What is AI? A Simple Explanation for Beginners | AI for Beginners Series

Sheridan Wendt (00:00)
I'll tell you this I have literally trained meta to show me business and success type content, right? Like literally every ad that I see is like here's a new way to make money online, right? Whoa, right?

Jon Foster (00:09)
Mm-hmm.

Yeah, same here with like the technology part,

like just a little dash of Bigfoot thrown in.

Sheridan Wendt (00:21)
Little dash of Bigfoot. I love it. All right.

Sheridan Wendt (00:25)
Welcome to the Answer Engines podcast. I'm your host, Sheridan Wendt

today I have my longtime friend, Jon Foster, here as a noob to participate in this series that we are calling AI for Beginners. And Jon is new to AI, so he will hopefully be asking all the questions that you might have in your head so I don't gloss over anything. Jon, can you tell them a little bit about yourself?

Jon Foster (00:51)
Hey everybody. My name is Jon Foster. I got into IT when I was 18 years old, was in the United States Navy. After getting out though, I went to school, got my bachelor's degree in cybersecurity, and that's kind of a little reason why I'm kind of new to me. Like, AI is scary. You know, personally, myself, probably to you. Pretty much I think that what I'm going to learn today, I'm hoping that it actually gets me

a little less paranoid and a little more open to making my life easier.

Sheridan Wendt (01:23)
Yeah, yeah, yeah, so.

Today we're going to be stripping away the hype and getting clear on what AI actually is and what it isn't. Let's start at the very beginning with what does AI even mean? So AI stands for artificial intelligence. And the idea here is that machines are performing tasks that normally require human intelligence. And there are three general categories for this. There is automation, there's machine learning,

And there's generative AI. And then that all branches out into even more when you want to dive into those, Automation, we're talking about just things that have rules, right? Rule-based automation, maybe firewall rules, in Jon's case. I bet you're familiar with those, right?

Then when it comes to machine learning, is machines actually learning from data. Great example of this is your Netflix preferences, right? The way that you are browsing on Netflix, the way that you look at different videos, when you rate them, when you watch them, if you click that up thumbs up or thumbs down button, that tells Netflix what your preferences are. And then Netflix can look at an individual

movie and take some of that attributes the data about that movie and say oh this is a Movie that's kind of dark and brooding. This is a light-hearted movie. This is a comedy. This is a romance, right? Jon Foster's Netflix feed is probably full of Rom-coms if I had to guess what do you what do you say Jon?

Jon Foster (02:57)
No, no, no wrong comes mostly just like yeah, I mean there's a little bit sprinkled in there You know nene, you know my my fiance she's you know kind of into that romance stuff But I mean like stranger things like mine's pretty pretty basic. It's pretty generic the like the top ten You know what's being watched is kind of where I get led

Sheridan Wendt (03:00)
No rom-coms?

Alright, actions, thrillers, and sci-fis, huh?

Jon Foster (03:23)
Yeah, kinda, ⁓ Mysteries.

Sheridan Wendt (03:25)
Okay

So netflix is learning What type of stuff that you like to watch and that's what pops up in your top 10 That's what pops up in recommended for you, right? And so that is machine learning then there's generative ai and generative ai is when Something new is created new text a new image a new video, right? a new summary of a blog post that you

pasted into your favorite AI model, your favorite LLM, and it asked it for a summary, right? The summary is new. Even though the information was already there because you pasted it in, the summary wasn't there before, and that was generated by AI. So that's generative AI.

Jon Foster (04:10)
Okay, so like when I'm on like social media and like I see all these videos like with Bigfoot and like the other crazy things like that's that's AI kind of like You know making that up on the fly

Sheridan Wendt (04:24)
First of all, have not, I don't know what social media you're on, but I'm not seeing videos of Bigfoot. So Facebook's been doing some machine learning about you. Somebody has.

Jon Foster (04:30)
I gotta send them to you.

I mean like

I know I'm bearded and I'm hairy, but I'm not that hairy like it's not I don't know the correlation But you know I mean maybe maybe an outboard Z. I don't know

Jon Foster (04:45)
Did you really not, like, see any of the Bigfoot videos?

Sheridan Wendt (04:49)
I mean, dude, everybody has seen like one or two Bigfoot videos, but they're not like pervasive in my social media feeds right now, you know, in 2025.

Jon Foster (04:59)
I mean, like, I don't know when you started watching them, but like 2025 is kind of what I have. I watch literally every freaking one. That's probably why freaking AI thinks that, ⁓

that I want more of it, I guess.

Sheridan Wendt (05:12)
Probably is man. It probably is. I have

I'll tell you this I have literally trained meta to show me business and success type content, right? Like literally every ad that I see is like here's a new way to make money online, right? Whoa, right?

Jon Foster (05:24)
Mm-hmm.

Sheridan Wendt (05:31)
So I don't know I get a ton of ideas for how to become a better Technologist and a better business person just from the ads that I see, you know, it's great

Jon Foster (05:41)
Yeah, same here with like the technology part,

like just a little dash of Bigfoot thrown in.

Sheridan Wendt (05:46)
Little dash of Bigfoot. I love it. All right.

That is a great time for us to break to our sponsors and friends over at Advantage Labs

Sheridan Wendt (07:03)
And we're back. I'm your host Sheridan on the Answer Engines podcast and this is Jon Foster. We are here today talking about what is AI, what is it not, and why does it matter?

it's possible that the videos you're talking about were generated by AI. Somebody went into a tool and said, make me a video of Bigfoot, right? And then they posted on social media. So, possibly, I don't know, it's hard to say.

Jon Foster (07:32)
Well, this Bigfoot's

real, you know?

Sheridan Wendt (07:35)
It could be a costume. Who knows what it is, you know, special effects on a video instead of AI. But, ⁓ the thing about the thing that I want to kind of clarify is what AI is doing is recognizing patterns. it's predicting it's learning from data. when you actually prompt AI, well, we'll get to that later, right? We'll talk about exactly what happens when you put a prompt into an LLM like Claude.

or chat GPT or deep seek, right? ⁓ So I do want to clarify while we're here though, that AI doesn't think and it doesn't feel, it predicts. It's predicting based off data, based off patterns. So you don't need to be correct. Yeah, that's it. Don't need to be technical to understand what it's doing. It's just looking at words and saying, I think this is the next word in the sentence based on math, right?

Jon Foster (08:16)
Pattern recognition.

Sheridan Wendt (08:28)
So what like, Jon, what's like the first thing that you think of when you think of AI? Besides Bigfoot.

Jon Foster (08:34)
Besides Bigfoot? Okay. First thing I think about when I think AI. definitely, definitely the scare factor. It's definitely the vulnerabilities and taking away jobs, stuff like that. know, making...

Sheridan Wendt (08:35)
I think we're good.

Man, you're real

doom and gloom guy.

Jon Foster (08:51)
I'm not trying to be. Well, there's some good things too, like, you know, being able to streamline simple tasks and stuff like that. Like, I don't really know how to implement much of that, but you know, I mean, I'm going to assume you're going to tell me at some point.

Sheridan Wendt (09:06)
Yes, I mean you do raise some valid concerns, right? I'm not trying to just dismiss them, right? There are job displacement is a real risk and Information security is a big risk, right? So I don't want to downplay them because they're very serious things and we're gonna talk about them But

going to do that on a later episode, because right now I want to come back to AI in our everyday lives. What is AI really? So we talked about automation and machine learning and generative AI. So if you have used Spotify or Gmail or chat GPT, you're already using AI. Spotify is recommending

Jon Foster (09:38)
Yeah.

Sheridan Wendt (09:45)
music for you based off of your listening preferences, just like Netflix is recommending movies based off of your watch history. So, Jon, have you seen, like, how have you seen some examples of AI creeping into your daily life in ways maybe you didn't expect?

Jon Foster (10:01)
besides Spotify thinking that I like Taylor Swift.

Sheridan Wendt (10:05)
But I mean, does Spotify think that? Why does Spotify think that is the better question?

Jon Foster (10:08)
I mean, they, then they, then they hops on to,

you know what mean? Like it's not, so apparently AI thinks, you know, my, girl is, is me, but anyway, um, I think, I think like mainly, mainly for me, I think I see it on YouTube. Um, like with the videos and stuff like that, that I watch, like a lot of stuff in like the tech field and like, you know, stuff like that, like dark diaries on like Spotify, like kind of like, I'm very, very like in my.

Currently right now I'm trying to stay up with the current technological culture. So, I can see that popping in there quite a bit.

Sheridan Wendt (10:46)
So recommendations on things to watch or listen to.

Jon Foster (10:50)
Yeah, pretty much. For me.

Sheridan Wendt (10:51)
Okay.

All right. Well, let's talk about what AI can't do, right? Talk about some common misconceptions and create some healthy expectations for what we can expect. Number one, it's at least for now, AI is not conscious or self-aware. It doesn't know or understand things, right? There are

Jon Foster (11:14)
Mm.

Sheridan Wendt (11:17)
there is a difference between being able to predict patterns and understand the data that it's been trained on and knowing something or understanding something. So it kind of seems that way when you ask it a question and it gives you an answer, but it's just been trained on that question and those answers before. Let's talk about hallucinations, right? Sometimes you ask a question to a model

Jon Foster (11:30)
Mm.

Sheridan Wendt (11:41)
and it gives you a wrong answer and it seems so confident that this is the right answer. Have you ever seen people post about that or seen it happen to you?

Jon Foster (11:49)
I see humans do that all the time on social media.

Sheridan Wendt (11:52)
Right. When AI does it, it's called hallucination. and it happens sometimes it gives a wrong answer. And there are a lot of people who have different opinions on how that happens, why it happens. it's because there's so much information on the internet that these models get trained on. It's hard to filter out incorrect things sometimes. and.

There may be other reasons besides that, right? But that is one of the reasons is the data that we train these models on has to be good data, right? Good data in, good data out. Bad data in, bad data out. So, the next thing is AI doesn't have emotions. it doesn't always have context and accept the context you give it. some other limitations might be ethics, maybe human judgment, right?

Any of those you want to talk about, Jon?

Jon Foster (12:43)
Well, I don't know if you're describing AI or ex-girlfriend, like, you know, I'm playing, playing. I don't talk bad about my exes. I don't. I'm just playing. No ex came to mind. But, you know, what I've noticed, say again?

Sheridan Wendt (12:47)
Hahaha

Ha

Everybody

go stalk Jon on Facebook, figure out which ex he's talking about.

Jon Foster (13:01)
yeah, right.

Well, no, no, the current one made me delete all of them. So don't worry. You won't find them. But yeah, but what I'm, what I'm starting to realize is, is, I mean, just like if you feed anybody wrong information, if they don't do their proper research and stuff like that, there's not enough data to, to, you know, debunk that.

Sheridan Wendt (13:09)
There you go.

Jon Foster (13:27)
People, humans believe this. AI, AI kinda is no different than a human. You know It makes errors too. You know what mean? With bad information.

Sheridan Wendt (13:37)
So that's something that you've heard is that it's not different from a human or you're saying it's not.

Jon Foster (13:42)
No,

no, what I'm picking up, you know what mean? Because when you described how AI actually operates and how, you know what mean, the hallucination that you said, you know I mean? It just, it sounds kind of like, you know, kind of what happens to, like, you've been told things like, like, ⁓ Simbad was in the,

Sheridan Wendt (13:53)
Right.

Jon Foster (14:03)
the movie Shazam, the Mandela effect, you know? You've heard that, you know what mean? I'm even comparable to that as well because I also thought Sinbad was in that movie as well. things like that, know, just misinformation and it believes so strongly that it is right. You know what I mean? So it's...

Sheridan Wendt (14:06)
Sure.

Jon Foster (14:25)
It's kinda like a human but not, but if you really think information in our brains is just us recalling information, it's stored, you know what mean? Or calling a friend to find out information and I guess that's how I would associate it with like when it refers and taps into its information on the internet.

Sheridan Wendt (14:43)
Right, right, okay.

I think I understand what you're trying to say though. You're trying to say that any given AI model is only as good as the information that it was given or that it was trained on, just like a person. You know, we might grow up our whole lives thinking that the world is flat.

Jon Foster (14:54)
Yeah.

Sheridan Wendt (14:58)
Ha ha ha.

Jon Foster (14:59)
Yeah,

right. Well, they did. They did at some point.

Sheridan Wendt (15:01)
Right, right. So, okay, let's talk about misconceptions or fears, anything that you've heard about AI that you want to talk about.

Jon Foster (15:11)
Well, it's really good that we're doing this right now because if you you if you've been on the internet, you've you've learned or read, let's say this, we got to do air quotes because we can't really trust the whole entire validity of the media sometimes. And these things that you read don't ever read every don't believe everything on the Internet kids. But like there was one situation where

They put self-learning into the AI algorithm, from my understanding. You know what mean? And they realized that it was learning too quickly and had survival mechanisms kind of ingrained in it to where it started blackmailing its creator that wanting to turn it off. And there was another one talking about how AI killed somebody, but...

You know, I like I said, like there's a lot of misinformation, propaganda spread out throughout the internet and it could just be bad stuff, you know, on the internet talking crap about AI, you know, but these are, these are things that are like people are hearing, you know,

Sheridan Wendt (16:18)
Right. Right. So, well, let's talk about that. Let's, let's go one at a time. Let's start with the situation where AI is trying to blackmail the, the creator, the developer, the, engineer who was working on it, right. To prevent him from shutting it off. that is definitely concerning, right? if AI, if a model can have survival instincts, then

that kind of changes things, right? It kind of goes back to the point where you said, what are we gonna connect this to? What data is it gonna have access to? What about information security? What about privacy, right? So I would say definitely concerning for people to think about that. And a lot of leaders in AI have been calling on regulation, right? From governments, from leaders around the world.

Jon Foster (16:45)
I

Mm-hmm.

Sheridan Wendt (17:11)
to say we need to really think about how we want to build these tools and how we want AI to work and what we want it to have access to and what we want its capabilities to be. And I think those are, I mean, those are good concerns to have, good questions to have, and I think people are still gonna be debating it for 100 years.

Jon Foster (17:32)
You know, it doesn't help like the Terminator movies like Skynet, stuff like that, like that. The AI fear was even put into place before even, you know, modern AI even existed. You know what I mean? But, you know, the scary part is, is that if it is self-learning and it does have those capabilities, then that could be a potential future. That is something because if you think of the human condition, we are the most destructive force on this planet.

Sheridan Wendt (17:53)
Right.

Jon Foster (18:02)
you know what mean? It would only make sense for it to be like, you know what? You're not needed anymore. But I think if we put proper regulations and security controls into place, you know, I think AI can be a great thing.

Sheridan Wendt (18:17)
Yeah, yeah, if we don't connect it to everything then and it only has access to some things then it's a whole lot safer, right? So we really got to think about that

right. I want to shift gears a little bit and talk about why All any of this matters for everyone, right? and the first thing is I think AI is the next wave of

digital transformation, right? Just like the internet or the mobile phone, right? AI is what's next. It's the next big thing. And it's happening now, right? It's already starting.

Jon Foster (18:49)
And it's not going away from, I don't think so.

Sheridan Wendt (18:53)
No, it's definitely not going away. It's just getting better every day. The transformation and the adoption rate of AI, if you look at some curves, has just been crazy. And its capabilities are getting better faster than we expected it to. There's a great... Well, I'll save it for another time. There's a great report I'd like to review on some of those capabilities. But another great thing about AI is you don't need to be a developer. You don't need to code. You don't need to...

go to school to learn how to use AI, right? What do you have to do?

Jon Foster (19:24)
You can probably just look it up, you know, or, you know, just listen to this podcast once we're done. Oh, I mean, yeah, I didn't even think about that. Yeah, could just, hey, how do I use you? It'll probably tell you.

Sheridan Wendt (19:25)
Yeah?

You just talk to it. Yeah, no, mean, listen to the podcast. Definitely.

No, seriously. So to use AI, you type in a question, you type in a sentence, you type in what you're trying to achieve. You don't even have to type in, you can speak to it and say, hey, I want to start a business about this, right? I want to start an online business, help me get started. And it'll help you put together a plan. And it'll tell you, it'll even offer you suggestions on would you like me to help you do this step next, right? So

You don't have to code. You don't have to go to school to learn how to use AI. You can certainly get better at using AI the more you learn about it. But yeah, lots of people are out there using AI right now to save time in their business, to get better at writing, to create more creative works. So I think... ⁓

It's just gonna get better. It's just gonna get more ingrained in our society and a lot of people are gonna start using it.

Jon Foster (20:35)
I've been using a little bit myself and it's a little weird, a little weird because it sounds human. You know what When you talk to it, it talks back to you and it's not like I am a robot, you know, kind of deal. But it's more like, you know, I'm just having a conversation. I'm like, man, if I do this long enough, I may think this is like a friend of mine or something.

Sheridan Wendt (20:56)
Right. Yeah. Yeah. Some people have gotten very attached to or close to some AI girlfriends, right? We'll get into that. We'll get into that in another. It's a thing. It's a thing. We'll do, we'll talk about another episode, right? So the, the, main point here is that winners in this new era aren't the most technical. They're the most creative.

Jon Foster (21:04)
⁓ that's a thing? Is that a thing?

lord.

Mm-hmm.

Mm.

Sheridan Wendt (21:21)
So anyways, now that we've kind of talked about what AI is and what it isn't, in the next episode, we're gonna be talking about where AI can deliver real business value.

So a few quick...

takeaways from today's episode. AI is actually pattern recognition and prediction. It's not magic. It's not human. It's not thinking and it's not feeling, right? There's a few different types of AIs. They serve different purposes. And another big takeaway, humans and AI are stronger together.

Jon Foster (21:43)
Mm.

Sheridan Wendt (21:53)
in episode two, we will talk about the real business value of AI, where we connect AI to measurable results.

appreciate you all for listening. I'm your host Sheridan. We will see you next time