Welcome to the Techstrong AI Podcast – Techstrong AI Podcast EP1

Hello and welcome to the Techstrong AI podcast. I'm excited to bring this to you today. This is our inaugural podcast.

I'm Amanda Ani, and with me is Mike Baard. And we're excited to bring you the latest and greatest in ai. So with that, I'm gonna hand it to Mike.

These are exciting times and almost every day something new happens, but I think we need some context. So hopefully with this podcast, we're gonna walk you through not just the big events of the week, but what it means and give you guys some sense of what can you act on, what's theory and most importantly, what's real. Hey, Amanda, as always, good to see you again.

Kick Us off. Good to see you too. Good to see you too, Mike.

Uh, well let's start with some of the AI governance issues of the week. I know President Biden and his executive AI order has been under a lot of scrutiny and um, the EU has also been putting together some things in regard to ai. So can we share a little bit about that?

It seems to me that bureaucrats have learned how to spell ai, but that doesn't mean that they necessarily understand how it works. And I wonder if some of these things are gonna be overly prescriptive and who helped them actually craft these regulations. Because while I'm all for regulations, I think we need them.

I just wonder sometimes if the folks who write these things don't necessarily understand how the technology works. And we're gonna run into issues down the way that are, well probably gonna make lawyers a fortune as we sort all this thing out. But what's your sense of what's going on here and, and will the US really pass something?

'cause right now we're looking at this and it's 50 different states with 50 different regulations and hopefully that will all get unified. But is that gonna happen this year, next year, next decade? What do you think?

Well, I don't think it's gonna take a decade, that's for sure. I mean, I think they're looking at this on a monthly basis and uh, we have a lot of the big players in, um, Google and Amazon and, um, various other players that are very invested and interested in this as well. Um, so they're, they're putting in their 2 cents and I think I would say maybe in a year or two there's gonna be a better sense because this technology is evolving so rapidly, they've really gotta get a handle on it or some sense of it and understand.

Um, especially in regard to DeepFakes, that's really the biggest issue right now is, um, cybersecurity and DeepFakes and how to, um, and also copyright issues and things like that. Um, so getting a handle on those. So I don't think it's gonna take too long, but I know there's, there's a lot of concern on being able, like there's really not real government um, regulations right now.

It's kind of very, um, uh, just broad sweeping and so everybody's trying to figure out, um, how to, um, really put some laws in place. Um, and I know the FTC also is, um, invested in this and they're looking into partnerships between some of these big companies I just mentioned. Um, and how that's gonna play out and um, if it's um, uh, meeting all the laws and requirements there.

So Let me tell you what worries me about all of this approach though. If I look at other industries, there's a nasty habit where the vendors come together and they create some sort of position paper. They hire a bunch of lawyers in Washington, typically lobbyists that live in a place called K Street.

And they cut and paste that document and shove it into a bunch of legislation and stick it under the nose of a representative or a senator and they sign it and away it goes. I think we need a better review process for ai 'cause there's just so much more at stake. It's a lot different than, uh, maybe, you know, a couple of regulations as to how we're gonna manage a chicken farm or whatever it might be that we have previously passed.

But I feel like this process may we need some more expertise at the government level who's gonna show up and kind of review these things. 'cause otherwise it's gonna be business as usual and we'll all be sorry later. Well, I wonder if that will open up some positions really of like what you just said at the government level.

There probably needs to just be some positions open up for people who are really knowledgeable with AI and its use and bring together a team and that's who they should be, um, referring to as far as laws and governance, Well, I'll tell you when I think everybody's gonna start howling about this is in this current election cycle, once they discover that AI is being used to manipulate votes and distribute disinformation, and of course that may be in the eye of the proverbial beholder, but politicians are gonna suddenly be like, wait a minute, this is directly impacting me. And suddenly they'll be paying a lot more attention to how this works. And uh, some cases they'll try to abuse it.

In other cases they will try to put the right guardrails in place. But do you think that AI is gonna be a major issue this election cycle? You know what, it's funny that you bring that up because I was very surprised.

I actually, um, had a doctor's appointment this morning and I was talking to the nurse and she was saying how, um, more and more people come in and they're talking about AI and um, before me there was a teacher, um, but she was like a 70-year-old teacher and she was all about AI and all her concerns and how she was afraid for our children and it was just gonna ruin our younger generations and all, all of this. And, uh, um, so it's surprising how many people are so hyper-focused on ai. Um, and yesterday I, I actually met with a friend and um, I was telling her all about the amazing use cases of AI for marketing and all kinds of stuff, and she scrunched up her nose and was like, no, I'm not gonna go there.

I don't think so I I do not trust ai, but she knew, you know, she had done a lot of research. So I know that people across all sectors and ages are very hyper-focused on ai. It's very surprising.

I wonder if people will stop believing whatever it is they're seeing. 'cause if they disagree with it, they'll automatically assume it was manipulated by some algorithm somewhere and they'll just basically, um, decide to go with their gut and whatever inherent biases they already have. And we'll have this situation where no one's gonna believe any content that they see unless they know precisely where it came from.

On the other side of that question, though, I could say positively, maybe people will be more skeptical of the content that gets presented to them. There's a lot of junk floating around on the internet today that people buy into and should be more suspicious of in the first place. So, um, what do you think?

Are we gonna be worse off or better off? I, I think in the end we are gonna be better off. AI is just amazing.

All the use cases, um, and how it's already infiltrated the enterprise in so many ways. Um, and now, you know, everyday people are utilizing it and I'm seeing class after class on AI in a variety of use cases and everybody wants to know how to use it. So I think in the end it's, it's gonna be embraced and it's gonna be a good thing.

But there is a lot of skepticism and concern and I mean, rightly so. I mean, what was it just like a couple of days ago, um, when Twitter had to stop the search for Taylor Swift, um, because of, uh, terrible DeepFakes. Um, and so I know there's gonna be, there is a lot of concern, um, in believing what you're seeing and I'm really not sure how that's gonna be addressed.

I know there's some watermarks that can be used and things like this, um, but really determining what's real is gonna be an issue and there are gonna people that aren't sure what they can trust and, and there is a lot of distress in the election this cycle especially, um, like worried about tampering, um, using AI to tamper with votes and, um, believing if the votes were accurate and going through that whole thing again. Mm-Hmm. Well it'll be interesting to see, 'cause you bring up watermarks and there's other people out there who are saying that they're gonna create AI models to monitor the AI models essentially.

So they're gonna use that to detect whether content was created using an AI model or not. But I ask you frankly, sometimes the content that the machine creates is more accurate, maybe even better written and certainly more engaging than what some of the stuff that humans are creating. So where do I draw the line between what is better and worse versus whether it was created by a human or a machine?

Because let's be honest, humans are creating a lot of crap out there. True. And, and also a lot of this software, I, I mean, I see, you know, I manage a lot of the content, um, as it comes in and it'll come in with software that, you know, apparently shows if it's authentically written by a human or, um, could have been written by ai.

Um, but there's a lot of problems with that software too. I mean, it, there's a lot of people complaining that they did authentically write these articles and the software is saying written by ai. So I mean, how, how, is there really a way to determine it?

I mean, um, and, and yeah, oftentimes it's better, uh, some things you need a little help from AI and it can really help tweak or improve the content. Um, it might just have a lot more knowledge, um, that you can grab in seconds. Um, but then again, it's, there's also hallucinations and made up stuff too.

So it's, it's just, it's kind of in a limbo stage I feel like. Well the whole thing seems to be revolving around whether or not it's fair use for generative AI to use the content in the first place. Um, authors and other folks have shown that uh, generative AI platforms are essentially copying their works and using it in some form of output that they're sharing with folks without any kind of attribution to it.

And it's pretty much one for one. But the folks who build a platform say, Hey, it's just fair use. We're scraping all this stuff of the internet and we're using it to create some other output.

And there were court cases about this issue that go everywhere. Some folks that I've talked to say, Hey, this is just a variation of the same issue we had about the debate over code. It's free speech at the end of the day.

And you can't really prevent people from uh, using it. And because you can't necessarily apply a copyright to that level of detail. I don't know, I'm not in a lawyer, but we'll see how this plays out.

But I suspect that we probably won't see these cases resolved until what the middle of the decade and that can create all kinds of interesting problems. What do you think? No, it definitely is gonna be a problem.

And we've seen like with the Rider strike and um, various other strikes, um, in this regard. And then we see others that are embracing it. So, you know, you see a lot of people that are embracing it and saying, yeah, here, use ai.

You can use my voice for all your creative stuff. And um, and then people very much unhappy about it. Um, and I don't know, yeah, it's gonna take a while to figure out how to even move forward and, and how to keep content copyrighted or, uh, it's just very hard now that everything's so open, Open and I'm sure that Microsoft means it when they say that they will indemnify folks for using content created from their platform.

And depending on the court case, we'll see how that plays out. But here's the part that I think we're overlooking. Let's say that I did create an application using code that was created by generative ai and then the court case comes along and you know, my side loses well, I may not have to pay the cost of that suit because Microsoft will pick that up, but I'm gonna have to go find all that code and rip it out and then I gotta replace it.

And the same thing if I created content or some sort of video or article, I'm gonna have to take that down and replace that. So the effort around this is gonna be considerable. And I think a lot of folks are in larger enterprises at the C level are having a moment here about cause for pause and do we need to think twice about where and when we're using this?

So I think there's plenty of use cases that won't involve that level of copyrighted content, but when it comes to that space, I think maybe caution will prevail. You agree, disagree? I agree.

Because I also think at the end of the day, if that is the case, is it really likely that they're gonna be able to find all of that code and all of it and and remove it? I mean, it's kind of a runaway train in a way, so it's gonna be very hard to go back and get it all off. I also wonder at the end of the day whether or not this is the year we're gonna see all of these wonderful benefits of gen ai or is this more likely to be the year where we operationalize AI in the sense that we're getting our processes together, we're figuring out how to use it, and a lot of it will be experimentation and we're playing around with this stuff, but we may not see it in mainstream processes from corporations tool when 2025?

Or are we there yet? I guess if, if I was mimicking the kids in the back of the car, what do you think? Well, um, in my discussions with executive leaders and the articles that I'm seeing come through, I mean, leaders are still trying to figure out how to implement and utilize this technology and wrap their heads around all the use cases for it.

And, um, change management is a real issue, um, in this regard. Um, so it is gonna be a little bit, it's gonna take a little bit. Um, but I think, um, I mean it's happening so fast like from a year ago till now.

I just, I I think within the year we're gonna see rapid advancements. I think there's also two sides of this. There's the projects that get funded by the IT organization and then there's the projects that get funded by the line of business that are not always one and the same.

I think the IT folks are more cautious 'cause they know where the data came from that's being used to train these models. I think the lines of business are about to discover that perhaps much of the data that they've been planning to use to train these A models is more circumspect than they would like to admit, even though they know how it was created in the first place and are actually probably the root cause for why there is so much data that's conflicting and just flat out wrong. Are we gonna have a moment here where maybe data management becomes the hot topic of 2024 because people are realizing, hey, no matter how smart something is, garbage in garbage still comes out.

Well, I think data management is the hot topic of this year. I think you have that right? Because I'm seeing so much content in, in this topic, data management and, um, how to get quality clean data, um, and how to harness it.

And um, so I think yes, that's definitely the case. And of course operationally operationalizing ai, that's a ton mouthful. Um, you know, we, that's why we have our new building with AI site on tech strong AI where there's a lot of articles, um, so people wanna actually use it.

Um, so yeah. All right. Well folks, this kind of brings us to the end of this particular episode as a man that just kind of noted there's a lot of articles on text drawing AI that you can check out that relate to this conversation.

They're all in the show notes. We invite you to go check those out. Amanda will carefully put those together for you guys.

And with that, we'd love your feedback. We hope you enjoyed this podcast. We plan to have, uh, many, many more of them.

And Amanda wanna say some parting thoughts? Absolutely. I, it's very exciting AI and I'm glad that you all are here listening to our first podcast and stay tuned because we plan to put one out on a steady cadence every week.

We'll see you all next time.