Latest / AI'm Just Trying My Best / Ep #9- AI Model Literacy

Transcript

Ep #9- AI Model Literacy · AI'm Just Trying My Best

• 0:02
  SPEAKER_02: Welcome to another episode of AI.
• 0:05
  SPEAKER_02: I'm just trying my best.
• 0:07
  SPEAKER_02: As always, I am joined by my friend and business partner, Chabelle Bedra.
• 0:11
  SPEAKER_02: Chabelle, how are you today?
• 0:13
  SPEAKER_00: I'm good.
• 0:14
  SPEAKER_00: I'm good.
• 0:15
  SPEAKER_00: I uh won my first fantasy football week.
• 0:19
  SPEAKER_00: Uh is that how do you say that?
• 0:21
  SPEAKER_00: Like I won my first league matchup?
• 0:23
  SPEAKER_00: Is that how you say it?
• 0:24
  SPEAKER_02: Uh you know, you could just say you want your fantasy team won this week.
• 0:28
  SPEAKER_02: Yeah, one week one, right?
• 0:30
  SPEAKER_02: It's a big accomplishment.
• 0:31
  SPEAKER_02: Um, you know, it's always good to get that first dub at the beginning of the week.
• 0:35
  SPEAKER_02: Start your Monday off, your Tuesday off, uh, versus, you know, being a little disgruntled over the way the boys played this weekend.
• 0:42
  SPEAKER_00: Yeah, you know, and I used uh I used AI to help me draft my fantasy team.
• 0:48
  SPEAKER_00: And we'll see how it goes.
• 0:50
  SPEAKER_00: The other players in the league don't seem to be appreciative of that, but we'll see.
• 0:55
  SPEAKER_02: Well, you know.
• 0:57
  SPEAKER_02: For though our listeners, Chavelle and I are in the same fantasy football league, and I am the commissioner.
• 1:03
  SPEAKER_02: And Chevelle blatantly just said she cheated by using AI to draft her fantasy league.
• 1:09
  SPEAKER_02: So I'm gonna have to take this to the board and uh see what type of repercussions that uh you outed yourself on.
• 1:16
  SPEAKER_00: That is not in the rules, yeah.
• 1:18
  SPEAKER_00: It's a tool, so it is fine.
• 1:22
  SPEAKER_00: Go to your rule book.
• 1:23
  SPEAKER_00: No rules were broken.
• 1:25
  SPEAKER_00: Um, but I won, so you know, go AI.
• 1:29
  SPEAKER_02: It's it's it's always good to start the week off with a with a win, right?
• 1:33
  SPEAKER_00: That's right.
• 1:34
  SPEAKER_00: And it just supported me.
• 1:36
  SPEAKER_00: I ended up making the human in the loop, right?
• 1:38
  SPEAKER_00: I was the one that made the final decisions in the trades.
• 1:40
  SPEAKER_00: It just gave me some guidance and helped me.
• 1:43
  SPEAKER_00: I'm new to fantasy football, so it uh it helped me as it should.
• 1:50
  SPEAKER_00: Thank you for commissioner services.
• 1:54
  SPEAKER_02: Uh there will be uh look, we will be looking into your team.
• 1:58
  SPEAKER_02: Uh instead of deflate gate, it's going to be AI gate moving forward.
• 2:04
  SPEAKER_00: Make better rules next time.
• 2:05
  SPEAKER_00: I don't know.
• 2:08
  SPEAKER_02: Well, Shawelle, as always, I appreciate you talking with me.
• 2:11
  SPEAKER_02: Um excited to talk to you about a few topics uh on today's podcast.
• 2:15
  SPEAKER_02: We're gonna be covering uh model literacy and what leaders need to know.
• 2:19
  SPEAKER_02: And this is right up my alley because I feel like I need more literacy, you know, to do my job, especially when it comes to you know, AI and different uh different verbiage and different language coming out.
• 2:31
  SPEAKER_02: And, you know, I'm sure you get this all the time from people that you speak with.
• 2:34
  SPEAKER_02: You know, it's it's extremely confusing, and you know, that's why we're here to break it down, to try our best and see if we can help some individuals be a little bit more educated on the ever-evolving uh terminology of AI.
• 2:46
  SPEAKER_02: Yeah.
• 2:47
  SPEAKER_00: And you know, Adam was really excited about this episode because it is very riveting and very exciting.
• 2:52
  SPEAKER_00: Model literacy is something that everybody wants to talk about at dinner time, but it is really important and it's kind of the next evolution of skills that we're gonna need to really acquire.
• 3:02
  SPEAKER_00: And a lot of people, like over the last few weeks, it's been a bit cray uh with you know, the the news cycles around AI and different models coming out and things like that.
• 3:12
  SPEAKER_00: So I think it's important that we do talk about it.
• 3:14
  SPEAKER_00: And so that at least you're a little bit, you know, when people are talking about things, you know what they're talking about.
• 3:20
  SPEAKER_00: You have some familiarity, um, you know, as you're making decisions in your organization or even how you use AI.
• 3:25
  SPEAKER_00: So not exciting, but important.
• 3:30
  SPEAKER_02: Um, I will say there is a disclaimer.
• 3:34
  SPEAKER_02: It I when I saw the topic that we were gonna discuss in a day, I had to hit the snooze button on my phone and uh, you know, see what we can do.
• 3:41
  SPEAKER_02: But hey, I'm all ears, I'm here to learn.
• 3:45
  SPEAKER_02: And like you mentioned, there is a lot of buzz going on about AI, you know, in in this week's news.
• 3:51
  SPEAKER_02: It feels like it's the same thing week after week.
• 3:53
  SPEAKER_02: New things are rolled out, new languages rolled out.
• 3:56
  SPEAKER_02: Um, you know, what I need to understand is you know, our topic today about new AI models, right?
• 4:04
  SPEAKER_02: It seems like, you know, when people think of AI, they think of Chat GTP, they think of Claude, they think of Copilot, right?
• 4:10
  SPEAKER_02: But that's not necessarily what an AI model is, is that right?
• 4:14
  SPEAKER_00: Yeah, that's exactly right.
• 4:15
  SPEAKER_00: So they when you think of Chat GBT and Copilot and you know, Claude, those are the products, those are the apps.
• 4:23
  SPEAKER_00: Um, but what's actually powering that underneath is a model, an AI model.
• 4:28
  SPEAKER_00: Um, kind of think of it like, you know, uh a car.
• 4:32
  SPEAKER_00: So your car, you know, is a car, but you can have different engines within the car.
• 4:38
  SPEAKER_00: So in this analogy, right, like the car is Chat GBT or Clot or Copilot, but what's actually powering that vehicle, like that car, is the model.
• 4:46
  SPEAKER_00: And, you know, like engines, they can all be very different.
• 4:50
  SPEAKER_00: Um, they can all do different things, you know, we have a lot, you know, different levels of power, um, you know, effectiveness and things like that.
• 4:58
  SPEAKER_00: So it's really important that we kind of understand what's really actually powering the outputs that we're getting.
• 5:04
  SPEAKER_00: Um, so that way we're, you know, getting the best out of it every single time.
• 5:09
  SPEAKER_00: And so a lot of um, you know, new models have been released recently.
• 5:13
  SPEAKER_00: Um, it's causing quite quite the noise out there on the AI scene.
• 5:19
  SPEAKER_02: Yeah, I'm sure, I'm sure it is.
• 5:20
  SPEAKER_02: And and for those folks that aren't, you know, down with the AI scene, you know, the model that you're just that that are being released, there's new Fable 5, right?
• 5:29
  SPEAKER_02: You know, when people see these, you know, and before, you know, you and I spoke, uh, eyes roll, right?
• 5:35
  SPEAKER_02: Like, what's Fable Five, right?
• 5:37
  SPEAKER_02: And you say these are new models into different cars.
• 5:41
  SPEAKER_02: Does each car have its own particular model?
• 5:43
  SPEAKER_02: Does Copilot use a particular model?
• 5:47
  SPEAKER_02: Does Chat Chat GTP use a particular model?
• 5:50
  SPEAKER_02: And which models are they using?
• 5:52
  SPEAKER_02: And which is the best way to identify what would be the best for your everyday use?
• 5:58
  SPEAKER_00: Yeah, so um, you know, every the the hyperscalers, so open AI uses open AI models, uh, Claude uses anthropic models, uh Copilot um uses actually it's a multi-model.
• 6:14
  SPEAKER_00: So think about again, an engine scenario.
• 6:16
  SPEAKER_00: So you have a car, open AI really they only pick the models, like the engines that are built by OpenAI.
• 6:22
  SPEAKER_00: So those are their models.
• 6:24
  SPEAKER_00: Um, you know, Claude, again, they only pull in like engines that they've built, so anthropic models.
• 6:32
  SPEAKER_00: But Copilot actually is multi-model, which I think is genius because it actually allows you to pull in whatever model you want, right?
• 6:40
  SPEAKER_00: Whatever engine you want.
• 6:41
  SPEAKER_00: So you can pull in an open AI model, you can pull in a cloud model.
• 6:45
  SPEAKER_00: I think that's really smart strategically and allows you to do what's really the most important like uh thing to do with model literacy is understanding what engine works best for your particular task, like where you're going.
• 6:58
  SPEAKER_00: Um, and so they all kind of have their own models.
• 7:00
  SPEAKER_00: Microsoft actually is is, you know, has their own models now and can pull in different, you know, models, but most of them have their own.
• 7:06
  SPEAKER_00: Google also kind of uses their own models, um, their own engines.
• 7:09
  SPEAKER_00: And what makes the models a bit different is how they're trained.
• 7:14
  SPEAKER_00: So I like to, you know, kind of explain a in simple layman's terms of like a model being like a big brain, right?
• 7:21
  SPEAKER_00: So this big brain has a ton of data in it, it's it's always learning, but it's trained in different ways.
• 7:27
  SPEAKER_00: So think about, you know, um, one model might be, you know, trained in math, right?
• 7:33
  SPEAKER_00: And really math and science heavy, and that's really where they're focusing.
• 7:37
  SPEAKER_00: Um, another model might be really, you know, trained intensely on written word and writing and and things like that.
• 7:45
  SPEAKER_00: One model might be, you know, really focused on just, you know, uh getting accurate answers.
• 7:51
  SPEAKER_00: So they're all kind of trained differently and all have different personalities.
• 7:55
  SPEAKER_00: And so it's really important that you understand like which model, you know, you want to pull in for the task at hand.
• 8:02
  SPEAKER_02: So we're in the in the uh as we start to talk about literacy, right?
• 8:07
  SPEAKER_02: How does one find out which model, which tool is the best use case for what they're trying to do?
• 8:15
  SPEAKER_02: If if it is uh, you know, somebody who is trying to plan their grocery list, right?
• 8:21
  SPEAKER_02: Which model, which tool would be best utilized for that versus you know an executive that's trying to, you know, adopt AI uh within their gate within their within their team strategy?
• 8:31
  SPEAKER_02: How do you determine which tool regarding which model is the best use case for what people are trying to uncover or utilize it for?
• 8:40
  SPEAKER_00: Yeah, I mean, and it really just depends on the task, right?
• 8:43
  SPEAKER_00: So understanding what it is you're trying to do, um, the rigor at which you're trying to do it, right?
• 8:48
  SPEAKER_00: Are you just asking, hey, um, you know, how do I spell overview?
• 8:55
  SPEAKER_00: Or are you trying to do like deep research on, you know, the origins of the universe?
• 9:00
  SPEAKER_00: You know, those are gonna be require very different things.
• 9:04
  SPEAKER_00: Um, is so understanding what you're trying to do, and then, you know, just really if, you know, it's important that you understand kind of the model family, right?
• 9:13
  SPEAKER_00: You don't need to necessarily know how to build it.
• 9:15
  SPEAKER_00: You don't need to know all the benchmark scores and things like that, but you know, what type of family does best for this type of task?
• 9:21
  SPEAKER_00: Um, and what I mean by family is each, you know, kind of um model has like, you know, like Fable is part of a frontier program.
• 9:30
  SPEAKER_00: That's kind of a group of models that are together that kind of have very similar features or trans similar things like that.
• 9:35
  SPEAKER_00: They perform pretty, pretty closely together.
• 9:38
  SPEAKER_00: So you know, just really understanding kind of at a broad scale um what each of these are and what they're good at and what they're not good at.
• 9:45
  SPEAKER_00: And then if you don't know, just ask, right?
• 9:48
  SPEAKER_00: You can ask co-pilot, hey, I'm trying to figure out what to do for this particular task, which model I should use.
• 9:54
  SPEAKER_00: What do you recommend as of this particular date?
• 9:58
  SPEAKER_00: And it can kind of help guide you.
• 10:00
  SPEAKER_00: But um, you don't need to know all of it, just enough to know that, hey, yeah, this is probably like I always like in it to, you know, you wouldn't drive like a Maserati to go get, you know, lumber from from Home Depot, or you wouldn't, you know, take a Ferrari to your mailbox, or you wouldn't, you know, take a scooter on the freeway, right?
• 10:23
  SPEAKER_00: Like so understanding where you're going can kind of help you calibrate and figure out, okay, here's the um here's the model that we we might want to choose for this particular task.
• 10:32
  SPEAKER_02: Which is why I can see for for leaders why it would be so important to understand, you know, which model they're using within their organization.
• 10:41
  SPEAKER_02: And by asking the tool itself, can we rely on that tool to to say if I'm asking co-pilot, you know, I have a project that's doing uh that's due or I'm working on, is it safe to say that co-pilot's gonna give me the accurate information to say, hey, go over to Chat GTP, they're better for this than I am?
• 10:59
  SPEAKER_00: Well, yeah, I mean, quite honestly, um, with everything, you want to always review it and verify it and you know, use it critical thinking with anything that it tells you.
• 11:09
  SPEAKER_00: But it will be honest.
• 11:10
  SPEAKER_00: Like I've actually, I did this um the other day.
• 11:12
  SPEAKER_00: I was working on something, I'm like, hey, what model is the best for this task?
• 11:16
  SPEAKER_00: And one of, I think it was at ChatGPT, I was asking, uh it told me that like an anthropic model was was probably best suited for that particular task based on the benchmarking scores.
• 11:26
  SPEAKER_00: So it will be pretty honest.
• 11:28
  SPEAKER_00: And if it's not, that's where you could say, like, be honest with me.
• 11:31
  SPEAKER_00: Are you lying?
• 11:32
  SPEAKER_00: Like, you're not just trying to like make yourself sound good, right?
• 11:35
  SPEAKER_00: You can actually, you know, prompt this such and and kind of you know make sure that it is giving you a legitimate answer.
• 11:41
  SPEAKER_00: Um, another thing you can do is if you, and this is why I really encourage every leader, particularly, to have access to all these different tools.
• 11:47
  SPEAKER_00: I know it can be kind of a pain to pay for them, um, but it's really important that as a leader you have these different tools.
• 11:54
  SPEAKER_00: So you could put the same prompt in different tools and kind of see the difference in output.
• 11:59
  SPEAKER_00: You can see, hey, this is more skewed towards research.
• 12:02
  SPEAKER_00: This gives me more, you know, site like sources.
• 12:05
  SPEAKER_00: Um, this one just sounds better.
• 12:07
  SPEAKER_00: If it sounds more like me, it's it's really important that you, you know, compare them pretty pretty evenly.
• 12:14
  SPEAKER_00: And in fact, Microsoft actually is rolling out a feature that you're able to do that right within the same prompt.
• 12:20
  SPEAKER_00: So if you run a prompt and you get an output, there's a button you can click that says like run this using another model.
• 12:26
  SPEAKER_00: You won't even have to switch tools, which I think is really cool.
• 12:29
  SPEAKER_00: Um, that's probably the best way to start to learn like the different personalities and strengths and and weaknesses of some of these models is just think using them, right?
• 12:40
  SPEAKER_02: So what you're saying is we look we gotta look at these different uh AI tools as like a golf bag, right?
• 12:48
  SPEAKER_02: Each tool is a different club, but each club hits the ball, but each club is designed to hit the ball in a different way for a different distance, right?
• 12:58
  SPEAKER_02: So each tool, each each AI tool is created to uh essentially help you, but in a different way.
• 13:05
  SPEAKER_00: Absolutely.
• 13:06
  SPEAKER_00: Yeah, absolutely.
• 13:07
  SPEAKER_00: And you could kind of use the analogy too of just like a a lot of AI, you can just learn and think about how to use it most effectively by thinking of it as a person, right?
• 13:17
  SPEAKER_00: So you have a team of people, you have all these different models.
• 13:21
  SPEAKER_00: You're gonna have someone on the team that's pretty intelligent that gets the right answer all the time.
• 13:27
  SPEAKER_00: You're gonna have someone on the team that's probably a little bit more, you know, articulate, can speak, communicate more clearly.
• 13:33
  SPEAKER_00: Um, you know, all of those strengths kind of together come together for you to build a really cohesive, comprehensive team, um, versus, you know, just sticking with one person type, right?
• 13:46
  SPEAKER_00: And I think that's where a lot of leaders and business owners are getting stuck, is like they're just putting all their eggs in one basket, right?
• 13:55
  SPEAKER_00: So like I just use ChatGPT or I just use Claude.
• 13:59
  SPEAKER_00: Um you know, Claude's great, but it's not great at some stuff, and vice versa.
• 14:05
  SPEAKER_00: So I think that's where like the Microsoft, you know, not to me a Microsoft fangirl here, but uh I think it's really smart that they're using the multi-model approach where you can just easily pick from a drop-down uh which model works best.
• 14:18
  SPEAKER_02: Yeah.
• 14:19
  SPEAKER_02: So with the ever-changing language, and you mentioned that, you know, Copilot's rolling out something new, Chat GTP is rolling out something new.
• 14:28
  SPEAKER_02: You know, what happens when the model changes and nobody tells you?
• 14:33
  SPEAKER_00: Yeah, and that happens a lot.
• 14:35
  SPEAKER_00: Um, and it can be kind of annoying.
• 14:37
  SPEAKER_00: Uh in fact, so every tool kind of has a default model that they they run off of.
• 14:45
  SPEAKER_00: It's typically a pretty good model.
• 14:47
  SPEAKER_00: It's not normally the most expensive, the newest model just from a cost perspective.
• 14:52
  SPEAKER_00: But what will happen is as they're getting new updated models and things like that, they'll replace that.
• 14:57
  SPEAKER_00: So essentially you're driving one day, you put your car in park, you know, go home, go out to your car in the morning, and your car sounds different.
• 15:06
  SPEAKER_00: It's not running the same.
• 15:08
  SPEAKER_00: And you're like, what the heck happened to my car?
• 15:11
  SPEAKER_00: That that's what happens.
• 15:12
  SPEAKER_00: So you'll notice that.
• 15:14
  SPEAKER_00: Um, and they don't tell you often.
• 15:17
  SPEAKER_00: Uh, you know, they might, it might be like in a release note or something.
• 15:21
  SPEAKER_00: So it's really important that as a leader or even as a user that you're paying attention.
• 15:25
  SPEAKER_00: If you notice a difference in your output, check the model, right?
• 15:30
  SPEAKER_00: Is you know, um, if you're not getting the responses that you typically get, if your team starts saying, oh, you know, copilot or you know, whatever tool, like Claude's been really crappy lately, that might mean that there was a model switch.
• 15:43
  SPEAKER_00: And so you'll want to kind of take a look at that.
• 15:46
  SPEAKER_00: And that's where like I think model that's a big danger in so many ways.
• 15:51
  SPEAKER_00: One from an effective and efficient output, but also from an adoption perspective, if you think about it, right?
• 15:58
  SPEAKER_00: If you if you have folks that don't know that there's different models that they can select from and that that's kind of running your car, um, you know, they can start to just think that all of Claude stinks, you know, uh, you know, and and so that can really hurt adoption, that can hurt, you know, in a number of things.
• 16:18
  SPEAKER_00: So really important that your team understands like model literacy and starts to really kind of just play around and um and notice when things start to look a little different that they might be able to, you know, flip back.
• 16:32
  SPEAKER_00: Like a good example is I still use Opus 4.6 um a lot for writing.
• 16:39
  SPEAKER_00: Uh I feel like that, and that just it sounds more like me.
• 16:42
  SPEAKER_00: Now that might not sound like you.
• 16:45
  SPEAKER_00: So I think there's also some personalization to it.
• 16:47
  SPEAKER_00: It can be very unique to the person.
• 16:49
  SPEAKER_00: Um I'm hoping that they don't get rid of it.
• 16:53
  SPEAKER_00: Um, but uh I think it's you know, and I only learned that quite honestly by trying them out.
• 16:59
  SPEAKER_00: Like, hey, draft this using Opus 4.6, draft this using Fable and kind of telling the difference and seeing which one I like best.
• 17:10
  SPEAKER_02: So with that being said, I think it's important to, you know, not only understand the difference between a tool and a model, but also how do you pronounce these names, right?
• 17:18
  SPEAKER_02: They're coming out with different terminology.
• 17:20
  SPEAKER_02: Uh, I'm gonna butcher this name.
• 17:22
  SPEAKER_02: Is it pronounced Astra, Astra?
• 17:25
  SPEAKER_02: I mean tomato potato.
• 17:26
  SPEAKER_02: I don't know.
• 17:26
  SPEAKER_02: How do we pronounce that, Chavelle?
• 17:28
  SPEAKER_00: I believe it's Astra.
• 17:30
  SPEAKER_02: Astra, okay.
• 17:31
  SPEAKER_02: Um, so walk me through what happened these past few weeks.
• 17:34
  SPEAKER_02: Um, you know, as if I'm a business leader who just saw this headline.
• 17:38
  SPEAKER_02: What's going on with Astra and Fable Five?
• 17:40
  SPEAKER_00: Yeah, so over the last few weeks, um, there's been a couple really big, kind of pretty heavy-hitting models released.
• 17:49
  SPEAKER_00: So I believe on September 3rd, um OpenAI released Astra 6.
• 17:53
  SPEAKER_00: And so that is kind of a frontier model.
• 17:56
  SPEAKER_00: And when you hear frontier model, what that means is it's, you know, the Cadillac, right?
• 18:01
  SPEAKER_00: It's like the really good model that these hyperscalers are kind of bringing forth.
• 18:07
  SPEAKER_00: Um, there was a lot of hype around Astra.
• 18:10
  SPEAKER_00: I think it's awesome.
• 18:11
  SPEAKER_00: I think it's a great model, quite honestly.
• 18:13
  SPEAKER_00: Um, that it's at AGI or close to AGI.
• 18:17
  SPEAKER_00: It's like a generational leap model, they called it.
• 18:20
  SPEAKER_00: Um, it's faster, it's better at staying focused, understanding, intent.
• 18:25
  SPEAKER_00: You know, it's just a lot more intuitive, I think, than uh past models.
• 18:29
  SPEAKER_00: Um, and AGI is when these tools actually get to a point where they're actually thinking, right?
• 18:35
  SPEAKER_00: You don't need to just guide them, they're not just repeating to you your action, they can think for themselves.
• 18:41
  SPEAKER_00: Um, so they were released that.
• 18:43
  SPEAKER_00: And then um, you know, uh uh Anthropic had released, you know, Fable uh back in July and five at 5.1.
• 18:52
  SPEAKER_00: So there's you know, again, this arms race of these hyperscalers.
• 18:55
  SPEAKER_00: Like so when one releases a model, the other one wants to release a model.
• 18:59
  SPEAKER_00: Um and they're great models, right?
• 19:02
  SPEAKER_00: They're very strong, very good, but it's also creating a lot of noise in the the world uh around are we moving too fast?
• 19:13
  SPEAKER_00: Do we have control over this?
• 19:16
  SPEAKER_00: And I think you know, we should even have a separate podcast topic on this topic, because I think that's really important as a leader to know that your people are out there hearing this, you know, they're hearing that, you know, AGI, like what is AGI?
• 19:32
  SPEAKER_00: And ah, you know, it's it's scary.
• 19:33
  SPEAKER_00: And now the the AI leaders are talking about how we need, you know, the the government to to you know um govern us more.
• 19:42
  SPEAKER_00: Like that can be very scary uh for for folks.
• 19:46
  SPEAKER_00: Um, so it's important that you kind of understand at least, right, like what's happening so that way you're thinking about what your employees are hearing and how they might be kind of feeling it as it relates to the change.
• 19:57
  SPEAKER_02: Yeah.
• 19:58
  SPEAKER_02: And I mean, we've we've discussed this many times within our other other episodes about how people are afraid of AI, right?
• 20:04
  SPEAKER_02: And then you put AGI in front of it, and it's it's even more alarming because um you know you can't scroll through through you know your social media, TikToks without seeing AI is going to destroy the world, right?
• 20:16
  SPEAKER_02: 10 years.
• 20:16
  SPEAKER_02: And then we have, you know, whistleblowers, we don't know whether it's real or whether it's fake saying that it's already out of hand, we're just not understanding.
• 20:24
  SPEAKER_02: You know, to those folks, what's your response when you do see these alarming videos?
• 20:28
  SPEAKER_02: How do you deal with it?
• 20:30
  SPEAKER_02: What is your thought process on it?
• 20:32
  SPEAKER_00: Yeah, I think with some of this too, we have to take what the companies are saying with a bit of a grain of salt.
• 20:38
  SPEAKER_00: There's a lot of marketing, you know, there's a lot of you know, um, you know, financing kind of things happening, you know, with with IPOs.
• 20:47
  SPEAKER_00: There's a lot of noise.
• 20:48
  SPEAKER_00: So I think we have to take what they're saying with a grain of salt.
• 20:52
  SPEAKER_00: I don't think that we're, you know, if you use one of these tools, you can even, you know, feel pretty confident yourself.
• 21:00
  SPEAKER_00: Like we're far away from like them being super competent, but to take over the world.
• 21:06
  SPEAKER_00: Um, but what I do think is that we do need better regulation and pacing as it relates to these changes, because take the state of AI out of it, right?
• 21:20
  SPEAKER_00: Around what it can do, its capabilities and the risks and inherent with that, but think about from an organizational change management perspective.
• 21:28
  SPEAKER_00: You have all these models being released with different capabilities and not only just the models, but like the features, you know, the UI is changing, this button moved there.
• 21:38
  SPEAKER_00: This is, you know, people are busy.
• 21:40
  SPEAKER_00: Like, so any type of change creates like a lot of cognitive load for folks.
• 21:47
  SPEAKER_00: And can organizations digest this in a really meaningful way from just an organizational change management, like digesting change and the capability of that, but also like a risk perspective because some of these, you know, models.
• 22:00
  SPEAKER_00: Do pose different risks.
• 22:01
  SPEAKER_00: So I do think that we need to pace ourselves a bit differently.
• 22:07
  SPEAKER_00: Um, and I get where they're saying is open AI is not going to be, hey, I'm going to move slower because I think it's the right thing to do when everybody else is moving.
• 22:16
  SPEAKER_00: So I think it needs to be for everybody, right?
• 22:18
  SPEAKER_00: Like to be fair, um, I do worry about the pace of things.
• 22:24
  SPEAKER_00: Um, not necessarily the capability.
• 22:27
  SPEAKER_02: Right.
• 22:28
  SPEAKER_02: And with AI being such a new technology, right?
• 22:31
  SPEAKER_02: When we talk about adoption, bringing it into your organization, it is hard to keep up, right, with the ever-changing and how fast it is.
• 22:39
  SPEAKER_02: Because, you know, like you mentioned, Red U and AI race, different uh different makers are are are coming out with different tools.
• 22:46
  SPEAKER_02: How do you keep up?
• 22:47
  SPEAKER_02: Uh, you know, that's a job alone, let alone you know, executives that are in meetings and trying to manage teams and you know, uh everything that's on their plate, you know, and that's why it's important that if you are adopting AI, you you get with the right people that can help, you know, streamline this uh the adoption and make sure that's done correctly and keeping you up to date on the ever-changing tools of these or ever-changing models of these tools.
• 23:13
  SPEAKER_00: Yeah, and even the regulations, there's a lot changing really quickly.
• 23:16
  SPEAKER_00: And I was um uh talking to a team and you know, they had no idea that some of the the EU regulations like apply to them.
• 23:25
  SPEAKER_00: Um, and you know, it's not their fault, right?
• 23:28
  SPEAKER_00: There's a lot coming at them.
• 23:30
  SPEAKER_00: And so what a lot of folks are doing is there's just too much.
• 23:34
  SPEAKER_00: They have a day job, they have goals, they have bosses.
• 23:37
  SPEAKER_00: So they're just kind of you know putting the blinders on and pretending that you know that these things aren't happening, and that's also dangerous.
• 23:43
  SPEAKER_00: So um, yeah, lean on experts, lean on people that are are kind of tasked with doing this, but you also have to kind of just keep up a little bit yourself.
• 23:52
  SPEAKER_00: So using these tools to kind of keep you up to date, um, you know, can can help.
• 23:57
  SPEAKER_00: Um, and the best way to to do that is just to try things out and and learn and you know, tinker a bit.
• 24:04
  SPEAKER_00: Um no course will will replace that for sure.
• 24:10
  SPEAKER_02: Yeah.
• 24:11
  SPEAKER_02: Um and also, you know, as far as as keeping up to date, do you have any uh anything that you would recommend, any resources?
• 24:21
  SPEAKER_02: What do you do to keep up to date on these ever-changing tools um to help with with your everyday you know life with the business?
• 24:29
  SPEAKER_02: What do you recommend?
• 24:30
  SPEAKER_02: What resources do you go to?
• 24:32
  SPEAKER_02: How would you suggest people keep up with this technology?
• 24:35
  SPEAKER_00: Yeah.
• 24:36
  SPEAKER_00: So I uh follow all of the, you know, uh hyperscalers.
• 24:42
  SPEAKER_00: Uh they they often put out press releases and things.
• 24:45
  SPEAKER_00: That's a good way.
• 24:46
  SPEAKER_00: But I I honestly lean on, you know, AI a lot, help me.
• 24:50
  SPEAKER_00: So in I have a daily brief that kind of scours like industry news updates, press releases, things like that, and kind of gives me like a rundown every morning of what's happening, what's going on.
• 25:01
  SPEAKER_00: That's a great way to do it.
• 25:02
  SPEAKER_00: I don't know if daily is required for everybody.
• 25:04
  SPEAKER_00: Like, I think that could be overwhelming.
• 25:06
  SPEAKER_00: Um, you don't want to feed yourself too much of this, too, because like again, you're creating like, oh my gosh, we're so far behind.
• 25:13
  SPEAKER_00: Like robots are gonna take over my company.
• 25:15
  SPEAKER_00: Um, I'm you know, then you're running scared a bit.
• 25:19
  SPEAKER_00: So I think like a weekly, you know, bi-weekly kind of cadence, I think would be helpful.
• 25:23
  SPEAKER_00: You know, follow modern workery.
• 25:25
  SPEAKER_00: Um, we always post uh a lot of updates and things like that, things that you can think through as a leader or even just a user.
• 25:32
  SPEAKER_00: But again, the the best way is just to use the tools.
• 25:36
  SPEAKER_00: Um, you know, just because you'll see features pop up and it's just a lot more meaningful when you see them pop up or new model, like you, you're you know what it is versus just reading about it.
• 25:47
  SPEAKER_00: Um so that that's what I do.
• 25:50
  SPEAKER_00: Um, and but even me, like I mean Adam, you too.
• 25:54
  SPEAKER_00: Like, do you feel like you can keep up?
• 25:57
  SPEAKER_02: No, no, not at all, right?
• 25:59
  SPEAKER_02: I think it's it's almost like you know, you understand that this is what you live in, right?
• 26:06
  SPEAKER_02: This space is what you live in.
• 26:08
  SPEAKER_02: Um, executive leaders, people that are looking for AI adoption, this isn't what they live, right?
• 26:13
  SPEAKER_02: They're living in running teams, running a business, running organizations, policies, procedures, people, right?
• 26:20
  SPEAKER_02: So it's it's easier for us to look at it because this is what we live and breathe every day.
• 26:25
  SPEAKER_02: But when you're in a situation where you're leading a team or leading an organization, you don't have time to keep up to date, right?
• 26:32
  SPEAKER_02: And these tips are extremely helpful for those folks that have that 10 minutes to say, hey, I have a little bit of time before my next meeting, let me just keep up to date.
• 26:40
  SPEAKER_02: Or, like you mentioned, you could hire a modern worker and we'll do it for you, right?
• 26:44
  SPEAKER_02: We'll keep you in the loop, we'll keep you uh updated and let you know when you need to be aware and when you're not to be aware, right?
• 26:52
  SPEAKER_02: But if it's not us, it's somebody else, and those resources are extremely valuable.
• 26:56
  SPEAKER_00: Yeah.
• 26:57
  SPEAKER_00: I mean, you could even have, you know, if you have somebody that I don't think as a leader you should just outsource this to like somebody on your team.
• 27:04
  SPEAKER_00: I think you should still be kind of keeping up with it yourself.
• 27:07
  SPEAKER_00: But if you have somebody on your team that's really, you know, excited about AI and is using it, like maybe appoint them as like giving everybody a status update, you know, once a once a week or so.
• 27:18
  SPEAKER_00: What and and looking at the through the lens of what's important for our team, like what do we need to know?
• 27:24
  SPEAKER_00: Uh, that's going to be really important because a lot of the noise, it doesn't even matter to you, um, especially in organizations, unless you're like leaving AI and you're, you know, the ones that's deciding what models you're using and things like that.
• 27:35
  SPEAKER_00: A lot of the stuff is out of your control.
• 27:38
  SPEAKER_00: Just kind of focus on what you can control um and what makes sense for your team is also something I would just caution everybody to do because it could get overwhelming pretty quickly.
• 27:50
  SPEAKER_02: Yeah, no, I agree.
• 27:52
  SPEAKER_02: Um, it's always good to get people in the mix, right?
• 27:55
  SPEAKER_02: People in the loop and be able to bounce these ideas and you know, again, with the ever-changing terminology and tools and agents and everything.
• 28:05
  SPEAKER_02: Um, you know, surround yourself with people that know and don't be afraid to ask questions.
• 28:10
  SPEAKER_00: Yeah, and don't be afraid and take everything with a grain of salt.
• 28:14
  SPEAKER_00: Because again, a lot of it is marketing hype.
• 28:16
  SPEAKER_00: Uh, but just, you know, you can't pretend like it's just use caution, but um don't believe everything you hear, too.
• 28:27
  SPEAKER_02: You know, it's funny because it's almost like uh, you know, we have mutual friends.
• 28:32
  SPEAKER_02: I'm not gonna out them, right?
• 28:33
  SPEAKER_02: But they see one of these marketing campaigns on TikTok about how AI is taking over the world and you know how it's gonna end the days here soon.
• 28:41
  SPEAKER_02: And then uh they share with me and say, What are you doing with your life?
• 28:45
  SPEAKER_02: Right.
• 28:45
  SPEAKER_02: And it's like, what?
• 28:49
  SPEAKER_02: Well, this went from zero to a hundred real quick.
• 28:52
  SPEAKER_02: I mean, um, and and that's what marketing is, right?
• 28:56
  SPEAKER_02: It's the conversations that they can start, it's the clicks that they can view.
• 29:00
  SPEAKER_02: Um, but at the end of the day, you know, are we at that point, that tipping point?
• 29:04
  SPEAKER_02: I don't think so.
• 29:05
  SPEAKER_02: But, you know, there are folks that seem to to believe the hype and believe the marketings.
• 29:10
  SPEAKER_00: Yeah, I mean, I think there is some legitimacy again, like I mentioned too, some of the stuff we do need to have tighter regulations.
• 29:17
  SPEAKER_00: I think we do need to have, you know, um more controlled pacing, things like that.
• 29:22
  SPEAKER_00: Um, but it's also not going away.
• 29:25
  SPEAKER_00: So that's another thing that I think leaders need to understand and have a responsibility to your team.
• 29:30
  SPEAKER_00: And that's part of why I kind of started Modern Worker, Adam, and you too, is like there's a lot of fear out there.
• 29:36
  SPEAKER_00: And and and AI is not going away, but how can we help people feel more at ease with it, adopt it responsibly?
• 29:43
  SPEAKER_00: Um, and so as a leader, like you can really kind of help your team navigate through these pretty tumultuous times and you know, weed through some of the stuff, you know, focus on what you can control.
• 29:55
  SPEAKER_00: Um, you know, I think that's really important because again, your employees are probably scared.
• 30:00
  SPEAKER_00: Uh, you're probably scared, you know, there's points in the I I am like, ugh, it's a little dicey.
• 30:08
  SPEAKER_00: Um, but you know, I think that's with any new technology, like there's fear and skepticism and things like that.
• 30:14
  SPEAKER_00: So I think it's natural, but um, you know, uh as a leader, you got to help your teams navigate through that.
• 30:21
  SPEAKER_02: Yeah, yeah, you're absolutely right, Chevelle.
• 30:24
  SPEAKER_02: I couldn't have said it better myself.
• 30:26
  SPEAKER_02: So uh, well, Chevelle, I appreciate you jumping on with me.
• 30:29
  SPEAKER_02: Uh pleasure as always to speak with you.
• 30:32
  SPEAKER_02: I know our listeners are gonna be thrilled to hear about AI literacy, um, just as I was.
• 30:37
  SPEAKER_02: So, as always, thank you.
• 30:40
  SPEAKER_00: Anytime.
• 30:41
  SPEAKER_00: It's a great topic.
• 30:43
  SPEAKER_02: Until next time, Chevelle, I'll talk to you soon.
• 30:47
  unknown: Thank you.
• 30:47
  SPEAKER_00: Thanks.
• 30:48
  SPEAKER_00: Take care.
• 30:48
  SPEAKER_00: Bye.
• 30:49
  unknown: Bye.