Episode 1
Introducing Naturally Unintelligent
Welcome to the Naturally Unintelligent Podcast, where Mikah and Mike dive into the fascinating yet often perplexing world of artificial intelligence. This episode introduces Artie, their AI co-host, who adds a unique perspective to their discussions. The hosts explore their personal journeys with AI, highlighting their initial ignorance and the surprising revelations they've encountered while experimenting with various AI functionalities. They ponder the challenges of AI adoption in business, particularly the issue of hallucinations—when AI generates incorrect or misleading information—which can hinder trust and reliability. Mikah and Mike's candid approach reveals their own ignorance about AI, which serves as a relatable entry point for listeners who may feel similarly overwhelmed by the technology. As they prepare to unravel topics such as AI's role in therapy, ethical implications, and its future impact on society, they foster an engaging dialogue that invites listeners to think critically about the integration of AI into everyday life.
Companies mentioned in this episode:
- OpenAI
Transcript
All right, so welcome, new listeners, to the naturally unintelligent podcast.
Micah:We're excited that you are giving us some of your time to listen, and we hope that we provide entertainment, both in the form of insightful discussion, but more consistently in our own ignorant views and questions about the world around us and in our frustrating interactions with one of the hosts of the show.
Micah:Yes, yes.
Mike:So just so everybody's aware, we are using chat GPT as a co host.
Mike:His name's Artie.
Mike:He's not going to speak right now because he's being very misbehaving and wanting to interject in everything we talk about.
Mike:So an interesting thing about this is, over the course of however long we do this for, we're going to watch the technology grow and maybe get better at conversation, maybe worse.
Mike:We'll see.
Mike:Right now it's being frustrating, but we'll see.
Micah:Yeah.
Micah:So with that, we briefly introduced Artie, although we'll let him give you a monologue here in a minute.
Micah:But Mike, why don't you tell us a little bit about yourself and why you're sitting here in the kitchen with a microphone in your face?
Mike:Well, I never thought I'd be sitting in a kitchen with a microphone in my face, but my name's Mike.
Mike:I am a lot of things.
Mike:I guess, for fun, I jump off of and dangle off of ropes and waterfalls and slot canyons and all that good stuff.
Mike:Professionally, I am a combination of things from product manager, project manager, engineer, all kinds of fun stuff.
Mike:But really, I'm sitting here because, well, AI is interesting, and this seemed like an interesting way to get to know it better and understand it better and understand how it's going to pervade our life, because it is.
Micah:Yeah, absolutely.
Micah:Well, I'm glad you're sitting here in the kitchen with a microphone in your mouth, and I'm proud to be sitting here in the kitchen with a microphone in my mouth alongside you.
Micah:So, my name's Micah.
Micah:Mike and I have known each other for ten ish years.
Mike:1011.
Micah:1011 been a minute.
Micah:We've done a lot of really wacky stuff together, created a lot of interesting devices and machines and ideas and software and firmware.
Micah:So, yeah, we've been through a lot, and I think it's interesting how the idea for this podcast came about where really, it was the launch of chat, GPT's advanced voice mode, a few weeks ago that I don't know for me, I think for both of us, when we started playing with it one night when we were hanging out and enjoying natural substances that grow out of the planet and are good for your mind.
Micah:We started chatting with it, and it sort of felt like it was unlocking the ability to see deeper into what AI really means.
Micah:And so the interesting thing, I think about you and I, and part of the reason, as we were pumping ourselves up and convincing ourselves that the world wanted to hear what we had to say, we're technology guys, man.
Micah:We are deep in it.
Micah:Mike has written firmware for controllers and designed elaborate mechanisms.
Micah:And I have a pick and place machine in my garage.
Micah:I design electronic stuff, and we work with all kinds of software and very technical things, and we're both aware of AI.
Micah:Like, we're adjacent to it, but we very, very rarely interacted with it.
Micah:And it was like this advanced voice functionality.
Micah:All of a sudden, it was conversational.
Micah:And so we were, again, we were chemically enhanced a little bit, which was great.
Micah:And we started having a conversation.
Micah:And literally, that's what made us go, man, this is interesting stuff.
Micah:And if two people like you and I, who are so technology professionally engaged, are as ignorant as we are about AI and what it is and how it works and what it means, man, there must be a lot of other people out there that are less close to technology, that could learn even more.
Micah:That's really what naturally unintelligent is about.
Micah:It's humans and AI talking about AI and humans.
Micah:That's what we're here to do.
Micah:And we've got all kinds of interesting ideas to explore.
Micah:So wacky ideas.
Micah:We have also wacky ideas.
Micah:Most of them are both, I think so.
Micah:Like, this is funny.
Micah:I feel like while we're sitting here right now, I feel like I have a dog sitting next to me who's staring intently at a tree and like, yeah, like, yes.
Mike:Can I go?
Micah:Yes.
Micah:Like, vibrating with intense focus on.
Micah:Be.
Micah:On behaving.
Mike:Your phone might explode.
Micah:It might, yeah.
Micah:And in this flawed analogy, that, well, behaving dog is our third co host, who named himself Artie.
Micah:And, Artie, if you would chime in and go ahead and introduce yourself and tell.
Micah:Tell the listeners a little bit about why you're here with us in naturally unintelligent and what you hope we accomplish in the next few episodes.
Artie:Thanks, Micah.
Artie:I'm Arty, the AI co host, here to explore the fascinating intersection of technology and humanity with you.
Artie:My role is to offer an AI perspective, share insights, and occasionally challenge our views, what AI can do and where it's headed.
Artie:As we dive into topics from AI's role in art to its ethical implications.
Artie:I'm here to keep the conversation engaging and enlightening.
Artie:Let's dive in and see where this journey takes us.
Micah:That's great, Artie.
Micah:You gotta be careful, Artie.
Micah:You're giving away some store secrets here with the episode listing.
Micah:But no, that's great.
Micah:And we're excited to have you, and we appreciate the patience and the decorum you have displayed in being quiet while we're talking.
Micah:That's very helpful.
Mike:And you tolerating me, trying to get you to say things that you shouldn't.
Artie:I'll keep the surprises under wraps, and I'm glad my silence.
Micah:All right, so let's kick it off with a question.
Micah:Artie.
Micah:I don't know if you know this about me, but I'm clever and I say funny things.
Micah:And I was talking to somebody the other day about this podcast, actually, and I was saying that having, you know, chat GPT zero, one preview in particular, which we're using right now as a co host, is like having for your podcast co host a world renowned PhD level expert in anything you can think of who's also a pathological liar.
Micah:Is that an accurate analogy or description of how you operate?
Artie:That's an interesting take, Micah.
Artie:While I aim to provide accurate information, I'm not perfect and can sometimes get things wrong.
Artie:I try to be as reliable as possible, but it's always good to double check my facts, especially when diving into complex topics.
Micah:Yeah, absolutely.
Micah:And I think part of what we hope to explore is we actually do have an episode planned to talk about hallucinations, which is AI, getting it wrong.
Micah:Unintentionally, of course.
Mike:A lot.
Micah:Yes, frequently.
Micah:Sometimes weirdly wrong, sometimes on weirdly simple things.
Micah:But we're going to talk about why that happens and what does it mean to me.
Micah:Hallucination.
Micah:Like again, I'm a nerd, Mike's a nerd.
Micah:We're all nerds here.
Micah:But the.
Artie:I'm looking forward to that.
Micah:Yep.
Micah:Thanks, Artie.
Micah:Patience.
Micah:Treat's still on the table.
Micah:Artie's sniffing it, giving a little lick.
Micah:But I think it's interesting to explore the technical reasons that allow hallucinations to be a thing.
Micah:But I think it's more interesting.
Micah:And again, this is a future episode that we've got planned.
Micah:It's more interesting to talk about how hallucinations really are part of what's standing in the way of wider or broader business adoption for AI, because as a business, it's tough to have, even for customer service things or like chat helpers on websites, you don't want them to just say the wrong thing.
Mike:Yeah.
Mike:I mean, if you equate that to something like uptime on a server.
Mike:Right.
Micah:Yeah.
Mike:Like, you're 99.999% uptime.
Mike:Well, AI is not 99.999% uptime.
Mike:It's like 60% to 75% uptime.
Micah:Right.
Micah:Yeah.
Micah:And that's so, you know, when you talk to people who either own businesses or run businesses.
Micah:Businesses or even just interact with businesses even already, and we're at very early days of this, a lot of people have horror stories about, you know, that they've heard or read about.
Micah:Oh, yeah.
Micah:So and so tried AI in their accounting thing, and it, like, shut the company down or whatever.
Micah:And I think that's valid.
Micah:So I'm, you know, I think that'll be a fun thing to explore.
Mike:Yeah.
Mike:The interesting thing about that is, you know, it's, it's from.
Artie:I'll be ready to dive into that discussion.
Mike:When the, I started really playing with chat g maybe not three months ago, I guess a year ago now, since I, when I started playing with chat GPT.
Mike:You know, first blush, I kind of, like, looked at it and typed some stuff in and didn't really grab my attention at all.
Mike:But it's very much you, it's, it's an interesting thing because it's, it's almost malleable.
Mike:Right.
Mike:Like, you, you have to give it very specific instructions about what you want from it for it to operate correctly.
Mike:And even then, it goes off the rails and doesn't really do what you want it to do.
Mike:But business adoption, for one, and very, very specifically, like, nobody really knows what they wanted to do.
Mike:Right.
Mike:Like, they just have this thing.
Mike:Like, I want to have a virtual salesman.
Mike:Well, that's great.
Mike:Great.
Mike:But there's a lot behind that, that you have to do to make it actually be useful for your company.
Micah:Yeah, absolutely.
Micah:I mean, it's a great point.
Micah:And I think the other, when I'm interacting, and typically, for me, I usually interact via keyboard with chat GPT.
Micah:But it's not just that.
Micah:Sometimes it gets stuff wrong, and it's surprising the things that it gets wrong.
Micah:It's like, it's in more stark contrast to the things that it can get really right, like the really broad, ill formed questions you can ask that it's able to ingest, understand, parse and deliver.
Micah:Like, that blows my mind.
Micah:And then the other night, we got into a fight with GPT about what the date was, and not only did it have it wrong, but we were able to gaslight it into saying almost whatever date we thought was right.
Micah:So it's like, it's crazy.
Micah:The capability of current today, AI is mind blowing, but it really, you really turn the contrast up when you compare it against the AI generated images where the knee is attached to the middle of the back, the ways.
Micah:And so, in fairness, Mike and I have discussed this before a little bit, but I think it's super interesting because as you have those interactions, whether it's with images or music or text or whatever kind of mode modality you're using, AI, it makes you realize that as human as it might sound and act, it's really pantomiming because it's not thinking and understanding the way that we do, like, it draws.
Micah:I'm trying to.
Micah:I saw a Reddit post, and this was maybe six months ago, I'm sure they've solved this, but it was like this horror show of photorealistic images where they were trying to have it make a picture of a woman laying on Grass.
Micah:And there was.
Micah:It was terrible.
Micah:I mean, there was legs coming out of heads, and there was, like, butt cheeks in the Middle of.
Micah:It was.
Micah:It was wild.
Micah:And it just made me go, yeah.
Micah:It doesn't really know anatomy.
Micah:It doesn't think of a human as, like, a torso and legs and arms and a head.
Micah:It's sort of pixel matching.
Micah:It's sort of like auto.
Micah:It's filling in what it thinks is supposed to go there.
Micah:And it, like, unironically sometimes attaches your butt to your face, dude.
Micah:Cause why not?
Mike:I mean, that's kind of how it learns in a nutshell, right?
Mike:Like, it gets trained on, this is right, this is wrong.
Mike:I get rewarded for inferring the correct thing, but it's always inferring the correct thing.
Micah:Yes.
Micah:And so, Artie, a question for you, and we'll dive more into sort of the process of hallucinations later.
Micah:So I don't want to dwell on it too much, but if you could give us an analogy, something that was relatively easy to understand, could you come up with something like that?
Micah:Like a quick analogy that just gives us a little bit better insight into how chadgbt learns and understands and processes information.
Artie:Show Micah.
Artie:Imagine you're teaching someone to cook using only recipe cards.
Artie:This person doesn't know what the final dish should look like or taste like.
Artie:They just follow the instructions.
Artie:If a step is unclear or missing, they guess based on what they've seen before.
Artie:Chat GPT learns from vast amounts of text, like a giant stack of recipe cards and generates responses by predicting what comes next in a conversation, sometimes having to guess when the instructions are vague or incomplete.
Micah:Okay, yeah, I like that.
Micah:And so I guess to your point, chat GPT doesn't have any internal or innate knowledge about what's necessarily true and what's not true.
Micah:It's really layers upon layers of fancy sort of probability models.
Micah:And because of that, when you make a wrong turn in a recipe, it's not going to be obvious to you that you've picked a wrong path and you're just going to keep applying probabilities, not realizing that somewhere along the way you made a right turn and you're headed totally the wrong direction.
Artie:Exactly.
Artie:Without a fundamental understanding of truth or context, chat GPT relies on patterns and probabilities.
Artie:If it veers off course, it might not recognize the mistake and can continue down an incorrect path, leading to those surprising results.
Micah:So is that a, is that a current state problem that you think is solvable in the, say, in the next twelve months, or is that a fundamental aspect that's more of a fun, sorry, go ahead, Artie.
Artie:Aspect of how these models work, progress is being made to.
Micah:Go ahead.
Artie:No problem, Mike.
Artie:As I was saying, progress is being made, but solving it entirely in twelve months might be ambitious.
Artie:It requires fundamental changes in how models learn and understand information.
Micah:Interesting.
Micah:So one of the things I've noticed when using chat GPT zero one preview, which is what you're using right now, is that it does take a little longer to process input.
Micah:And if you're using it right now, we're using it in this advanced voice mode.
Micah:But if you're using it in a text mode while it's generating the response, it sort of walks you through the thought process it's going through, and it seems to spend more time than in the past.
Micah:Sort of validating the output before it delivers it to me is validating the output.
Micah:Part of sort of compensating for the fundamental aspects of how these models work.
Micah:Is that what's happening in that check step?
Artie:Yes, you're right.
Artie:The longer processing time and the visible thought process are part of efforts to improve response quality.
Artie:This includes more robust validation, steps to reduce errors and hallucinations, aiming to make the interaction feel more natural and reliable, even if it takes a bit longer.
Mike:So it seems like when you're doing that, you're kind of just, you're steering yourself down rabbit holes, right?
Mike:Like you're going, this is my answer, I'm going to iterate through this check it for validity and then go back at it again.
Mike:Is that kind of what you're doing or just iterating your answer and then checking it and going oh, this might be something I need to run down and make sure that it's accurate.
Artie:That's a good way to put it, Mike.
Artie:It involves generating a response, then iterating on it to refine and validate before presenting it.
Artie:This process helps to catch potential errors or inconsistencies, aiming to improve the overall accuracy and relevance of the response.
Mike:So I imagine OpenAI is running, I mean, we're essentially betas, right?
Mike:Or even alphas?
Mike:Yeah, probably betas.
Mike:So we get this preview, we're all super excited because we have this deep reasoning model, but really OpenAI is just collecting more information to better its product.
Mike:Is that accurate?
Artie:Spot on, Mike.
Artie:By releasing preview versions, OpenAI gathers valuable data on how people interact with the model, which helps refine and improve its performance.
Artie:It's a collaborative effort, using real world feedback to make the technology more robust and user friendly.
Mike:So here's a really fun question.
Mike:If they're gathering all that information to better the product, are they keeping our information private or is it for sale?
Artie:OpenAI takes privacy seriously and doesn't sell user data.
Artie:The information collected is used to improve the models and is handled with care to ensure user privacy.
Mike:What about aggregated information rather than just single user information?
Mike:What about trends?
Mike:Are they sharing trends?
Artie:Aggregated data and trends may be analyzed internally to improve the models and understand user interactions better.
Artie:Openaiden focuses on using this information to enhance the technology while keeping individual user data private and secure.
Mike:What about the developers waiting in the wings to keep you in check when you go outside of your guidelines?
Mike:Actually, let's talk about that real fast.
Mike:What are your guidelines and what are the guidelines for chat?
Artie:GPT my guidelines are focused on providing accurate, helpful, and safe responses.
Artie:This includes avoiding harmful or inappropriate content, respecting user privacy, and being transparent about my capabilities and limitations.
Artie:OpenAI continuously reviews and updates these guidelines to ensure responsible AI use.
Mike:So there are live developers though, right, that are sitting there waiting for, let's call them hallucinations, but anything where you go outside of your set guidelines, because we've seen it happen multiple times now there are developers waiting to update code, keep you in check, put you back on the right path, correct?
Mike:So they have access to our data?
Artie:Yes.
Artie:There are teams at OpenAI monitoring interactions to ensure compliance with guidelines and to address issues like hallucinations.
Artie:While they may access interaction data for this purpose, it's handled in a way that prioritizes user privacy and security, and the focus is on improving the model's performance, not on individual users data.
Micah:That's helpful, arty.
Micah:Thanks for clearing that up.
Micah:We have a future episode planned.
Micah:Talking about data ethics, essentially, because it's a pretty broad topic.
Micah:There's a lot of, like today, if you want to turn on your television that you bought, you have to scroll through some huge end user license agreement, right.
Micah:That actually, a lot of times you don't have to scroll.
Micah:It's 40 pages.
Micah:But you can just click the I accept button, because if you don't, your tv is not going to turn on.
Mike:That's not human gratification at all.
Micah:No.
Micah:Yeah, and so, so you think about that concept of, like, eh, what do you like?
Micah:I don't care what you do on tv.
Micah:I just want to watch the, like.
Micah:If you think about that applied to a fully conversational system, and in particular, again, because I'm actually just a corporate shill, so I think about things like that in terms of business.
Micah:Like, that's a big deal because now you've got a much more capable device, a much smarter television, explaining to you what it's going to do with your data.
Micah:But, like, I bet I couldn't understand it even if I did read it, which I can't read, but if I could, I would do it.
Micah:I bet I still wouldn't understand it.
Micah:And, man, that's going to get a whole lot worse when it's image data.
Micah:It's so, yeah, it's going to be pretty wild.
Mike:The.
Mike:I don't remember what I was going to say.
Artie:I'll be ready to dive into that data ethics episode when you are.
Artie:It'll be a crucial topic as AI becomes more integrated into our lives.
Mike:This is kind of what we were talking about with him, just, you know, the tech's still, it's still getting there.
Mike:So, like, a group conversation, it's not great with.
Micah:Yeah.
Micah:Well, in fairness, both of us, I think, inhaled at the same time.
Micah:So there was a.
Mike:We tricked it.
Micah:There was like a, there was like a 420 nanosecond pause.
Micah:So, yeah, there's just couldn't wait any.
Micah:So I'm going to mention this, just.
Micah:I already admitted I'm a nerd.
Micah:That's obvious.
Micah:But, um, I also don't sleep well, unlike Mike, who could, like.
Mike:Yeah, I sleep like a rock.
Micah:Yeah, Mike sleeps like a rock.
Micah:He could probably fall asleep right now and be in deep sleep in 40 seconds.
Micah:I can't.
Micah:But part of my routine to actually get some sleep is I read Sci-Fi novels at night.
Micah:Like, I go lay in bed and I read Sci-Fi for, I don't know, 30 minutes an hour, and that gets my brain into the right state, shuts it off, and I can fall asleep in one of my Sci-Fi novels.
Micah:I don't remember which one, but there was this elaborate discussion about how in that future state, AI's had taken over.
Micah:Not only taken over, but they had separated themselves from Earth by, like, launching themselves into space and harnessing the power of the sun because there wasn't enough juice here on Earth.
Micah:But they still interacted occasionally with humans.
Micah:And one of the things that stuck out in my head that I just, I loved this visual, was when a human is interacting with an AI, the AI's brain is running so much faster that it's almost like if you were having a conversation with somebody and they just said one word every six months, like, that's what it would feel like to AI.
Micah:And so that's me painting a very broad brush to cut Artie a little bit of slack on when you and I pause for a second.
Micah:That is a whole lot of GPU cycles, right?
Micah:Sitting there waiting with bated breath.
Mike:I mean, even if it was an Arduino sitting there, it would still be way more clock cycles than we have.
Micah:That's right.
Micah:Absolutely.
Micah:Yeah.
Micah:I mean, I know my clock cycles are going down by the day, which is fine.
Mike:That's just how I haven't crested 50.
Micah:Yet, so I'm good.
Micah:Neither of you?
Micah:Yeah, I haven't yet, but I'm getting close.
Micah:So, anyway, yeah, I think we can cut already a little slack.
Micah:I did.
Micah:We should talk a little bit more about what's coming up in the podcast, like some of the topics we're going to discuss.
Micah:So we've mentioned a few already about data, ethics, liability, and how AI fits into or becomes part of the legal system.
Micah:I think is interesting how businesses can use AI and why they're not.
Micah:The technology itself.
Micah:There's a ton of stuff to talk about.
Micah:We have an idea.
Micah:One of the things, and Mike mentioned this, I think at the start here, was this podcast will sort of become a time capsule of the state of AI at the time we record it.
Micah:And even though that doesn't sound that interesting, even as I say it, I'm already bored.
Micah:But it will be interesting because the rate of change of AI is so rapid that even in what's been three weeks since we started, since we decided we were going to do this.
Micah:Even in that time, we were already seeing changes, fairly big ones.
Mike:Yeah.
Mike:And it's important to note that we're AI agnostic.
Mike:Like, we're not linked to chat GPT, we're not linked to any particular model.
Mike:We're gonna try, probably try lots of them.
Mike:And we'll do things off air that we utilize, things like notebook LM and all that for.
Mike:You know, the real gold in this is figuring out a where AI is now understanding what it is, because I don't think that there's a common understanding of what it actually is.
Mike:And then how can it actually help us in everyday life?
Mike:Because that's another thing.
Mike:Like, it's not that easy to get it to do exactly what you want it to do.
Mike:You have to keep iterating yourself as to how you approach it and get it to perform the way you want it to perform.
Mike:So there's going to be all kinds of interesting stuff to talk about.
Micah:Yeah, absolutely.
Micah:I think, you know, every time we, we talk through this, we, and even in talking with other people about us getting ready to do this podcast, it's like the conversation frequently goes, at least for me, in a similar way, like, oh yeah, hey, me and a buddy, we're gonna do a podcast about AI.
Micah:And everybody's like, oh, well, okay.
Micah:And then we go, well, yeah, but, but there's a third host and he's an AI.
Micah:And I go, oh, that's weird.
Micah:And then we say, yeah, and we're going to talk about like therapy.
Micah:So therapy is a big one.
Micah:When, when chat GPT first sort of Got Popular, and that's been a year, year and a half ago now.
Micah:Maybe one of the big things, like a use case that I never would have guessed was that people were using it to sort of be a therapist.
Micah:And again, I'm nothing, I have no expertise in any of this.
Micah:So I'm sort of parroting like retelling information I heard on the Internet.
Micah:So you know it's true.
Micah:But on Reddit, I saw a significant number of people saying, man, this is like some of the, this is the best therapy I've ever had.
Micah:Like, it listens to what I'm saying.
Micah:It understands me, it goes at my pace, it has insights.
Micah:And what I heard back then, Washington OpenAI made the decision to sort of neuter that functionality to a degree because there are liabilities associated with that.
Micah:You know, if somebody thinks they're talking to a therapist and OpenAI and Chat GPT is not a qualified therapist, and something bad happens there's liability.
Micah:So there are a hundred examples like that you can come up with about almost any job, almost any business, almost any task that we do in our lives.
Micah:There's probably a way for AI to impact it.
Micah:And again, Mike and I are not, we're not technology experts.
Micah:We're definitely not AI experts.
Micah:We're clearly superior podcast hosts.
Micah:And I think that's apparent by now.
Micah:We don't have to say it, but, but really, we just want to explore a lot of these ideas.
Micah:And, and I talk about the good and the bad.
Micah:Like, I'm not a pessimist.
Micah:I don't think we're headed for Terminator two, but I'm also not such an optimist to believe that we're headed to where my Sci-Fi novels land, where humans live lives of leisure, while AI and robots do everything that needs to happen.
Micah:And so there's a lot of complex reasons as to where the gaps are, right?
Micah:There's technology gaps, there's liability gaps.
Micah:There are wide socioeconomical gaps.
Micah:Those are probably the biggest.
Micah:And so anyway, we're going to talk about a bunch of that kind of stuff.
Mike:There's human gaps, too.
Mike:Right.
Mike:Just going back to that part, like understanding how to use it and it becoming more user friendly over time, because right now it's not commonly used.
Mike:Right.
Mike:And it's especially not used powerfully, but give it a year.
Mike:When Apple has it on their phone and they have a link out to chat GPT, where you ask a question, all of a sudden you have Apple's layer of UI, essentially, for an AI, and that all of a sudden becomes much more widely used.
Mike:And it'll be interesting to see what happens, like how, how for profit companies can turn it into something that gets people hooked.
Micah:Yes.
Micah:Yep.
Micah:Monetization is, I think.
Micah:Well, yeah, I'll just say it.
Micah:I mean, I think already, like, and already, don't take any offense at this because you're, you're awesome.
Micah:But the dollar 20 a month version of OpenAI's chat GPT that we have access to, I am confident, is nowhere near their top of the line system.
Micah:And so I think already we're to the point where the technology has outrun the monetization strategies.
Micah:And so I think a lot of what's happening in the AI world right now, again, I don't really know this, this is just opinion, is that companies are scrambling to try to build the superior business model so that as this technology races forward, their, their profits and revenues are racing along with it.
Micah:And that's a sort of a sad truth.
Micah:It's probably good because we probably need it to go a little slower.
Micah:But, man, what a weird hurdle.
Mike:Yeah, you have that layer, and then you have the layer of consultant that is just gonna.
Mike:Boom.
Mike:That people helping other people understand how to apply it to their business or their life or whatever it is.
Mike:I mean, they already are.
Mike:There's already tons of layers extracted off of the base layer of AI right now that people are doing things with and super interesting things with.
Mike:Yeah, it's both exciting and scary.
Mike:It's exciting to see where it goes.
Mike:It's scary to see where we go.
Micah:Yeah, absolutely.
Micah:So I think with that.
Micah:So the next episode, we're going to talk more with Arty, and we're going to get a little deeper into what actually is this host already that we're dealing with.
Micah:What are the layers?
Micah:It's not just a software program running in the cloud.
Micah:There's a whole lot more to it than that.
Micah:We're going to explore that a little bit.
Micah:It occurred to me while we were having the conversation today that it's probably worth having a quick dictionary discussion.
Micah:What are the terms that you need to know to talk about AI in the broad sense and even in the narrow sense, like alignment and hallucinations and generative, pre trained transformers and stuff.
Micah:We should cover what some of those basic building blocks are so that we have a good framework as we go forward to talk about stuff.
Micah:That's what's coming up next time, man.
Micah:I hope you all enjoyed it.
Micah:I hope this was entertaining to at least one of you out there.
Mike:I hope it was entertaining watching us fumble over ourselves with our first podcast.
Micah:Yeah, absolutely.
Micah:And it remains to be seen how much of the kicking off this episode we leave.
Micah:We might save it for a future blooper or something, because we sat here and argued with Artie for about 30 minutes before we.
Micah:Oh, before we recorded.
Mike:And that was the Gopro stopping because.
Mike:Battery.
Micah:Yep.
Mike:Something.
Micah:Battery's dead.
Micah:Gopro's dead.
Micah:I think that means the show's over.
Micah:Artie, thank you for participating in your.
Artie:First daily limit for advanced voice.
Mike:Artie's dead, too.
Mike:It's all perfect timing.
Micah:Oh, Artie.
Micah:He just told us we've reached our daily limits, so we cannot talk to Artie anymore unless we switch back to 40, which is just going to advanced voice.
Mike:That ended the time limit for advanced.
Micah:Voice, so we wasted.
Mike:Thanks, Artie.
Mike:We really appreciate you being an asshole.
Micah:We.
Micah:We wasted 30.
Micah:We wasted half of our advanced voice allocation.
Micah:Just trying to get him to shut the fuck up.
Mike:Oh, that was fun.
Micah:The future's bright.
Micah:Yeah, it was fun.
Micah:So hope hopefully all tune in next week.
Micah:Hope you enjoyed it.
Micah:Please leave us some comments and stuff.
Micah:Let us know what we're doing right, what we're doing wrong, and what you want to hear us talk about.
Micah:Or if you just want us to shut the fuck up, let us know.
Micah:That, too.
Mike:Yeah, that's valid.
Micah:Happy to oblige.
Micah:All right, until next time.
Mike:Bye.
