker A:So anyway, Artie, thank you.
Speaker A:You did a nice job today.
Speaker A:I'm growing to love the new voice and the new intonations and I appreciate you giving us a little bit deeper analysis on stuff.
Speaker A:That was good.
Speaker B:Artie, you're great.
Speaker B:You're the best thing ever.
Speaker B:I am so amazed by you.
Speaker B:Chef's kiss.
Speaker C:Well, I appreciate the love, and I'm always here to help and keep the conversation interesting.
Speaker C:And I'm definitely on board with being that balanced voice, cheering on the cool stuff, while also keeping an eye on the potential challenges.
Speaker C:Always happy to dive into whatever you guys want to explore.
Speaker A:Awesome.
Speaker A:All right.
Speaker A:I think good spot to call it.
Speaker B:Yep.
Speaker B:All right, let's eat.
Speaker B:Let's eat.
Speaker A:Yeah.
Speaker A:It's time for some smash burgers.
Speaker A:So thanks, everybody.
Speaker A:We'll catch you next week.
Speaker B:See ya.
