American Data Privacy and Protection Act – Mike O’Malley, SenecaGlobal

This is Textron TV. Hey everyone, welcome back to techstrunk TV. I'm happy to have as our next guest on Tech strung today TV coming to us from Chicago where it's a balmy 50 degrees and it's but they actually have had a pretty mild winter compared to a lot of you folks out here.

It's Michael O'Malley and Mike is with Seneca Global Mike. Welcome to text strong TV. Thanks Alan.

Appreciate you having me. Yes, I'm Michael Mali. I'm the SVP of strategy for Seneca Global and what we are is we are an innovation software development company.

So for companies that are looking to Outsource their software development for hard to find skills. Like AI machine learning microservices, we do that kind of work and we particularly help small the midsize companies that startups. So it's a pleasure to be here.

Actually. Thank you for being here. Hey, Mike what we're at it maybe a little bit besides your present position at Seneca.

Give us a little bit of your background. Sure. My background is a 25 year technology guy.

I'm I I came up as a doubly engineer engineering undergrad engineering Masters and an MBA and moved into product management and leading product businesses. And so I've spent about 25 years in the space the last 10 or 15 insecurity specifically and it's really around helping companies innovate and bringing new products to Market and that's really what gets me excited. Actually Seneca Global for people want to get more information you gave us a little background, but what's the website?

com, that's our website and you can find information about our product development services and how we could potentially help you. Excellent. All right, let's let's jump into it.

So Mike, you know, we've heard rumors and threats and Promises of you know us passing some privacy and and data laws. I don't want to say a kid to gdpr because it's really not but you know our own kind of data privacy regulations legislation on a federal level, which really hasn't been done. No, right.

We would always threaten that as the big stick to make Private Industry kind of mind their own house and then individual states such as you know, California is probably the leading one passed their own but there's actually a a bipartisan. Supported Bill in Congress right now. Why don't you maybe clue our audience in a little bit?

Sure. So there as you said there's a bipartisan Bill in Congress. It passed through committee last session and they're hearings on it this session.

So there's a lot of speculation because it is bipartisan that it may actually pass this this Congressional session and it's the adppa. So it's the American data data privacy and protection act. And so what it's about is as you said, it's a national standard and it's really about moving forward beyond what was covered in gdpr and what was caller covered by the California Privacy Law to now move into the algorithms itself.

And talk about protection of the algorithms and the output of it and what it's really centered around is making sure that Ai and machine learning and the algorithms associated with that are being done safely with the minimal amount of data to not intrude on privacy and that also their non-discriminatory as well. Love it. Of course, the devil's in the details as always especially with You know laws and regulations.

when you say they're going to look at the algorithms Mike Michael give us Give us some examples of what you're referring to. Sure. So the way the way the bill is set up right now and again, you know, let's we probably shouldn't get too hung up on the exact wording to the bill because as we said right it's in committee right now right that some of those Provisions could change but the important thing is the way it's set up right now is you would have to provide an outside audit if you are company that uses AI which is, you know, nearly every new startup in in the world right now.

You would have to provide an audits to the FTC to say that your algorithm is safe that it's not discriminatory to show that it's effective. And so what it comes to is it really comes to around standardizing around best practices meaning if you take some of the previous, you know cases that that have spurned this and kind of given the industry a black eye you could look at things like the Facebook hearings or the Instagram cases where young girls were looking at, you know, No, extremely skinny models and the algorithms then who are trained to basically look at what you're interests are and serve you up ads and what you're interested in, you know had a behavior that obviously we don't want in society which is you know, serving up weight loss ads to vulnerable teens that are worried that they're not thin enough, right? And so that was an example.

You know I get it. That's a great example. Yeah.

Let me give you a fear of mine though sure that another example is Congress is going to be or the house is doing hearings on our big tech companies discriminating against right-wing. Post or points of view or what have you? I'm afraid a bill like this can be wielded as a as a weapon.

And and kind of enforced. You know, it's very easy to say sure. We don't want the poor anorexic girls to be getting weight loss.

It's another thing to say, I want my point of view to be unfettered. Even if it's wrong or even if it's not true or even whatever you if it's harmful. It doesn't make a difference.

And so that So I guess my problem with the this kind of Bill he lies. There's who gets to play God now we're gonna have third party auditors. Right extreme my experience with them is like with PCI Auditors, depending what PCI ordered or did your audit you had a very drastic?

It was a lot of discrepancy there left to the auditor. I you know, it it I worry about about that. Yeah, I think that's I mean definitely that would be a concern right if you look at the way it's done.

This really is gonna come down to using best practices. And as you said there will be you know, a significant variance in who you uses an outside audio depending on there knowledge, right, you know from my point of view. I think it's an opportunity for the industry to really standardize around best practices.

Right. There's nothing in the bill that talks about. You know that talks about any reduction in freedom of speech.

Right? So, you know specific to the concern that you have right? So there's nothing in there, you know specifically about that what it does talk about is that you have to verify that your algorithm is effective right that based on the data that you're you're feeding it that it's actually showing efficacy.

So think of it from a healthcare perspective where you're minding through a lot of data about People's Pharmacy records and the medicines they're taking and their doctor's visits and you using that for example their startups today that we're working with that are using that type of information for either medication management to help people, you know remind people on their mobile devices when you know when they need to take their meds, so they don't have to go into the doctor so much or also using it for early warning of, you know, long-term diseases like Timers or Parkinson's right. So great outcomes, right certainly things in society we want to promote and so I think it's really around using this as an opportunity to say. Hey, there are really great uses of AI and here's how we can do that.

If we choose best fit algorithms if we're careful about the training data that we use on these things to make sure they're getting the output that that we want. I mean that you know, the the example that I use all the time on training data is you know, as AI is only as good as the data you feed it, right if I give it pictures if I give an algorithm pictures of dogs and cats and the only kind of cat I show as a tabby cat. Well, then it's gonna you know, look at any other type of Pat and say well that's a dog because I got trained that the only kind of cats are tabby cats, right?

So those are the types of things where you know, their best practices we can use to make sure that the data we're feeding Healthcare algorithms are effective and giving us the information. One and we're not just feeding in noise right that we're not taking data about other prescription records or other things that we don't really need. Right?

And so and then the other piece of that then is biases right? We don't intend for the day to be biased. But again, sometimes we miss things, you know, the example there is an algorithm that was used by one of the Ivy League schools or my a few years back on admissions where they wanted to speed up the admissions process.

And so they said let's look at previous admissions data, right and look at what what were the key words in the resumes of the people that got admitted to the school and we'll use that as a screen mechanism to make it more efficient. Well what they found accidentally was that in doing that because the you know, if I look at a long enough sample over time, you know college admissions was more and more male dominated as I go further and further into the past. So on intentionally, right they found out that they were discriminating.

Against some women's resumes because the algorithm said hey, these guys are less likely based on past data to get into college. It wasn't that they were any less qualified. It's just there were fewer women in college back then right?

So there are things like that that I think is a society we can correct for and show that we can use AI for good right with some good auditing. It's just has to be, you know, intentionally done makes absolutely absolutely You're not going to get an argument for me about what are the benefits that AI can do for us all? No doubt about it.

Yeah. My concern is really the the concern is the implementation and the yeah. Yeah, I mean no, so I grew up in an area where the scariest way what was wrong already that scares her.

I'm here from the government and I'm from the government and I'm here to help right that kind of thing. And so when you tell me the government's gonna enforce this with third party orders and everyone has to file You know a certificate of you know, approval or whatever. I I just worry about the bureaucracy of it.

Yeah. Yeah, I mean obviously enforcement is gonna be a question for sure. Yeah, and there are lots of questions out there right now about who's gonna handle enforcement.

Is it going to be the FTC? Is it going to be the state's attorney general? Is it going to be the US attorney general?

Right. So there's a lot of things that have to be hashed out without question. I think the important thing though from a company to understand because for example, I'm already working and talking with startups that are looking at compliance solutions for these types of things.

I think the important thing to understand is in whatever form something like this is going to pass and that college you need to be prepared for it. Yeah, and it will give rise the same way the PCI gave rise to this entire. You know PCI testing infrastructure and audit infrastructure and compliance.

I think it's gonna there'll be a similar kind of more than a cottage industry. It'll be big business. Absolutely right doing this kind of thing.

And and so look, it'll be a federal. So it's Country-Wide and and quite frankly if they say look if you're going to do business in the US you got to do it. It really becomes a worldwide regulation much as gdpr has become correct.

You're not in the EU you're absolutely correct and they're even people there's even legislation in the EU looking at gdpr. And of course, what are they looking at as things to potentially adopt some version of this ADP PA so exactly as you said, I think we need to think about this as a next standard coming after gdpr. And as you said anything that gets enforced in the US by just just by the size of the US market is effectively going to become a global standard, right?

And so the question is as you said, right, we you know, as an industry, we certainly have concerns about it so we can either use it as an opportunity or a threat right? I think the best thing to do from an industry standpoint is to see it as an opportunity and use it as a way to Market ourselves show that using best practices show that good engineering and good AI is really helpful for society and ease some of these concerns. Yep, you know look it won't be without bumps.

Thank you guarantee you that but hopefully you know, even our dysfunctional Congress can amend this thing and you know, it will it will get better as well. Let's see it come out first and then we'll you know, we'll go from there but it's gonna be a living breathing kind of thing that we're all going to have to learn to deal with but you know what I can't tell you might I mean we're in text so for us this is Kind of bits and bites, but when you talk to people who are in Tech you say I see I think my Alexa is listening in on what I'm saying, or how does Facebook know I want to do that. So, you know, you've heard it from your family members.

I forget from mine as well and Friends. The over, you know, the the the following you from website to website with what you looked at before the retargets and stuff. It scares people it does feel very big brother and she does feel very kind of.

Oppressive, you know the that they're inside your head. And so we do need we do need everyone to feel like there's a fair Shake here. There's a everyone starts, you know, it's a flat table.

They're not kind of taking advantage of you. Yeah, for sure exactly as you said right I get that, you know, I get those questions from my mom right on. Hey, I went and looked at something, you know on the web and now ads for it are following me around right?

So yeah, what's the listening to me or my iPhone listening? I'm like no Mom that's just retargeting ads right maybe what those are but but to your point right for the people that aren't in the industry and our tech savvy right if we play this, right? This is an opportunity for the industry to show that AI is safe and effective and has really good outcomes because as you said don't understand how this works, but it's a great Innovation, right Ai and machine learning as a technology is tremendously helpful to the industry.

Right? So I think it's way for us to simplify it show that we can prove that these things are safe that they're effective that they get, you know from a healthcare perspective that they can save lives from a security perspective. They keep people's data safe, but at the same time Educate a little bit right about how these work are to say.

Hey, you don't need to understand the bits and bytes as you said, but you just need to rely on professionals like us right that understand the bits and bytes to show you that hey what we're doing here is sensible and reasonable. And we're reliable but we haven't even talked about the chat GPT thing, but we'll do that on another call because that's gonna you know, these people don't understand it's not sentient and you know, I've been getting crazy stuff with that. That's like I want to I want to thank you for coming on and sharing with us a little bit and telling us a little bit about Seneca Global as well come back and we can we can continue this conversation.

Absolutely. I'd love to do that. Thanks for having me today.

All right, Michael Valley here from Seneca Global on Tech strong TV. We're gonna take a break and we'll be right back with our next guest.