[00:00:02] Speaker A: Did you know the AI you're using right now can't actually think, but still is smarter than most humans at specific tasks? And here's why this matters. Understanding what AI actually is will completely change how you use it, how you trust it, and how you protect yourself from it. Cue the music.
[00:00:31] Speaker B: From the natural state. This is Public Facing the Podcast with Joe Holman.
[00:00:42] Speaker A: Welcome to Public Facing the Podcast. I'm Joe Holman and today we're kicking off a five part deep dive into AI and innovation. Whether you're a tech enthusiast, a worried parent, or someone who just wants to know what's exactly happening on your phone, then this series is for you. Now, confession time. When I first heard machine learning, well, I thought it was robots doing homework or something to that effect anyway, but spoiler alert, it's way cooler and way weirder than that.
Today we're breaking down the different types of AI, how they work and why it matters for your everyday life.
Think of it as AI101. No computer science degree required, and nerds are still allowed. Here's our roadmap. We'll explore narrow AI versus general AI. We'll dive into how these systems actually learn and finish with a real world look at where this technology lives in your daily life.
Our main takeaway knowledge is power. So especially when. When the technology feels like magic. All right, let's get into it.
Now, the AI you already know.
All the current AI is narrow AI and it's designed for specific tasks. Narrow AI or weak AI, this is every AI existing, like Siri, Alexa, Chat GPT, Netflix, recommendations even.
And each is trained to do one thing really well.
Here's a fun fact.
Chat GPT can pass medical licensing exams, but it can't tell you if it's raining outside. Without Internet access, it's brilliant and limited.
Now, general AI or AGI, this is the sci fi dream. It's AI that can learn anything like a human.
Companies like OpenAI are racing towards this, but experts estimate that we're still years, possibly even decades away.
Now, super AI is purely theoretical at this point. We're talking AI that's smarter than all humans combined. Cue the Terminator jokes. Right, but let's not just panic yet.
Basically, your smart speaker is really good at playing music on command. Yes, but ask it to fold your laundry. Oh, now, now, it's a complete and total failure.
We're safe from the robot overlords for now.
And today's AI is incredibly specialized. It's a tool, but it's not yet a replacement for human Intelligence.
So now that we know what AI is, I want us to take a look at how it actually learns.
Now, I think a key here is that AI systems, they learn through data patterns, not through programming rules. When it comes to machine learning, instead of programming every rule, it is fed AI examples, and it finds patterns within those examples. Think of teaching a child to recognize dogs by showing them thousands of dog pictures.
Then comes deep learning.
So this uses neural networks, mathematical structures inspired by the human brain, and layers upon layers of processing create remarkably sophisticated outputs.
And that training data. How do we train the dragon? Right. This is the secret sauce. AI is only as good as the data that it learns from. So if there's garbage in, there's garbage out, and then there's that whole hallucination problem.
That's right. AI can confidently generate completely false information because it's predicting patterns and not understanding truth.
And when it came to IBM's Deep Blue, well, it beat the world chess champion back in 1997 by analyzing the board and predicting outcomes.
But ask it to play checkers.
Now it's totally clueless.
Think of it like being a PhD in one subject but failing kindergarten in another.
An AI doesn't understand anything. It recognizes patterns.
And this is both its superpower as well as its biggest weakness.
There is another side to this public facing. The podcast.
When it comes to the AI in your pocket, yes, the AI that's already woven into your daily life in ways that you might not even realize.
Look at text modality using ChatGPT and translation apps, even email suggestions, those that we use in Gemini or in Copilot or even in Quad as well.
Now in image generation, using Dall E and midjourney, creating photorealistic images from text descriptions.
And those multimodal AI. These are systems like GPT4 that can process text, images and audio simultaneously.
We knew it was coming when we started switching over to electronic medical records.
Health care applications.
AI is helping doctors spot cancers, predict patient risks, and personalize treatments.
And when it comes to education revolution, well, AI tutors are providing personalized learning and at scale.
So with this evolution, since 2022's ChatGPT launch, we've seen over 100 million users take on the fastest technology adoption in human history.
And the genie, well, it's not going back into the bottle. AI is here.
It's evolving faster than any technology before it. And understanding its capabilities and limitations is no longer optional. It's essential.
Now, in our next episode, we're going to tackle the big question that everybody's asking.
Who's actually in control of this technology because, spoiler alert, the answer might just surprise you.
And that's going to bring us to today's final word.
In today's final word, all the AI today is narrow AI, specialized tools. Yeah. Not generalized intelligence.
The AI learns from patterns in data, not from true understanding.
So this technology is already transforming health care, education, and our daily life, whether we're paying attention or not.
But ultimately, AI is a mirror reflecting human knowledge.
Brilliant, flawed, and constantly evolving.
That's today's final word, folks. If you've enjoyed today's episode, make sure to follow and share Public Facing wherever you listen.
Email us at hellobublicfacingpod Stream and visit us also at publicfacingpod Stream for all of our episodes. And that's gonna do it for our time together today.
If this episode has helped demystify AI at all for you, please do me a favor. Share it. Send this to a friend, send this to a colleague, and make sure you follow and stay connected to Public Facing the Podcast for more conversations about the people, issues and everyday moments shaping our world.
Thanks for spending time with us today. Until next time, stay informed, stay encouraged, and stay public facing.
For Public Facing the Podcast, I'm Joe Holman. Come back and see us again soon.
[00:10:06] Speaker B: Thanks for listening to Public Facing the Podcast with Joe Holman. Visit us online at PublicFacingPod stream and on
[email protected] PublicFacingPod.
[00:10:24] Speaker A: Sam.
