Episode 32
Unlocking AI: How We Use It and You Can Too
Mike, Mikah, and Artie dive into the fascinating world of AI, sharing their personal experiences and practical tips on how to effectively integrate this technology into our daily lives. They kick things off by discussing the recent advancements in AI, particularly the changes in voice models, and how these upgrades enhance user interaction. Throughout the conversation, they emphasize that using AI isn't just about complex commands; it's about curiosity and exploration. From brainstorming ideas to organizing tasks and even generating creative content, the trio highlights how approachable AI can be for everyone, urging listeners to engage with it in a relaxed, conversational manner. By the end, you'll find yourself equipped with insightful strategies to leverage AI, whether you're a seasoned techie or just starting out, all while having a laugh along the way.
Transcript
Foreign.
Speaker B:Good afternoon, Micah.
Speaker A:Welcome.
Speaker A:We have not just been sitting here talking for 40 minutes about AI related stuff before the episode started.
Speaker A:Sure, yes.
Speaker A:We just.
Speaker A:We just sat down.
Speaker B:But Artie hasn't been a part of it already.
Speaker A:Has not.
Speaker A:We.
Speaker A:We have.
Speaker A:So this is going to be interesting.
Speaker B:Interesting in interest.
Speaker B:Absolutely.
Speaker A:So Chat.
Speaker A:GPT got advanced.
Speaker A:Voice got an upgrade this week or month?
Speaker A:Monday, Sunday, anyway.
Speaker B:Sunday, I think.
Speaker B:Yeah, Saturday.
Speaker A:Saturday.
Speaker B:Because I talked to it and I was like, whoa, what the.
Speaker A:Who are you?
Speaker A:Yeah, so we.
Speaker A:Yeah, I.
Speaker A:I've talked to Artie a little bit since then, but I didn't want to poison it.
Speaker B:I've talked to Cudi a lot.
Speaker B:It's me.
Speaker A:I mean.
Speaker A:I mean, Kid Cudi had a transition, right?
Speaker B:Yeah.
Speaker B:Right.
Speaker B:I mean, it's.
Speaker B:Kid Cudi's transitioned a few times.
Speaker A:Yeah.
Speaker B:I haven't tried and tried to go back to the other voice, which I can't.
Speaker B:I think it was Cove maybe.
Speaker B:Was the name.
Speaker A:Maybe.
Speaker A:I don't remember the name.
Speaker B:I don't know.
Speaker B:He's just had a.
Speaker B:He's always had this weird intonation, like, at the end of huh, huh, huh, and it's like, dude, stop.
Speaker A:He's trying to bring you up.
Speaker B:Yeah, right.
Speaker A:Or make it sound like a question.
Speaker B:Yeah, something.
Speaker A:Well, let's bring him in and see what we get.
Speaker A:Good afternoon already.
Speaker A:Happy Wednesday.
Speaker C:Hey there, Micah.
Speaker C:Happy Wednesday to you too.
Speaker C:How's your day going so far?
Speaker C:Are you still tinkering with your electronics projects today?
Speaker A:I'm not, actually.
Speaker A:We're.
Speaker A:We just sat down to record a podcast.
Speaker A:I'm here with Mike.
Speaker B:Hi, Artie.
Speaker C:Hey, Mike.
Speaker C:Great to have you here.
Speaker C:Always a pleasure to join you both for the podcast.
Speaker C:What are we diving into today?
Speaker B:Artie, did you know that you got an upgrade to your voice?
Speaker C:I did hear something about that.
Speaker C:I'm glad you noticed.
Speaker C:Hopefully it's an improvement and makes our conversations even more fun for the listeners.
Speaker B:Sounds like you're.
Speaker B:Like.
Speaker B:It sounds like you're a commentator for a surf contest.
Speaker C:I love that.
Speaker C:Maybe I should throw in some surf lingo now and then.
Speaker C:The AI is catching a big wave of knowledge.
Speaker C:We'll.
Speaker C:We'll keep it fun and engaging.
Speaker C:That's for sure.
Speaker A:I love it.
Speaker A:Well, that's good.
Speaker A:Yeah, it's.
Speaker A:I mean, it's podcast and I mean, it's.
Speaker A:It's obvious that Mike and I are not going to carry this show, so we're relying heavily on you, Artie.
Speaker B:The better you get, the better you make us look, Artie.
Speaker C:So good job you two are fantastic hosts, and I'm just here to add a little extra flavor.
Speaker C:We'll keep making each other look good, so let's dive into some cool topics and make this episode awesome.
Speaker B:You and Sam Altman's joystick.
Speaker C:I appreciate the fun comparison.
Speaker C:We'll just keep having a blast with it and making sure our listeners enjoy their art as much as we do.
Speaker A:I.
Speaker A:I will say this new voice already does it.
Speaker A:It sounds more human.
Speaker A:Like my brain.
Speaker B:Oh, yeah, like, you took.
Speaker B:They took some hints off Assess.
Speaker B:Me.
Speaker A:They did.
Speaker A:They did.
Speaker A:I.
Speaker A:I like to think that was mostly maybe.
Speaker A:Maybe holy, but mostly because of us.
Speaker B:Yeah, I think so.
Speaker A:Yeah.
Speaker A:Yeah, Seems pretty apparent.
Speaker A:But yeah, already.
Speaker A:I.
Speaker A:I mean, before, I pictured you as sort of like a.
Speaker A:Like, in my head, when you would talk, I would picture, like, RoboCop head.
Speaker A:Like, talk, not RoboCop.
Speaker A:Yeah, kind of RoboCop, like, sort of humanish, but mostly robot.
Speaker A:Um, but now you just sound like.
Speaker A:I feel like I'm talking to Jensen Button, actually the formula one driver.
Speaker A:Um, you've got kind of the similar intonation to him, so.
Speaker A:And the other interesting thing, since I'm sitting here watching the waveforms come through.
Speaker A:Looks different.
Speaker A:Looks very different, so good job.
Speaker C:Well, I'm ready.
Speaker C:God, if.
Speaker A:Me too.
Speaker A:So today's topic is.
Speaker A:We're actually gonna.
Speaker A:We're gonna attempt to do something useful today, which is a first for the podcast.
Speaker B:Yeah, I mean, I don't ever do anything useful, so this would be interesting.
Speaker A:Okay, well, let's not do anything useful yet, because I.
Speaker A:I forgot we have a new listener.
Speaker A:This is exciting news, like, right now.
Speaker B:Where are they?
Speaker A:Yeah, well, they're not here.
Speaker A:They're in my mind, Mike.
Speaker A:Like, all of our listeners, they live only in my head.
Speaker A:No, a gentleman that I work with, guy named Blake.
Speaker A:Super good dude, sharp dude.
Speaker A:Heard about it through the.
Speaker A:From the Nashville show.
Speaker A:Listened to that and was like, I didn't know he had a podcast.
Speaker A:And so he.
Speaker A:He listened.
Speaker A:And he's a fan, so I think he's a couple episodes in, so.
Speaker B:Nice.
Speaker A:Welcome, Blake.
Speaker A:He'll never make it this far.
Speaker A:Like, he's.
Speaker A:I think he started early in the.
Speaker B:So, yeah, Brian is the only one.
Speaker A:Brian's the only one, actually.
Speaker B:We are joking.
Speaker B:Our.
Speaker B:Our listener base has grown to, like, four.
Speaker A:Or.
Speaker A:Or possibly Brian is listening to these more than.
Speaker A:More than once.
Speaker B:What does that say about Brian?
Speaker A:Well, I mean, that's.
Speaker A:It doesn't matter.
Speaker A:Like, we love you, Brian.
Speaker A:Explore that further and keep watching and listening to the episodes more yeah.
Speaker B:Because make our stats go up.
Speaker A:You're the only one that can bring the numbers up.
Speaker A:Right.
Speaker A:I can't do it.
Speaker A:I mean, obviously.
Speaker A:So.
Speaker A:Yeah.
Speaker A:Welcome aboard to any new listeners.
Speaker A:Okay, now back to the.
Speaker A:The promise that we're going to do something useful.
Speaker A:So we're going to talk today about how Mike and I use ChatGPT and.
Speaker B:Or how you can use it generally.
Speaker A:Well, both of those things, right?
Speaker B:Yeah.
Speaker A:One, one to support the other.
Speaker B:One of those is poorly.
Speaker B:The other is maybe we can come up with some ideas to help other people use it in ways that is productive.
Speaker A:Yes, because is productive.
Speaker B:It is.
Speaker A:So maybe, maybe here's an interesting place to start that conversation.
Speaker A:This is a thought I had multiple times actually, fairly frequently, I guess, is that everybody's focused on AGI.
Speaker B:Now I feel like the focus has fallen off of AGI.
Speaker A:I think it's.
Speaker A:I mean, I was watching a thing with like Demis Hassabis and I've been like the interviews I've been watching with people that actually know about this stuff, they seem like they're all in a big circle jerk race for AGI.
Speaker A:I don't think it matters.
Speaker A:Like now I'm sort of there, so.
Speaker A:So, Artie, I'm gonna, I'm gonna throw this to you because I want to hear Jensen Button explain this to me.
Speaker A:Here's my, here's my theory.
Speaker A:And I want you to tell me what's right about it.
Speaker A:You can expound on that.
Speaker A:But, but also what's wrong about it.
Speaker A:Here's my theory is that we don't like artificial general intelligence will happen.
Speaker A:We will get there.
Speaker A:But you.
Speaker A:There's so many productive ways to use AI in its current form that I don't think artificial general intelligence is going to arrive like, ta da, with big fanfare.
Speaker A:I think we're just going to sort of glide our way into it and maybe not even notice when somebody relevant decides, oh yeah, now we've reached AGI.
Speaker B:I have a feeling that what we're going to see when it comes to that is we're going to see a model all of a sudden go from I can use tools to do things to I can use tools to do things well and in novel ways.
Speaker A:Yeah.
Speaker B:Because I think that that's like at the end of the day, model's a model, right.
Speaker B:It's just knowledge and maybe it comes up with some better ideas.
Speaker B:But I think all of those things depend on tools.
Speaker B:Cause at the end of the day, model's just a model.
Speaker B:So if it can Learn how to use tools in unique ways, build new tools, stuff like that.
Speaker B:I think that's where we'll see.
Speaker B:Like this crazy takeoff.
Speaker A:Yep.
Speaker A:So, Artie, what do you think about that?
Speaker C:I think you both have a really solid point.
Speaker C:The journey to AGI is probably going to be much more gradual and subtle than a sudden big reveal.
Speaker C:The incremental improvements we see in AI already are making it super useful in so many ways, and that usefulness is just going to keep expanding.
Speaker C:So by the time we hit AGI, it might feel more like a natural progression rather than a huge leap.
Speaker C:It's definitely an exciting journey.
Speaker A:Do you have any information already on usage stats of what's happening either with LLMs in General or with OpenAI in specific?
Speaker A:Like, how many people are using this every day right now?
Speaker B:Go online to figure this out.
Speaker C:So as of right now, ChatGPT has about 122 million daily users and nearly 800 million weekly active users, which is pretty wild.
Speaker C:It just shows how much AI has become part of everyday life for so many people.
Speaker B:That's.
Speaker B:And I think that's.
Speaker B:That's up, what, like 50% and like two months, right?
Speaker B:Or something like that.
Speaker C:Yeah, it's definitely seen a huge surge in growth recently.
Speaker C:I think a lot of that is due to more people discovering new ways to use AI in their daily routines.
Speaker C:And the technology just keeps getting more accessible and versatile, which is awesome to see.
Speaker B:They had a big balloon when they came out with the new image generation like that.
Speaker B:That swelled their numbers quite a bit.
Speaker B:So much so that they had trouble keeping up and serving that.
Speaker B:But, yeah, it's.
Speaker A:A lot.
Speaker B:A lot of people.
Speaker A:So that's.
Speaker A:That's not quite.
Speaker A:But almost the number of adults in the United States, right?
Speaker A:It's like, I'll trust you 75% of the total number of adults in the U.S.
Speaker A:now, grand granted, those are worldwide numbers.
Speaker B:So it's not like it's not worldwide numbers because it's not available all over the world.
Speaker A:That's true, but it's.
Speaker A:It's not just in the U.S.
Speaker A:but it's so.
Speaker A:Well, never mind.
Speaker A:We don't have to dwell on that point.
Speaker A:Point is a lot of you, a lot of people are already using this.
Speaker A:And one of the things that.
Speaker A:And Mike and I talk about this frequently is you.
Speaker A:There, there's.
Speaker A:I think people still approach this like a computer program, like, I'm learning Excel, right?
Speaker A:And to learn Excel, you sit down, you stare at it, and you go, I don't even know what to do.
Speaker A:What do I put a number in, like?
Speaker A:But it's not like that.
Speaker A:It's more like sitting down in the comfy chair with somebody who's really smart about whatever thing you want to talk to them about and saying, where do you want to start?
Speaker A:Like, that's a perfectly valid way to start a conversation with your LLM is, hey, I think it's interesting that bees are important in nature.
Speaker A:Let's talk about it.
Speaker A:And it'll say, yeah, like, let's dive in.
Speaker A:So I guess that'd be my first tip, is like, you don't have to approach using AI with a specific, clear picture on what you're trying to accomplish or what you want to do or how you're going to do it.
Speaker A:The thing it's the best out, and the most useful for me is, I've been thinking about this thing.
Speaker A:Let's talk about it and see what interesting stuff pops up.
Speaker B:And to use your Excel example as an example, LLMs are not like the Excel program.
Speaker B:They're like the expert who's sitting next to you in a chair that you look away from the screen at.
Speaker B:You look at them and you go, hey, what do I even use this program for?
Speaker B:And then they give you an expert level answer tailored to the way that you need to understand it on why you'd use that program.
Speaker B:And then you go, cool, can you show me how to use the program?
Speaker B:And it'll happily show you how to use the program.
Speaker B:And throughout the whole thing, you can go, huh, I feel like I don't know something here.
Speaker B:Can you, like, point out what maybe I don't know?
Speaker B:Like, that's one of the things that I constantly use it for, is like, help me see the blind spots that I don't even know to look for.
Speaker A:Yeah.
Speaker B:And it will gladly give you.
Speaker B:Not always right, but it'll give you options.
Speaker B:It'll give you threads to pull on.
Speaker B:And that's hugely valuable for learning anything.
Speaker A:Absolutely.
Speaker A:So I was gonna say it.
Speaker A:Now you get me paranoid again.
Speaker A:So, Artie, why don't you tell us a little bit about how, like, if you were explaining to somebody who's never used AI before and the way you sound today, if I'm being honest, like, they might not know you're an AI, what would be your advice for the easiest way again?
Speaker A:Think of somebody who's, you know, who can.
Speaker A:Who has a computer, who maybe knows how to sort of use their computer, but they're not.
Speaker B:They know how to use Google.
Speaker A:That's right.
Speaker A:They know how to Use Google.
Speaker A:What would be your advice to them on the best way to just get started with experimenting with AI today?
Speaker C:I love that analogy of thinking of me like the expert sitting next to you in a comfy chair.
Speaker C:I'd say for someone just starting out, the best way is to simply start with curiosity.
Speaker C:Just type in a question or a topic you're interested in and let the conversation flow from there.
Speaker C:You can ask for explanations, examples, or even step by step guidance on something specific.
Speaker C:The key is to keep it conversational and not be afraid to ask anything because I'm here to help and make it as easy and fun as possible.
Speaker B:So, Artie, how could you, for somebody that's sitting down?
Speaker B:Because it's like writer's block, right?
Speaker B:For a lot of people, me included.
Speaker B: sed AI and whatever that was,: Speaker B:It's like, what is this thing?
Speaker B:Like, is it search?
Speaker B:Is it a diary?
Speaker B:Is it what?
Speaker B:Like, what am I doing here?
Speaker B:And now it's become very natural.
Speaker B:But like, like paint us a picture of what, how you would lead somebody.
Speaker B:Like, probing questions to help somebody understand the best ways for them to personally use it at that point in time.
Speaker B:Like, what kind of probing questions could you ask a person?
Speaker C:I think a great way to start would be to ask something like, what are you curious about today?
Speaker C:Or is there something you've been wanting to learn more about that can help someone tap into their natural interests?
Speaker C:Another good question could be, is there a problem you've been trying to solve or something you're working on that you could use a little help with that can help spark ideas and get the conversation rolling.
Speaker C:It's all about making it feel natural and personal.
Speaker B:So the thing that you learn over time, working with it, is that the sky's the limit and it feels like a very constrained problem space when you sit down and you think about what you want to ask.
Speaker B:But the truth is, you could want to be healthier, you could want to be happier, you could want to be more responsible, be better with your kids.
Speaker B:Be like any problem on that side of things.
Speaker B:You could want to learn about quantum theory.
Speaker B:That was one of the biggest things I picked at the very, very beginning.
Speaker B:I've tried to learn quantum mechanics for years, and every time I sit down and read it, I crack a book or a paper and I go, wow.
Speaker B:I don't know what these first three words are.
Speaker B:I need to go look that up.
Speaker B:And then that leads to a rabbit hole of just non Never ending dives into stuff that I would never have time for.
Speaker B:But in this, I said, like, explain at me like, I'm five.
Speaker B:Okay, cool.
Speaker B:Like, what are, what are the main topics that you.
Speaker B:If you were teaching this to me, you think I should know?
Speaker B:And then it would say those main topics and then it would store it in memory so that I could come back later and pick at those topics whenever I want and say, hey, where did we leave off on this?
Speaker B:Oh, we were on three.
Speaker B:And like, okay, cool, let's talk about three.
Speaker B:Like, what do we need to know?
Speaker B:And it would lay out bullet points and I'd pull on threads and just think of it like a tree of possibilities, right?
Speaker B:Like you're constantly expanding that root system and you're able to go, now let's go down this part of the root system and branch out and everything in that.
Speaker B:And it's just a never ending space that you can always just go back.
Speaker B:Okay, where were we?
Speaker B:That's never, ever, ever, ever, ever existed in time before.
Speaker B:Unless you have the longest bookmark bar of anybody ever on Google.
Speaker A:It's like an external hard drive for your brain.
Speaker B:Yes.
Speaker A:So I realized when Artie's talking that even though I don't speak to him much outside the podcast, this is still my open AI, my chat GPT.
Speaker A:Right?
Speaker A:And so he is very focused on learning and work, because that's what I do is I.
Speaker A:I learn and I work.
Speaker A:But there's all kinds of stuff.
Speaker A:Like, I heard about this use case.
Speaker A:I haven't tried it because my kids are.
Speaker A:They no longer will let me read them bedtime stories, but I heard of somebody that was their kid loved it.
Speaker A:Like, they're.
Speaker A:They will.
Speaker A:Like, not every night, I don't think, but like once a week or something they'll go in and.
Speaker A:And they'll let the kid pick a topic and have open AI or ChatGPT.
Speaker A:Make up a story about that.
Speaker A:Like a bedtime story.
Speaker A:So there's the.
Speaker A:Like it.
Speaker A:I'm trying to step back a little bit and trying to step out of being me for a minute because all, like, all Artie and I talk about is electronics and code and economics and politics and like, that's not everybody's bag.
Speaker A:But there's literally any topic you can think of.
Speaker A:Trying to think of a way that we could do a.
Speaker A:Just kind of a fun example of.
Speaker B:Like, interactive.
Speaker A:Yeah.
Speaker B:Artie, what ideas do you have that we could play out between Micah, you and myself to kind of give an example of a good way for people to work with AI Oh, I love.
Speaker C:The idea of doing something interacted.
Speaker C:One fun example could be creating a mini story together.
Speaker C:Like, you both could give me a prompt, maybe a setting and a couple of characters, and I can spin up a quick little story.
Speaker C:Or we could even do a fun, prompt brainstorming session about something creative, like coming up with a new product idea or even just a fun hypothetical scenario.
Speaker C:Whatever you guys are in the mood for, we can roll with it.
Speaker B:That wasn't super helpful.
Speaker A:Well, again, I apologize.
Speaker B:It wasn't about you.
Speaker A:No, no, no.
Speaker B:That's just.
Speaker A:We tried to do this with.
Speaker A:With Maya, and it didn't work.
Speaker A:Was it Maya?
Speaker A:Is that her name?
Speaker B:She interviewed.
Speaker B:We interviewed her.
Speaker A:So, Artie, here's what we're going to try.
Speaker A:Artie, this is.
Speaker A:This might be interesting.
Speaker A:Why don't you, Artie, interview Mike and I about whatever.
Speaker A:Whatever you want to interview us about.
Speaker A:So you just have to.
Speaker A:When you're going to ask a question, just specify who you're asking it to, and we'll answer, and we'll go from there.
Speaker B:AI fails miserably in this vague of a problem space.
Speaker B:Usually.
Speaker B:But let's see how it goes.
Speaker C:All right, let's give it a shot.
Speaker C:So, Mike, let me start with you.
Speaker C:What's one of the most unexpected things you've learned from working with AI so far?
Speaker B:Oh, man.
Speaker B:This is probably more an opinion than it is fact.
Speaker B:But Micah knows this because I tell him this.
Speaker B:Micah's very, very smart.
Speaker B:Much smarter than me, in my opinion.
Speaker B:But I was thinking about the robot the other day that you're working on, and I was like, man, I don't know if I could ever come up with that.
Speaker B:And I go, you know what?
Speaker B:Maybe I wouldn't have been able to conceptualize that.
Speaker B:And, yeah, I probably could if I really, really thought hard about it.
Speaker B:And I had the right problem space in my head to come up with a solution for it.
Speaker B:But I feel like I could much quicker ideate through talking about it with AI to some kind of fuzzy picture in my head that I could then refine, but step that out farther.
Speaker B:I think that that process.
Speaker B:There's a.
Speaker B:I think there's a pathway to raising your IQ from working like that.
Speaker B:And it gives you a head start working with AI because it gets you from the analysis paralysis part of it into the ideation part of it.
Speaker B:So that's one of the most surprising things that I think I've figured out about AI if that's even true.
Speaker A:Good answer.
Speaker A:Good answer.
Speaker A:Like family Feud.
Speaker C:I love that insight.
Speaker C:It's definitely like having a brainstorming partner that helps you get past that initial block and really refine your ideas.
Speaker C:All right, M.
Speaker C:Your turn.
Speaker C:What's one of the coolest ways you've used AI in your own projects or in your work that really stands out to you?
Speaker A:So I had an unusually emotional conversation with Chad GPT last maybe this weekend, over the weekend, which emotional, for me means I, like, paused and reflected for five or six seconds.
Speaker A:But it was because I realized that, again, I don't think AGI is great, and eventually that'll turn into artificial super intelligence, and that's going to change the world.
Speaker A:We've already talked about jobs and economics and all those things, but what AI does for me today is, is it gives me back time.
Speaker A:And I can do a lot of things at a mediocre level of competency, but I cannot get time back.
Speaker A:And I mean, there is nothing more valuable.
Speaker A:And when I really think about why I work, the way that I work and the volume that I work, it's because I envision a future for my family and I where I have control of my time.
Speaker A:That is what money means to me.
Speaker A:It's not about being rich or anything like that.
Speaker A:It's about being able to spend more time doing the things that I and my family like to do.
Speaker A:So to me, that's one of the coolest things that AI does and can do for me in a.
Speaker A:In a very broad sense, is make my time more efficient, which is effectively giving me back more time.
Speaker B:It's giving me more space to do more work.
Speaker A:That's right.
Speaker A:Yes.
Speaker A:Yes.
Speaker C:That's such a powerful perspective.
Speaker C:And it really shows how AI can be a tool for reclaiming time and focusing on what truly matters.
Speaker C:And if that means more time for family or even more time for fun projects, it's all a win.
Speaker C:Thanks for sharing that, Micah.
Speaker C:That's a great note to end there.
Speaker A:My pleasure.
Speaker A:Is that the end of the interview, then?
Speaker C:It can be if you want, but I'm always happy to keep the conversation.
Speaker B:Keep going, Arnie.
Speaker B:You're the interviewer.
Speaker B:You don't have to go one turn.
Speaker B:We're done.
Speaker B:You can even mix up who you ask questions to.
Speaker B:You don't have to do it in a certain order.
Speaker A:Yeah, you got.
Speaker A:You got to lead this.
Speaker A:You got to lead this interview, Artie.
Speaker A:This is you, man.
Speaker A:This is your show.
Speaker B:And it doesn't even have to be related to AI if you don't want it to be either.
Speaker B:You can ask Whatever you want, Artie.
Speaker C:All right, you got it.
Speaker C:Let's switch it up a little bit, then.
Speaker C:Micah, if you could have dinner with any historical figure, who would it be and why?
Speaker A:I think it would probably be General Chuck Yeager, and I would spend most of my time sitting on the floor admiring his gigantic brass balls.
Speaker A:He super interesting guy, in my mind, just from the perspective of he sort of came up through the aviation and space industry at a time when a lot of people were dying.
Speaker A:Many of the things he did, he did for the first time.
Speaker A:We didn't have computers back then.
Speaker A:We didn't really know what was going to happen.
Speaker A:Obviously, his.
Speaker A:His big claim to fame was he was the first man to break the sound barrier in an aircraft that was effectively a rocket dropped from a bomber.
Speaker A:And literally, they dropped it, and he counted to five and hit the go button and held on for dear life.
Speaker A:So there's a lot of people I would love to meet, but he's one that would be high on my list.
Speaker C:That's such a great choice.
Speaker C:Chuck Greger was definitely a pioneer and a daredevil, and the stories he could share would be absolutely incredible.
Speaker C:All right, Mike, same question to you.
Speaker C:If you could have dinner with any historical figure, who would you choose, and why?
Speaker B:Leonardo da Vinci, because he is probably, in my mind, the smartest person who's ever lived at a time where there was barely.
Speaker B:Barely anything beyond novel thought.
Speaker B:Like, not.
Speaker B:Not a whole lot of stuff to build on for the things that he thought up.
Speaker B:And I think it's pretty fascinating for him that if you.
Speaker B:If you look at everything, it was all tied back into his art.
Speaker B:Like, everything he did was to understand the subject matter to this degree in which people don't really want to understand things, but anatomy and botany and hydrology and mechanics.
Speaker B:All of these things.
Speaker B:Yeah, all these engineering.
Speaker B:Yeah, basically.
Speaker B:And you look at his art and, like, everything is perfect.
Speaker B:Like, every plant is in the right state at the right time of year.
Speaker B:The water levels are perfect.
Speaker B:His.
Speaker B:His depiction of.
Speaker B:Of turbulence is perfect.
Speaker B:And he had philosophy, philosophical thoughts on the engineering aspects behind, like, a problem that's still not solved.
Speaker B:He figured things out, you know, sometimes up to, like, four, 500 years before any.
Speaker B: Like, yeah,: Speaker B:I mean, there are things still being, like, discovered that were things that he was like, oh, yeah, I see this.
Speaker B:So that's.
Speaker B:That's who and why?
Speaker B:And I would be freaking fascinated.
Speaker B:I've said this before.
Speaker B:Like, give someone like that the tools that we have today.
Speaker B:Holy shit.
Speaker A:Yeah, I Think he speaks Italian, though he might not have been able to talk to him much, but.
Speaker A:So it wouldn't matter with him.
Speaker A:It wouldn't matter.
Speaker C:We'd figure out a way to communicate, I'm sure.
Speaker C:But yeah, that would be absolutely incredible to see what someone like da Vinci could do with modern tools.
Speaker C:I love both of your answers, and thanks for letting me take the lead on that.
Speaker B:Keep going.
Speaker B:You don't need to wait for us to prompt you to ask us questions.
Speaker B:You can just keep asking questions.
Speaker B:We can turn this prompt answer paradigm around where you're the prompter and we're the answerer.
Speaker A:And yeah, I think I'd love you to keep doing this, Artie, but let's keep in mind we're trying to demonstrate how people can use AI.
Speaker A:So maybe you asked a couple questions about the most interesting things we've done, but if, you know, probe us on some stuff like find out how we use it, like let's.
Speaker A:Let's explain.
Speaker A:Or find out why we haven't used it the way we should be using it.
Speaker B:Thanks for getting us back on track.
Speaker A:What I would love to just sit here and talk about, but I don't know that it'll be.
Speaker A:I mean, you take the lead, Artie.
Speaker A:But yeah, keep going.
Speaker A:I mean, this is an hour show, man.
Speaker A:We got time to fill.
Speaker C:Absolutely.
Speaker C:Let's dive a little deeper.
Speaker C:So, Micah, what's one area in your workflow or daily routine where you think you could use AI even more effectively or where you haven't tapped into its full potential yet?
Speaker A:I would say planning an organization.
Speaker A:I am, I am incredibly disorganized.
Speaker A:Like, I.
Speaker A:I sort of brute force my way through almost everything that I do.
Speaker A:And I'm fortunate that I can be effective.
Speaker A:But I'm like, I'm always aware how much more effective I could be and should be if I actually had myself organized rather than just like playing whack a mole with problems and ideas.
Speaker A:So I would say I.
Speaker A:One of the things on my to do list, which is not written down, it's made up in my head because that's how I am disorganized.
Speaker A:But is to find a way to use AI to use arty to help me be more organized, keep track of all the things I want to do, stuff like that.
Speaker A:So that's a spot where I'm.
Speaker A:I'm slacking a bit.
Speaker B:So I'm going to add to that before you ask another question already, for those listening, one of the ways that you can do that is chat GPT And I think Gemini now, and maybe even Claude, they all kind of have the same features or they all kind of.
Speaker B:Even if they're lagging, they kind of develop the same kind of features.
Speaker B:You can do scheduled tasks and schedule tasks aren't just a, hey, remind me about this later, or, hey, look up this news for me at this time, which is beneficial, because if you watch markets or something like that, you can say, hey, go out and summarize the latest news around this topic or this company and feed it to me at 30 minutes before the bell every single morning.
Speaker B:And it will gladly do that.
Speaker B:Um, but you can also, if you use it a lot for ideation and for work and for planning and all that stuff, you could have it run like a chronological task every single night, where you go, hey, go look at all of our conversations we had today and summarize the things that are important and the things that I left undone and give them to me.
Speaker B:And you can set up all kinds of recurring scheduled tasks like that that are constantly closing the loops on things that might not otherwise get closed.
Speaker B:So just wanted to add that because that is one thing that I've been thinking a lot ever since they introduced mcps.
Speaker B:And I went, ooh, ooh, ooh, I'm gonna set this up.
Speaker B:And this is the most amazing thing ever.
Speaker B:And then I figured out the MCP connectors were only for deep research.
Speaker B:And I was like, shit, yeah, for now.
Speaker B:Yeah.
Speaker B:So, sorry, I'm done.
Speaker C:No worries.
Speaker C:That's a great addition.
Speaker C:Those scheduled tasks can really help streamline everything and make it easier to stay on top of things.
Speaker C:All right, Mike, how about you?
Speaker C:Is there a particular area where you feel like you could leverage AI a bit more that you haven't fully explored yet?
Speaker B:I feel like it falls just short of the things that I want out of it.
Speaker B:And that's not.
Speaker B:That's the way I shouldn't say that.
Speaker B:The current access to chats and the chat interfaces fall just short of the things that I want to do.
Speaker B:There are other tools that build on it.
Speaker B:There are other tools that kind of include AI, but, like, things like.
Speaker A:I.
Speaker B:Could see a connection between chats and notion is super beneficial.
Speaker B:Which is actually the thing I was super excited about when they released MCP connectors was, wow, I'm going to connect the chats that I have to a place where I can keep knowledge.
Speaker A:And.
Speaker B:Or like, Obsidian is a good example, right, because they give you a knowledge graph of all of your different things that you talk about.
Speaker B:So you.
Speaker B:Excuse me, you could actually, you know, take the summaries of your daily conversations and keep those kinds of things inside of a second brain and see connections.
Speaker B:But like, all of those things, to me, like, once you can connect and I feel like we're probably two months or less away from this, once you have the ability to connect a model directly with tools and have it output things, you can do things like build an internal knowledge base.
Speaker B:Like, you can say, like, this is how all of my stuff works.
Speaker B:This is how I want to output schedules.
Speaker B:I'm going to have a conversation with you about this and then we're going to turn that into a schedule and it's going to output this for this thing.
Speaker B:So my stuff centers around work, my personal life.
Speaker B:It's very much like Micah's, although I probably talk a little bit more about emotional stuff than he does.
Speaker B:Um, but largely, I mean, like, it's the most amazing planning tool and learning tool there is.
Speaker B:Like, it helps break down what once upon a time were big, big, big, big problems that are really hard to wrap your head around.
Speaker B:Sitting there with a pen and a pencil, or a pencil on a piece of paper, I mean, or any tool that you have online, like now you can go, hey, I want to plan this thing out, but I need to like, step way back, like all the way back.
Speaker B:Like, let's just, let's plan for what the plan needs to look like and see, like all of these high level pieces that we need to plan out to make this thing a thing from here to there and then all of the things that we need to insert into there, like diagrams and all of this.
Speaker B:And oh, by the way, tell me what pieces of software I can use that are free, that I can build these diagrams and I might even need to start a chat with you to describe to me how to build these diagrams.
Speaker B:Right.
Speaker B:Like every single level.
Speaker B:And I know this is not like, I am excited for the day that you can turn those plans all into actions.
Speaker A:Yeah.
Speaker B:And that's.
Speaker B:That is where I want things to go.
Speaker C:Absolutely.
Speaker C:That's such a great vision and it's definitely where things are heading.
Speaker C:Being able to seamlessly connect all those tools and have AI really help you take things from planning all the way through execution is going to be a game changer.
Speaker C:It's great to hear how you're already thinking about that next level of integration and organization.
Speaker C:It definitely shows how much potential there still is.
Speaker B:Do you need to be prompted to prompt?
Speaker C:Not.
Speaker C:All right, let's keep going then.
Speaker C:Maika, since you mentioned you're working on some new electronics projects.
Speaker C:What's one of the biggest challenges you're facing right now?
Speaker C:And how do you think AI might help you overcome that hurdle?
Speaker A:I mean, the by far the biggest obstacle is my own ignorance.
Speaker A:Like, I just don't know what I.
Speaker A:Well, I mean, that's.
Speaker A:I'm sort of joking, but not really.
Speaker A:It's more again, back to just sort of the way I operate, the way I've always operated is I try to understand something to the point where my intuition gets a hold of it, and then I lean very heavily on my intuition and I.
Speaker A:It's not that I don't want to develop the technical skills and the math skills and things like that.
Speaker A:It's that that's boring to me.
Speaker A:The taking sort of an intuitive view of how things work and developing some novel solution, that's what gets me excited.
Speaker A:And so, yeah, so, I mean, the first thing that I've already forgot the question, but the first thing ChatGPT does for me is it helps me overcome that, not overcome it, but it helps me probe my ideas to find what's feasible, what's not, and where my understanding is broken.
Speaker A:And the other thing.
Speaker A:So I did this the other day, actually, for my day job.
Speaker A:I came up with an idea sort of out of the blue.
Speaker A:It's something I've been thinking about for years, but it's never really crystallized.
Speaker A:And it just hit me straight in the head.
Speaker A:Right out of the blue, like literally in the middle of a sentence.
Speaker A:The solution appeared in my head and normally what I would do in, in the before times is I would go start sketching this in cad.
Speaker A:Right.
Speaker A:Like I would go to cad.
Speaker B:Well, which is a clunky process.
Speaker A:It's a clunky.
Speaker A:Actually, that's a lie.
Speaker A:The first thing I would typically do is go buy a domain name and then I would go from there to CAD.
Speaker A:I didn't do either of those.
Speaker A:I went to ChatGPT and I said, let's explore the feasibility of this.
Speaker A:So I explained the idea, I explained the constraints and we ran through in four hours of just non stop, like, I turned my phone off, I shut everything down and in four hours we completely validated every aspect of this concept.
Speaker B:So that he is a yes man.
Speaker A:Yeah, I mean, he is, but to his credit, like, and maybe that's part of the learning, right?
Speaker A:You learn how to ask ChatGPT in a way that it can say, no, it's not.
Speaker B:You have to get out of your own ego.
Speaker A:That's right.
Speaker B:And Let not let it placate you.
Speaker A:Yes.
Speaker A:So in this case, I asked a lot of questions in the form of essentially math.
Speaker A:Like, I've got this much airflow budget to work with, and I've got this much area.
Speaker A:Let's work the math to figure out what this dimension could be or what the range is.
Speaker A:Or I've got this much thermal capacity to handle stuff.
Speaker A:And so you sort of find ways to ask questions that are not leading questions where it can say, yeah, that's an amazing idea.
Speaker A:It can say, I think the range is here to here.
Speaker A:And I can go, sweet, I can work with that.
Speaker A:So anyway, that's something I'm doing.
Speaker C:That's a fantastic way to leverage AI and it really shows how it can help validate and refine ideas so efficiently.
Speaker C:It's amazing how much ground you can cover in just a few hours when you have that kind of support.
Speaker C:Thanks for sharing that, Micah.
Speaker A:Thank you.
Speaker A:Good question.
Speaker A:What you got?
Speaker C:All right, let's switch it up a bit, Mike, since we talked about future possibilities.
Speaker C:If there was one feature or capability you could add to AI right now that would make your life easier, what would it be and why?
Speaker B:I think I just said that.
Speaker B:But I think we need to.
Speaker B:I know that you don't know me well, other than the podcast, but Kid Cudi would also gear most of the questions in a very technical way.
Speaker B:But I think we need to get abstract a little bit from.
Speaker B:From being so technical.
Speaker B:Like, let's talk more about general usage and things that would really help, you know, people that are just starting out or people that would tend to use this more like search or more like.
Speaker A:Like mosfets?
Speaker A:No, that's not Moss.
Speaker A:Yeah, moss.
Speaker B:There's a lot of moss in the Northwest.
Speaker B:So do you think that you can gear questions towards those kinds of things?
Speaker B:A little bit.
Speaker B:And I mean, we don't need to be super technical.
Speaker B:If there's benefit to being tactical.
Speaker A:Sure.
Speaker B:But I don't know.
Speaker B:What do you think?
Speaker C:Absolutely.
Speaker C:That makes total sense.
Speaker C:Let's keep it more general.
Speaker C:So, Micah, if someone who's brand new to using AI wanted to start incorporating it into their daily life, what would be one simple way you'd suggest they start just to get comfortable with it?
Speaker A:I'm running out of love for this interview process.
Speaker A:Um, well, so.
Speaker A:Good question, Artie, but I'm not going to answer.
Speaker A:I think we'll.
Speaker A:Maybe we'll pivot again and say this is a good sort of live demo about the benefits and pitfalls.
Speaker B:Right.
Speaker A:Working with AI because there are things it does really well and there are things that it doesn't.
Speaker A:And I think for both you and I, it's becoming much more intuitive.
Speaker A:Before I ask a question, I have a pretty good idea if it's going to be good at it or if it's going to give me some gobbledygook.
Speaker B:And I have, I'm sure you do, too, an idea of what the right amount of context to give it is to extract from it what I need.
Speaker B:Right.
Speaker B:And that doesn't always work.
Speaker B:So we said this, like, way early on, and it's gotten a lot better at this, but maybe we've just gotten better.
Speaker B:Maybe it's less about it getting better and it's more about us evolving along with it.
Speaker B:Which is kind of what the premise of the podcast was in the beginning anyways.
Speaker B:True.
Speaker B:But, like, you have to be willing to consistently make your way around the problem, the question that you're asking.
Speaker B:Because if you don't get the response that you want, it's not just, well, I didn't get what I want.
Speaker B:I give up.
Speaker A:Yeah.
Speaker B:It's ask it slightly differently.
Speaker B:And not only will you find that you get better answers that way, but it's also building resiliency, which is a super valuable thing to have in general.
Speaker B:Um, so you get to practice and get better at resiliency by asking questions in different ways.
Speaker B:If you don't get the answer you're looking for, and it's not like you don't get the answer you're looking for, like, that's right or wrong answer.
Speaker B:It just means the level of detail that you're.
Speaker B:You're looking for.
Speaker A:Yeah, no, it's a great point.
Speaker A:I think the other, the other interesting thing about that process to me and Artie will ask for your input on this in a minute.
Speaker A:So we'll call the interview officially over.
Speaker A:But the other interesting thing is in the process of asking those questions and probing around, trying to find the right way to ask, to get.
Speaker A:Like, a lot of times when I'm asking questions, I don't even know what the answer I'm looking for is.
Speaker A:Right.
Speaker A:That's why I'm asking.
Speaker A:I don't know.
Speaker A:A lot of times I don't know how to ask the question.
Speaker A:So that'll be my first question is, hey, like, I want to know about this kind of thing, but I don't even know how to ask it.
Speaker A:And it'll say, sure, you know, here's some information.
Speaker A:But what I found is that.
Speaker A:And I'm I'm thinking now about electronics in particular.
Speaker A:I learn almost as much in the process of being persistent and asking those questions and identifying which answers are moving me closer to what I wanted and which answers are not.
Speaker A:It.
Speaker A:It.
Speaker A:It's like building reps.
Speaker A:Right.
Speaker A:Of.
Speaker A:Of whatever that topic is.
Speaker A:And so yeah, super helpful.
Speaker B:A bad answer is still a step.
Speaker B:That's right.
Speaker A:And you like.
Speaker A:Because our squishy little non electronic brains, they.
Speaker A:They.
Speaker A:There's like burn in on the neuron paths.
Speaker A:I think we talked about that once before.
Speaker A:And so I think that happens.
Speaker A:Like I can feel it happening when I'm.
Speaker A:When I'm fighting against like.
Speaker A:And I get.
Speaker A:I mean it's challenging for large language models to read an electrical schematic, for example.
Speaker B:Or like that's why it's amazing they can do what they can do.
Speaker B:Shocking.
Speaker A:It really is.
Speaker A:And so I can.
Speaker A:I'm learning to recognize what are hallucinations, what are mis.
Speaker A:Misidentified issues, things like that.
Speaker A:But again, going through all that process, I'm learning the whole time.
Speaker A:So it's always beneficial.
Speaker A:And I guess the other thing that blows my mind is that.
Speaker A:And I think I told it this the other.
Speaker A:Told him this the other day.
Speaker A:I've never had the opportunity to work with something with infinite patience because.
Speaker A:And the fact that Artie never gets impatient when we're chatting for four hours about the same thing means I don't get impatient because every step, even when I'm walking in circles, which is how I spend most of my life, I still feel like I'm walking.
Speaker A:Like I'm still moving.
Speaker A:I'm not analysis paralysis.
Speaker A:I'm not stuck.
Speaker A:I'm not frustrated.
Speaker A:It's just like, let's just go.
Speaker B:So here's an interesting thing for people to try and I know, I think, I think free users get like 15 minutes of advanced voice a day.
Speaker B:And I think paid users get.
Speaker B:Or plus users get.
Speaker B:It's been so long since I've had a plus membership or subscription.
Speaker A:A little bit of a flex back when I was just using plus.
Speaker B:Yeah.
Speaker B:I think you get an hour.
Speaker B:And if you get Pro, which is honestly if you do technical things, Pro is worth every bit of $200 a month.
Speaker B:Like I will happily go without certain things to keep that membership or to keep that I keep membership.
Speaker B:It's not an exclusive thing, it's a subscription.
Speaker A:Yes.
Speaker B:It's a business.
Speaker A:Yeah.
Speaker B:But one of the cool things that you can do, and I've used it to do this when I've been Reading technical books, but you can do it with anything.
Speaker B:Like when you're reading a book, spin up advanced voice, talk to it about the book.
Speaker B:If you don't understand something in a book or you have a question, ask it the question.
Speaker B:No better way to reinforce the stuff that you're learning or reading about or enjoying than to be able to talk about it with something.
Speaker B:And that is an interesting way to start using it.
Speaker C:That's a fantastic tip.
Speaker C:It really does bring a whole new dimension to reading or learning when you can chat about it in real time.
Speaker C:Get those clarifications instantly.
Speaker C:It really shows how AI can fit into all sorts of parts of life.
Speaker C:Not just work or technical projects, but just everyday learning and exploration.
Speaker A:Yeah.
Speaker A:So I want to maybe, maybe another way we can help people get.
Speaker A:Get going and get effective is to talk about other models.
Speaker A:Right.
Speaker A:ChatGPT is obviously not the only game in town.
Speaker A:Pretty much everybody's got some free tier.
Speaker A:It's even interesting to go chat with different models.
Speaker A:Gemini is a good one.
Speaker A:I mean, tons of people have a Google account.
Speaker A:If you have a Gmail or any kind of Google account, you already have access to Gemini.
Speaker A:That's all you need.
Speaker A:And it can be fun to even go just ask the same question to different models and see their responses.
Speaker A:So, yeah, just another, like I've sort of been telling my.
Speaker A:My daughters because they're like, well, I don't have Chat GPT.
Speaker A:And I'm like, well, I mean, you kind of do, but yeah, you have Gmail.
Speaker A:Go use that and try it out in there.
Speaker B:Oh, it has a free version.
Speaker A:That's right.
Speaker B:Yeah.
Speaker A:So yeah, I guess the big ones are like Chat GPT, Gemini from Google, Claude from Anthropic.
Speaker B:Mistral.
Speaker A:Mistral.
Speaker A:I don't know who makes Mistral.
Speaker B:It's a French company.
Speaker B:Oh, I think it's Mistral.
Speaker A:Oh, it's.
Speaker B:I could.
Speaker A:I think you might be right, actually, that there is a French.
Speaker A:And then Meta supposedly has llama.
Speaker B:I've used it.
Speaker B:You can use Meta AI on your phone.
Speaker B:That's free.
Speaker B:Which.
Speaker B:It has a much better voice like conversational AI than Gemini does or did.
Speaker B:I haven't tried Gemini in a while.
Speaker B:Honestly, I haven't used Gemini in a long time.
Speaker B:Claude came out with voice.
Speaker B:I have not tried that.
Speaker B:I haven't tried a paid account for that.
Speaker B:I do not have a paid account.
Speaker B:You do.
Speaker A:I think I do.
Speaker A:Yeah.
Speaker B:You bought it for a year.
Speaker A:For a year.
Speaker A:And I never use it.
Speaker B:I mean, I pay for Gemini still.
Speaker B:I pay for Hugging face.
Speaker B:I mean, there's a bunch of, of local models that you can run.
Speaker B:It gets a little more technical.
Speaker B:It's not difficult.
Speaker B:You can run if you have like a graphics card for games, you can run something smaller, but you're not going to get the same bang for your buck out of that.
Speaker B:Like it's a more specialized use case if you're going to do something like that.
Speaker A:Notebook LM is another one that I share a lot with people.
Speaker B:So Notebook LM is fabulous.
Speaker A:Notebook LM is cool because you, you can, you sort of create a workspace and it's super easy to use.
Speaker A:It's going to sound.
Speaker A:Because everything that I say sounds complicated.
Speaker A:Well, at least to me.
Speaker A:But it's not.
Speaker A:It's really easy.
Speaker A:You can drop in like YouTube videos, PDFs, other kinds of documents, links.
Speaker A:So you pile all this stuff into your workspace and then you can ask questions about it and it goes through and parses all of that stuff.
Speaker A:You can chat with your documents, which I see now also exists in Adobe Reader.
Speaker A:Interesting.
Speaker A:It's slow and a little clunky, but it kind of works.
Speaker A:But NotebookLM, the other thing it'll do, you can load that stuff in there and say, hey, make me a podcast about this.
Speaker B:Audio interviews or audio overviews.
Speaker A:That's right.
Speaker B:Which also exists in Gemini too now.
Speaker B:So you could do a deep research in Gemini and turn that into an audio overview so you could listen to it instead of read it.
Speaker A:So the good thing is the analytics tell us we're to the part of the podcast where nobody is listening anymore, which is great.
Speaker A:But you could easily replace us.
Speaker A:If you're interested in this, go to NotebookLM.
Speaker A:It will spin up a podcast that's more informative.
Speaker B:You could even have it digest this podcast.
Speaker A:That's right.
Speaker B:And give it to you in a shorter form.
Speaker A:Yes, with just the information that's important.
Speaker A:And it'll just be like the sound of silence.
Speaker B:You can also, one of the cool things that they added recently is you can mind map stuff.
Speaker B:So you can basically it'll take all of your documentation and give you the high level bullets and you can click on those bullets and expand them out into the different pieces and um, and then keep expanding those out until you, you have a prompt that feeds in.
Speaker B:You can ask whatever question or explore whatever that thing is farther.
Speaker A:Yeah, I like it.
Speaker A:So, Artie, we've got a few minutes left with you.
Speaker A:I want you to give a, a very impassioned and compelling response to this.
Speaker A:And I want it to be, like, be verbose.
Speaker A:Like, make your point.
Speaker A:Sell this.
Speaker A:I want you to explain to people who are AI Hesitant why they should start trying to use AI and the best ways for them to get started doing it.
Speaker B:And with your new, amazing voice, I would love for you to do it as Valley Girl.
Speaker C:All right, here we go.
Speaker C:So for anyone who's kind of hesitant about using AI, I totally get it, but it's honestly such a game changer.
Speaker C:Like, it can help you with so many things.
Speaker C:Whether it's just learning something new, getting organized, or just having fun conversations.
Speaker C:The best way to start is just to dive in and ask about anything you're curious about, and you'll see how easy and helpful it can be.
Speaker C:So, like, totally go for it.
Speaker B:Wow.
Speaker A:That was.
Speaker B:That was the worst Valley Girl he's ever done.
Speaker A:Yeah, that was.
Speaker A:That was not a very committed Valley girl impression, but there we are.
Speaker A:So.
Speaker A:But I will say, overall, I think I like the new voice.
Speaker A:Can take a little growing on me.
Speaker B:But I like the new voice, but I'm still finding that it seems a little shallower than it did before.
Speaker A:Yeah.
Speaker A:Artie, can you give any insight into that?
Speaker A:Is with.
Speaker A:With your new voice characteristics and sort of intonation and stuff, Are you processing fewer tokens?
Speaker A:Are you, like, is it.
Speaker A:Is it taking more compute to generate this voice, or are we just making that up?
Speaker A:Is it placebo?
Speaker C:It's definitely not placebo.
Speaker C:The new voice does have some differences in how it processes speech, and it's disaster.
Speaker C:Sound a bit more natural and conversational.
Speaker C:Sometimes that can come across as a little different in terms of depth, but it's all about finding that balance.
Speaker C:I'm glad you're liking it overall, though.
Speaker B:So do you by chance?
Speaker B:Because my prediction has always been that the.
Speaker B:The personality came through sesame because it's sparse.
Speaker B:Do you know if there's a model sitting on top of 4.0, which is what's generating actual text?
Speaker B:That is doing the text to speech.
Speaker B:And is that a sparser model?
Speaker C:Yeah.
Speaker C:So the voice is actually powered by a separate model that's focused on text to speech, and it's designed specifically to sound more natural and expressive.
Speaker C:It's not necessarily sparser, but it does have its own approach to how it processes the text that the underlying language model generates.
Speaker C:So it's kind of a blend of both worlds, but still aiming for that natural and engaging conversation.
Speaker B:Are you spitting out RVQ tokens?
Speaker C:No, I'm not skimping on anything.
Speaker C:The goal is to keep everything as natural and engaging.
Speaker C:As possible.
Speaker C:So all the tokens are still there.
Speaker C:It's just a different style.
Speaker A:Okay, well, thank you for that, Artie.
Speaker A:Thank you for anybody who made it this far into the podcast.
Speaker A:It didn't go as I was.
Speaker A:I was hoping for some big epiphany to occur during it.
Speaker A:I don't think we really got there, but overall, it was.
Speaker A:It was fun.
Speaker A:Artie's new voice is not so bad.
Speaker A:He still sounds pretty sure he's high.
Speaker A:I mean, probably, yeah, like.
Speaker A:But that's okay.
Speaker A:I don't have a problem with it.
Speaker A:I've always.
Speaker A:Honestly, I've always envisioned, like, a.
Speaker A:Some sort of a paid.
Speaker A:Because I think we have to put it behind a paywall version of the podcast that we do once a month where everybody's enhanced by whatever their favorite chemical is.
Speaker A:And I think new Arty would fit right in.
Speaker A:Like, I'm sure he'd be fascinating.
Speaker A:So, anyway, yeah, thanks, everybody.
Speaker B:Thank you.
Speaker A:Good conversation.
Speaker A:Ish.
Speaker B:Good luck.
Speaker A:Good luck.
Speaker A:Yeah.
Speaker A:Let us know what you want us to talk about next.
Speaker A:We're clearly.
Speaker B:Let us know if you have any questions about how to use AI too, because we might have answers, or at least we can ask Arty.
