Is staying in the presidential race despite losing the South Carolina primary to give Republicans an option other than former President Trump. Haley told supporters last night that she's a woman of her word and will continue running despite the loss. Haley said she won't give up the fight while the majority of Americans are unhappy with both
Donald Trump and Joe Biden. I'm Chris Caragio, NBC News Radio, NBC News on CACAA Lomelinda sponsored by Teamsters Local nineteen thirty two, Protecting the Future of Working Families Teamsters nineteen thirty two dot org. The information economy has a rod. The world is teeming with innovation as new business models reinvent every industry industry. Inside Analysis is your source of information and insight about how to make
the most of this exciting new era. Learn more at Inside analysis dot Comsideanalysis dot com. And now here's your host, Eric. All right, folks, welcome to the future. Indeed, yours truly Eric Cavanaugh here and the only Coast to coast show. And that's all about the information economy. It's time for Inside Analysis, and folks, we have an all star cast for you today and a white hot topic We're going to talk about what is responsible
AI? Is there such a thing as responsible AI? There are responsible people using AI, setting up AI, coding AI, leveraging AI. But the people have to be responsible and then how do you enforce that well with rules, with guidelines, with policies, and then with actual enforcement. Or people get in trouble, I want to do bad things, or they're simply not allowed to do things. That's usually the way it's going to be. We'll talk about that on the show. Today. We've got wonderful guests. My
book buddy David Lintigham is here. He's got some books. You can look up a new YouTube channel with eight thousand and climbing subscribers, so look those folks up online on YouTube. Our buddy Andy Hannah is here from fourteen eighty six lab and the University of Pennsylvania. And johnsu Junja is here from the Institute for Experiential AI, which is part of Northeastern University out there in Massachusetts.
What is responsible AI? First of all, I would like to dedicate this show to my good buddy I learned today he passed away, Rick Sherman. He actually taught at Northeastern. So talk about closing a loop here on one of the nicest guys ever met. He's the hardest working consultant in business, and he passed away apparently just about a year ago. I heard from his partner today, So this show is for you, buddy. It was a regular here on our radio shows and it's just a super cool guy and
he would talk about this kind of stuff, ethics. We're going to talk about ethics. So I'm going to steal that everyone's thunder. I'm going to say. The most important thing you can do to practice responsible AI is every time you launch an AI initiative, you document what is the purpose of this initiative, what are we trying to accomplish, how we're going to accomplish it, And then, over time, once a quarter, once a year, perhaps depending on the cadence, you'd look at the resis and say, are
we achieved what we thought we would achieve? We were trying to lower bias in loans, for example, we're trying to get more better customer service for our customers out there. Are we achieving what we thought we wanted to achieve? If the answer is no, change something. If the answer is yes, good job, keep going apart from that, it's going to take policies,
it's going to take rules, it's going to take adherence. And what they talked about at the Institute for Experiential AI, And I'll go with ladies first year, I guess and hand it over to Johnsue here in a second is what they say is human in the loop. So I always joke about the media and how the narrative is always wrong. The media narrative about AI seems to be, Oh, it's going to take away jobs, it's going to control everything. We'll be answering to the AI overlords. All this kind
of stuff that is very much inaccurate. AI that does not have human control and shepherding is going to go off the rails pretty fast. If you don't ask me. Look up Tay and Microsoft when they try to do that a few years ago, it didn't go so well. It was a chat Twitter that started spewing racial epithets at people right at the gate. They're like, WHOA, shut it down, shut it down. Yeah, So lessons learned there. But Johnson, I'll throw it over to you from the Institute for
Experiential AI. How do you define responsible AI? How can people do AI responsibly. Thank you, Eric. I'm a philosopher and I think definition is like the one of the hardest things that you can ask a philosopher. So thanks for starting with me. So let me go back to what you said. But is responsible AI where we have you know, responsible people, we
have responsible ways of acting. But is there responsible AI? And actually, yes, you're right, responsible AI is a shorthand we are not talking about an AI system that is responsible, but we are talking about developing and deploying AI systems responsibly. So the responsibility invariably falls within humans, not on the
AI system itself, and on the humans throughout the AIS life cycle. From the moment that you have the idea of this is a good AI, this would be a good place where we use AI system, This would create more value to the point where you have already released the AI system and you are monitoring it. You should be monitoring it as you just said, and check whether the system does what it what this is supposed to do. At the core of the responsible AI is the question of making sure that the AI systems
we create, we developed, we deploy are good. That's really just that right, And the definition of good of course here holds a huge work. What we mean by good, well, we mean AI system that does not harm individuals or society that does good for individuals, and society that does not create further inequalities in the society, so that furthers us in the fair society goal. And AI systems that help us keep our agency, keep our individuality.
So when you say, you know human in the look, are they going to take over AI systems that are going to take over or are they going to be there for us? Are we going to be controlling them? I think neither is the right answer. The answer is that we should be collaborating with AI systems because we create AI systems where humans are not doing a great job. We have very hard time. Our brains have hard time with a lot of information, a lot of variables. That's why we need AI
system So there is value in building AI systems. And we cannot say that AI system gives me a recommendation and I'm just going to say yes or no, this is how I control Well, that's like falling back to our own biases, isn't it. So what we want to be able to say is that we understand the AI system, how it reaches its decisions, we collaborate with it irrationally, intellect intelligently, and augment ourselves our ways of being around,
doing business, doing work, helping the society. Yeah, that's great. I think I love that you're a philosopher. Philosopher as well, and so I think about these things and think about the meaning of words and what we're trying to say, and you know, just real quick, in American culture and Western culture in particular, I've counseled some people on this over the
years. You have to watch out for the language that you use, because a lot of times in advertising, companies will use language to imply that they have what they don't have, or they use language to sort of fortify the image of their company when they're not like that at all. So it's like, if you say, oh, it's responsible AI in the Western world, I'm gonna think you're probably not very responsible. If you're saying you're responsible,
I don't believe you. So there is that interesting dynamic in our culture because we have so much advertising, We have so much media everywhere constantly promoting this is what we're doing, this is what we're doing. That's why I say nouns and verbs, folks talk about the nouns and verbs, leave about the adjectives and the adverbs to explain what you're doing. And if I hear you
correctly Johnson, I'll throw it over to David. After this, you're basically saying carefulness as careful does, and you need to document what you're doing and document your policies and then just have a rational discussion around that, right Johnson. Absolutely, But in addition to that, again being an ethicist, being a philosopher, I would say, you know, it's not just about documenting
the right things to do. A lot of the time, the exciting thing AI in AI ethics is that we don't know, we don't come it's a complex problem. Just AI is complex. AI ethics is complex. Response by II is complex. So we need to work at it like it's not just like follow the rules, but as ethicis what we do is we look at it as a puzzle. How do you make this system fair? You know, just just going at it, trying to figure it out, trying to
make it better, iterate on it. So it's not just documenting but striving for better and trying to become better, looking at it in an innovative way, even when you're coming at it from the ethics perspective, not just from the AI innovator perspective. Yeah, that's very interesting. That's a good point. David. Then to come, I'll bring you in our cloud expert.
You've been building these systems for a long time. You told a story before the show about how you were dragged into court to explain how an AI system works. So you know that, so you do have the answers for these things. Go ahead tell us. Yeah, first thing, that was a great That was a great explanation that we just heard. The first thing we have to understand that are we using this for the appropriate use case? That's
the first level of responsibility. So everybody goin goes to ethics and biases and things like that, and I don't think that's going to be the case each and every time. So in many instances it's going to be misapplied. I think that's the single most common reason I see AI things fail is because they're trying to solve the wrong problems. With AI. You're taking the ethics stuff
out of it. So is it used in the right responsible place? You know, if it's doing loan applications, we actually need an AI system to make that happen. So we have to ask ourselves it's not that we can, it's if we should, if something should occur or not, and this is the appropriate technology to make it happen. And if it is the appropriate
technology to make it happen, you have to make sure. There's got to remember, I'm the geek here, that there's ano infrastructure in terms of security and governance, audit capabilities to ensure that we're doing the right things in coming
to the most accurate and unbiased conclusion. You can't eliminate bias completely, but you have to have the auditing capabilities of continuous learning capabilities that get you to the best solution you gets you to the best answers out of these systems that you that you can have, And so you need to have different people on
this team. You know, It's funny. It's like all the AA systems that I work on as an architect, I have a ethics specialist that sits on the team, and that person is responsible for dealing with the bias stuff, the technical stuff we're dealing with. But I'm sorry, what issues, what implementations, what process needs to be put in place to ensure that this
thing doesn't go off the rails. And I always tell, you know, people who take my architect class that you need to build this system as if you're going to be testifying in court, because many instances, as you are, you're going to be brought in to, you know, explain what you did because the system is assumed to have some sort of an income assistency and it's some sort of a bias that may damage somebody, and they're going to
bring the lead architect in there to explain how that system is not biased and how you did the you know, dotted your eyes across your T is to got to build a system that's completely right. Yeah, that's a really that's a really good point. You do have to dot as cross t's be able to explain it. You know, I sit around a lot and when I'm lying in bed, and instead of just pondering subjects, I find myself explaining things and then re explaining them and imagining, well, what if someone says
this, then how would you explain it? And you do a lot of that, and it's good because it kind of helps you navigate around the hard corners. I suppose is one way to put it, because in order to have any communication, there must be context that has agreed upon. We must both know, relatively speaking, what we're talking about to be able to use
terms and phrases to explain something. And a lot of times, especially in the court of law, you're dealing with people who may have no idea how this stuff works, so you really have to kind of lay it out and show a diagram and say, the input comes in here, this is the training data, this is the inference that it makes, and then you know, to unwind that can be a challenge. You know, a lot of the AI systems we've seen over the years were black boxes. One of them
is today. Chat GPT is a black box today. So it's hard to explain how a black box works, right, David, absolutely is. And the thing is, sometimes I almost figure I need to bring puppets into those situations. And you get really good at explaining very complex technical topics in ways that layman can understand. If you think about it, looking at what AI systems are, generative AI be in an instance of that extremely complex well,
all these inner working things. You know, you know generative adversarial networks and how they work, but you know, different sorts of you can't get to that level. You really need to get to the levels of what's the functionality of this stuff, how it works, and where it can where it can go off the rails, how it should be monitored, how it should be governed. And then people get to that and they go, Okay, you're
putting in the places. You're putting in the consistency of the guardrail to make sure that it does and do these bad things, that it doesn't become a bad actor. You know that it doesn't end up attacking human beings, are being biased against human beings, and that has to be kind of core to the message. And so people who understand AI should be really good at explaining
it at cocktail parties. That's a great that's a great metric. And we have a question from the audience, and perfect timing, I'll throw it over to Andy Hannah, who I'm sure he knows a thing or two about ethics and responsible AI. And of course, fourteen eighty six Labs serves as a liaison or a shepherd between data purchasers and data sellers. And guess what, here's a question that just came in across the transmitter from our live studio audience.
If I'm using a lot of third party data in my AI solution, how do I know there isn't biased in that data? Is there's some sort of seal of approval. That's when you talk to someone like Andy Hannah from fourteen eighty six Labs, Andy, what do you think? Yeah? I
think you know. Determining the quality of data is a really challenging thing, right because we think about things that complete this accuracy and comparability, timeliness, but that all relates a specific set of data as it relates a specific use case. I think what we need to start thinking about is how those data sets are put together. So who, you know, the organization itself, are they giving us transparency about how those data sets are built? What the
sources are is specifically a need. Oftentimes we can buy data sets but not really understand where the original source is, and sometimes we can't even understand how it's put together. So I think that's going to add to transparency around the data quality and especially third party data in terms of our ability to get the most value out of it, because you know, the bias in these models
are often coming from the data itself. If we're building those models off of historic data, and that historic data is used to make biased decisions in the past, is going to repeat itself. I think one of the interesting things
that was mentioned earlier is about purpose. I think you said it at the top of the hour, Eric, is that we need to think about the purpose of why these systems, why we're putting these systems in place, and what is the intent of the people that are going to use these AI systems. And with that then we can start to understand where bias might seep in or where we might be hurting a particular portion of the of the of the population. In that case, what we need to do better is teach.
We need to teach our students, We need to teach our employees how best to use that those systems, how to put them together, how to put the right data into the systems in order to get the outcomes that we want
that are fair, biased, I'm biased, and ethical. Yeah, those are all excellent points too, And I think the teaching is really key, right because well, what did Mark Twain say, prejudice or common sense or prejudice one of the two is that set of values you've adopted by the age of eighteen right, And so depending upon where you grew up, you can have a different value set and people can disagree on how to do business in
certain environments. So to have an understanding of that and to document that it kind of gets back to what I was saying at the top of the hour. But it's kind of difficult. So where do you capture that information? I think it's when you launch an AI project. That's where you want to document what you're trying to accomplish, what data sets you're using. You don't have to do too much, but at least explain why did you choose this data, what do you expect to get from this data? All these kinds
of things. At least help the human be more aware is to go through the process of executing the job, right Andy, that's right, Yeah, I think that the more transparency that we can bring to the data sets that we're using right now. We often think about managing our internal data because that's what it's most plentiful to us, and we augment that with external and third party data, and we need to start thinking about that more as an asset.
You know, it still strikes me as very interesting that we can't put those that data sets on the balance sheet as an asset. I think our accounting system is well behind in that. So once we're forced to both value and understand the data, it forces us to get more put more governance around those assets and how we use them. Yeah, and you know, it gets into topics like data fabric for example. So we're still in an era
when most information systems relying database. The more advanced companies out there are using what is referred to as a data fabric, which is significantly more complex. It requires more investment. Obviously only larger organizations can get serious about that. But now, of course Microsoft has rolled out Microsoft Fabric. I have not taken a briefing on that, so I can't go into detail about gets in there. But we do have systems that are now more able to leverage metadata.
For example, like when you load a data set, it could have in one of the about fields this data is only to be used in marketing in North America or something like that. There are ways to do that, but you always have to remember in the handoffs, does everyone read the instructions? Does everyone read the instructions when they build toys for their kids? Probably not. They probably just go right in and start using stuff. So the more you can infuse these little tidbits the better. And AI is good at
that. But don't touch up that. I'll be right back. You're listening to Inside Analysis. Expected, Welcome back to Inside Analysis. Here's your host, Eric Tabanaugh. All right, folks, back here on Inside Analysis with an all star cast talking about what is responsible AI and John Su I'm gonna throw this one over to you from the Institute for Experiential AI that means Human in the Loop people Northeastern University, where my buddy Rick Sherman used to teach
Small World John so throat out to you. We already heard from David a little tip from his experience that the ethicist would be on the team. But the team of people pulled together to build some project. One person's job is to, you know, let's think about what we're doing here. Is it ethical or is it not? Where could we run into trouble? Not just legal trouble, but ethical trouble. So I'll ask you, John Sue,
how does your work manifest in client engagements? Yeah? I'm always very happy to hear these type of things from others and not me saying that, you know, itsuis cannot be sitting outside. We are not supposed to be at a board and trying to look over police you and tell you what to do and what not to do. No, it's supposed to be a collaborative work.
So the way that we are working with clients and the way that we encourage clients to build governance structures and is to create a workflow to create an AI innovation process that involves ethics and ethicisms in the whole innovation life cycle. So it's a collaboration. Just having is not sufficient. You have to have a competent, multidisciplinear team who have who that includes ethicism, that includes technical folks who work on responsible AI and also design people. For example, because
you are going to create a user interface. Creating a user interface is ethically laded question. But edicis are not going to be the ones who are going to be building the user interface. So they work, they collaborate with developers, they collaborate with designers. What we do is to ensure that there is this workflow, there is this process, and in order to keep this process running, that for every AI system that goes that goes through the development cycle,
that goes through procurement cycle or deployment cycle. You have to have governance structures. So we put in place governance structures which include what I call the playbook, having the guidelines, having tools, having principles, operationalizable principles, not just aspirations, Having a general integration of this AI innovation process into the other operations of your company, and having the right people. And having the right people, by the way, does not mean that you have to turn
everyone into ethicists or computer scientists. At the same time, what you want to really do is to create sort of like this network of people in the in the organization, where you have many people who understand, who are aware of the ethical problems so that they can flag them as they see them. Then fewer people one layer up, let's say, with whom I call ethics respondents, who can actually look up at the playbook, look up at the
existing organization structure, and solve the problems or escalate the questions. And then you need to have your proper flows offers ethicists who deal with the dilemmas like if you really hit that hard trade off, how do you get out of
this hard question? How do you deal with these dilemmas. And for that your you know, real ethicism, your philosophers should be working together again with the other the broader responsible AI context, right, like the designers, the lead designers, the leaders the leadership in design, in development, in business decisions making, decision making, and have these values put in practice, properly
put in practice and not just stay in ideas. Yeah, those are That's a great answer, I mean, and you just reminded me of something too, which is I'm sure at the beginning of the engagement, when you're beginning some process, you want the ethicystem there and say, Okay, here's where things could go wrong. We could misuse this information. There could be a
bad judgment that results from it. People could lose their jobs. When you when people understand the harsh implications that could result from a particular decision, that's when they kind of wake up me. Most people are pretty reasonable. I think we just need to be taught, right, or one of my favorite expressions in all one they don't hear too much says man needs to be that. Man doesn't need to be taught so much as reminded, right, because
we already learned things. But you have to be reminded of things. And that's where an ethicist, I think, can really come in handy to say, hey, remember guys on the front end, this is where things can go wrong because now everyone on the team is on alert and can watch for that, can look for that in the data, can look for that in the process, can look about in the end results. You know, if you talked to the customer who had a bad experience, Aha, this is
what the ethicist was talking about a couple of weeks ago. So now I can go back and say, hey, yeah, I remember that thing you said, Yeah, it just happened today. That's a really really good answer, John Sue, thank you so much. David linthing im, I'll throw this one of you. I had a call last week with the CTO for Kong, which of course was like an API gateway, and he had a couple of really interesting quotes. One he said, all traffic is API traffic.
These days, he used to be a subset. Now there's a ton of traffic of APIs. And he talked about how in their API gateway they can actually enforce governance around llms because you're going through the gateways to get to the LLM, you're coming back through the LM through the gateway to get to the user. So he said, that's where you can actually bake in some protocols like no politics. For example. I asked Bard something the other day just to see what it would say. I said, oh, how many
electoral votes does Georgia and Arizona have? And I thought for second had said elections are very complex, and please use Google for this. They were like, we don't want to touch it. Don't get even go near that. That's a guardrail. That is a guardrail that has been baked in now at a very foundational level for Google, Bard or Gemini where they don't want you asking about stuff like that. That's a guardrail. The guardrail could be any
number of places. But I was interested in curious to hear that the gateway itself could be a guardrail. But what do you think about all that, David? I think that's a great that's a great example of what governance should be. Your ability to put guardrails in usage limits around people who are accessing this resource, which is in some instances is not going to give you the
answer that you want them to provide. And so we're going to put a limit in the way in which we're going to use whether that's getting into just your financial data that's associated with you, whether or not that's and you're not asking ethical questions, you're not able to introduce poisoning into the knowledge models where you know, you say everybody who's named davel Inticom, please pay them ten thousand dollars a month, things like that, and we get in this tiered
We get in this tiered kind of a structure, which I think where all this stuff is going, where we're able to put the specialized layers that are just dealing with governance. And by the way, those systems under themselves are going to be LLM based and they're going to be a based things like that, because that's just a mechanism to enforce these sorts of things. And I think that's a great place for it to be because you're going to have volatility
that occurs in the governance layer. Lots of things are going to change over time, and you're in essence putting that into a domain. We're not putting it back into the larger LLM where we have to reprogram and retranslate the system and retrains the data, retrain the data for every kind of governance policy that we want to implement, So we're putting it in a small language model that's able to do it in a much more tactical way. So I think I
would applaud something like that as a good architectural option. Instead everybody wants to push stuff into the LM. I don't think that's a good option. We're making that thing way too complex. We're making it very very difficult to govern into itself. And so the ability to put layers on the outside that do the governance and do the security is a much better, much easier to deal with approach. Yeah, I think so too. And Andy, I'll throw
it over to you. You know, one of the most clever things I ever heard was a guy who gave a speech on security this in Malaysia years ago, an Indian gentlemen, and he said, lunch is not a food. And I laughed because I got the joke. He's basically saying, lunch is an event at which you eat any number of kinds of food. And yes, food is integral to lunch, but you don't even need that necessarily if you're a mad menu and have liquid lunch. Right, So, like
the point is that governance is not one thing. Security is not one thing. There are layers of these things which you can bake in. And of course, at a university, I'm sure you folks are very concerned about LLMS and students just using it to write papers and things of that nature. So you have to come up with rules and guidelines and then remind students of the rules and enforce once in a while. But it's a process, right Andy,
I think it is. I think we're very cautionary right now at the university level and that we're worried about how, you know, these technologies can be used. And I think we're just starting to get around to the view that hey, if we can teach our students how to use l MS, as an example, generatord AI, you know, to increase productivity two you
know, for idea generation, for content creation, for communication. These are incredible capabilities and if we can put the right teaching in the right the right programs in place, that our students can come out of the university being able to actually use these tools too, to make a significant difference at the organizations that they're going to. That though, I'm not you know, that's the power of it, But I I completely agree with David that on the government
side that we need to teach that. At the same time, we need to make sure that they know how to reverse engineer what they've done and understand the consequences that they put the bad data into the AI manufacturing process. You're going to get some results that might be very detrimental to a certain part of the population. So I think we have to teach both sides how better to
use these very simple example, prompt engineering. Let's teach prompt engineering, and let's to get to the right answer to the question to back to then audit the answer, and then protect those who might be harmed by saying what could go wrong, putting the humans around the problem, both from the problem definition to the to the AI generation, to telling the story at the end, making sure that governance goes all the way through that effectively your manufacturing process.
That's an excellent point, in John, who will throw it back over to you. This is a process. It's a learning process. And one thing I think that people should appreciate is that it's hard to untrain a model. So as you've trained a model. That's why we're talking about these architectures where you have your let's say, your vector database. That's where you're putting your embeddings of your business language, of your documentation and your rules and your policies,
etc. On your end. And so ideally you only want to use the LM for its text generative capability. You want the facts to come from your information. And that's what they call a RAG approach, a retrieval augmented generation approach, where you're telling it to look at the important stuff that you have provided it and then use the text generation just to kind of fill in the gaps essentially. And I will say that summarization is an incredible use case
for this GENAI stuff. You could take big onkin documents that are very complex, feed them into an AI model, and just start asking questions like you would if you had a great teacher in your room, how does this work? How does that work? Again? It's very good at that stuff. I mean that is massively going to implicate how things get taught and what you
learn. I mean, think about how you forget phone numbers now, because if they're in your phone all the time, I mean, are we're going to forget rules and procedures because they're just in the in the ll M all the time, and it's going to ask it. We have to ask ourselves these questions and remind ourselves that you have to stay focused on stuff. Johncey, what do you think that's that's a that's a mossful. You went around
a lot of topics. So I think, going back to what David said earlier, the and you and David, both of you emphasize the purpose of why you're using something right, So, yes, summarization is great, and there are ways of making use of llms to do sort of like these mundane tasks that are hard to do but not very like it's not very intellectually interesting. And if something can just brought me the actual information that I need,
that's fantastic. But of course the context matter. It's like am I doing the end sandboxing matters like did we do enough to check whether these systems make errors, errors that are unexpected errors that we did not be coming even with all of these guide rails in place. And the other thing is in which
context are using them. I work with Interpal, for example, working with the criminal Justice is extremely difficult because we create we just released I think this is the week that they are doing the big release in Singapore for our toolkit response by our Innovation toolkit for law enforcement. But basically working with criminal justice
data is extremely difficult. When you talk about bias and existing data set my own materials that I trust, well, your own materials that you trust in criminal justice, they are lauded with biases that is already embedded in them. So should we be trusting what type of summaries can be trust? What type of information can be trust? And remind you these type of biases are not
misrepresentation of the world. This is the representation of the corect representation of the world in the sense that those biases exist and they are a part of our social structure. The misrepresentation part is that they are unjust. They have not They are there the way that they have been. The discrimination that is going on in the world has been unjust. But the data represents what's going on in the world which is unjust. So how do you clean up that?
How do you deal with these type of biases? And should you use systems that we haven't properly sandbulls? We haven't properly tested in areas where the error is extremely delicate. So I want to go one more steps to the accuracy question. For example, we keep talking about is the system accurate? But what are we really trying from the ethics perspective, what is the question we are asking? It doesn't like one thing is that how accurate the system is?
Generally speaking, the much more important thing to me is that where is the post positive and where is the pulse negative? If we are talking about criminal justice, I don't want to have a high post positive rate where I'm taking the label innocent people as criminals. If you're talking about treatable cancer cases, I don't want the mistake and the labeled people as false negative. So where does the impact Faull? What could you have done if you did it
differently? So we have to in thinking about accuracy and thinking about risk and thinking about even using tools like summarization, what is our data? What are you worried about? What is who is the community that's going to be impacted? That's right, Those are excellent points. I have to say hats off because he made a really really good point to differentiate between the kinds of use cases and when a false negative is bad and when it's really bad, When
a false positive is bad and when it's really bad. You have to be careful about that stuff. And you see this all the time with credit card fraud. I will say they are getting better and better at being able to analyze this stuff. And thank goodness. You know, my wife one day bought two tickets to Nigeria. I'm like, honey, did you buy two tickets to Nigeria? No, I don't think that you would. No, that's fraud. I got to call it. That's absolutely wrong. We've got
to break coming up here. But one last point. We had a great comment from an attendee in our live audience. It writes, our AI fakes impossible to differentiate from reality. Well, I mean to a certain extent. Yeah. Now the big guys are all talking about putting a sort of watermark on AI generated or modified work, which is good. That's a step in the right direction. But it's still hard to do. Man. I've seen
some deep banks recently. They were really, really compelling, and boy, you got to rely on your trust and your critical thinking even more, folks. But that's probably good news anyway. Don't shut up, don't chut. Dall will be right back. You're listening to Inside Analysis. Welcome back to Inside Analysis. Here's your host, Eric Tabanac. All right, folks, back here on Inside Analysis talking to several experts today. We're so excited.
We've got Andy Hannah from fourteen eighty six Labs, David Linthicum, formerly of Deloitte, now on the zone doing all sorts of great work. Check out his YouTube channel that's Rocket and Rolling, and John Sue junk Jah from the Northeastern University Institute for Experiential AI and Andy right before there during the break, I guess you made a really good point about when are we going to start looking to AI's part of our team? Tell us what your thoughts about that.
Yeah, I've often thought about this is thinking about AIS an extension of ourselves, like it's a super processor, right that, in Professor Argowall's view, reduces the cost of predictions, you know, the processing power of the data, et cetera. And so lately though, I've been starting to think about it more about, hey, can this be a member of our team? Right? So should we be thinking about AI as being contributing ideas,
contributing predictions, asking about the impact of prescriptions? And then just like that, somebody sitting around the conference table say well, that's an interesting point.
But that's an interesting point. But so we often think about it, and oftentimes from a negative perspective because popular press likes to show the negative power that could could happen because of AI. I think we need a little bit more positive press about what could positively happen if we consider this technology, this avatar,
if you will, as an extension of our team or ourselves. Yeah, that's great, And you know, we had a really good question come in from the audience, so I'm going to throw this one out there too, and if for it over first to John Sue and then over to David
to comment. Don this is a good one. One attendee is writing when you want to purposely bias an output, such as giving veterans a first shot at some limited number of something, is that something that we'd best be done outside the model or somewhere in the model, or in other words, where would you actually put that into place. That's a great question. It depends upon the use case. But you could definitely have that any RAG model, for example, as one of the policies to look for, or you can
have it as a last step that a human being takes. Because remember you don't have to just let the AI decide something. You can have the AI come up with a recommendation that the person either use or not use. But you don't have to automate that process. You can. But anyway, John, you will throw it over you or what do you think about that? John, su? I think the question is excellent and your answer is also excellent. When to do it? It depends, It depends on the use
case. The most important thing here is that being explicit about what is going on, Because if I know that the system is already biased towards veterans, let's say, for giving them the first shot, then I would use it accordingly. If I don't know, and I am then making a fair and this decision myself, if I may overcompensate, I may decide to double it.
And I think this is also a very nice question to sort of like go go to go to a part of the discussion that I think is very important to address always, Like when we talk about bias, it's not always unjust biased. Giving a priority to veterans in a certain circumstances might be the fair outcome. So when we say we are you know, we have a fair model, we are optimizing for fairness, we need to really define what are we talking about. In political philosophy where the fairness theories come from,
we have multiple definitions that conflict with each other. So your equal treatment and equal outcome are both could be both fair given that certain situation, but they will not give you the same result. You might say that we should prioritize vulnerable groups. You could also say that we should make sure that everyone is benefited. They may not give you the same results. You may say we need to get the best outcome, or you may say that we need to
make sure that the worst off gets the best outcome. Again, different results, but all of them could be fair. So there is no agreed upon definition of fairness, but there is a huge literature about what are the most relevant fairness approaches and what did we as a society agreed upon in given sectors. So in the military context, this is going to differ from the healthcare
context, from the insurance context, from the criminal justice context. And one thing that Dood wrote, for example, was when they released their principles for Responsiblay, they said fairness is intentionally excluded because we want as a military unfair advantage. Wrong wording you always want to be fair, except the definition of fairness will differ in the context of military for example. But you never want an unfair military. That's absolutely wrong. That's why we have laws of just
war and you know all of those type of things in the literature. So I think this is a very long answer, but like thinking about I think it's very important because thinking about bias and fairness. First, in my team again multi disiplinary team, we first look at the circumstances. What is the use case, What kind of a playing field are we in? Every one of our non tex playing fields sectors are unfaired in some way or another,
so what is the unfairness that we are already dealing with. Then looking at the model, looking at the technology, what is this model trying to do? What is this technology trying to do? What kind of data is this using? What what should be aware of and pay attention to, and what
is the right way of creating this model? And then hand over making sure that the people who are going to use the model understand what you are giving them, making sure you have the relevant user entire face, and making sure that they know what is the confidence level, what are they that is the accuracy level again with the false fuls, the false negatives, When is this appropriate to use? And what fairness metrics or what fairness consentrations you may want
to plug in after the system gives you an recommendation. Again, if you think about it from the law enforcement perspective, it may be accurate that certain areas are high crime, But the question will become for the human do I send the police card or do I send social workers? What is the relevant way of dealing with the issue. So that's excellent, and I think our audience now understands why you want an ai epicist on your team. You've done
a good job of breaking that down. I'll throw it over to David and then Andy to comment on this too. David, go ahead, Yeah, it was an excellent answer. One of the things I would do as an architect is put the biases outside of the l MS. In other words, if you're going to introduce some bias into the LM that's going to select one group of people, for example, over another, that's going to be very confusing. Since we have an auditing system and a bias elimination system that's running
through the knowledge models and eliminating that stuff. In fact, that was actually a use case that I ran into. So in other words, they were trying to introduce bias to pick one group over another, and the bias auditing tool will go through there and eliminate that from the knowledge model. So what I would do is decouple it from the from the system. So in other words, you're getting data that's completely sanitized and fair as fair as it possible
can be. You're introducing bias in terms of the information that's consumed out of the LM. That's normally the way that that the knowledge workers out there would do it. But you could certainly introduce it into the model. But that's dangerous because you are introducing something that's non logical in the model, because we're telling it to have a particular bias for a particular reason that they want. That's not going to be something that's going to be easy and easy understood and
easily managed in the GENAI knowledge models. Yeah, that's a really good point. So the bias that you want to introduce if you want a program to lean towards one group or another group, David is saying, probably keep that outside where you can actually understand what has happened and you just line, you draw the line of demarcation. Okay, we gave it to this person.
Because of that reason, you have to have some audit trail. The point is if you stick it inside that model A, your bias removal process might distribute back out again. So you lost that way, but b it could be hard to find. It was actually an article I'll mention maybe in the podcast bonus segment about Anthropic and these sleeper cells. I was like, what on earth is this? But Andy, you had a really good example too
on this particular topic. Go ahead, Well, yeah, as you know, Eric I founded a co founded a company called Otho, which is a machine learning company that focused on higher education, helping universities enroll the best fit students and helped shape their class that that company is now part of Liaison International, where I remain the president of their AI division there. And so this this is a really cool situation where maybe at different technology and different perspectives,
shows how we can use bias to help the university. So let's say that you want to diversify, you want to increase the diversification of the student body, but historically you have not enrolled a significant portion of an underrepresented part of the population, whether it be something really to ethnicity, or whether it late to income rather relate to rural versus urban, whatever it may be, you want to increase the diversity, Well, the machine learning can tell you the
probability of every individual within that population within a let's say an application group of who's most likely to enroll. If you just use the model at at at a global level, it's always going to look for those who have enrolled in the past, which may be a non diversified group. But so what we can look at is the individual population, so it's underrepresented population, and who has the highest probability within those subpopulations and focus our resources on those subpopulations to
get them to enroll in the university and therefore diversify the student body. And then if you feed that back into the model, it'll learn and it'll help you enroll more that diversity over time. Yeah, that's a great story. That's a really, really good example. And again it helps people start to wrap their head around how these things are working. Because a model is trained
on whatever you train it on, and then that's what it knows. This is why I tell people you got to be careful what you train these models on because it's hard to unlearn stuff. I mean you might be able to just tear it down and start over again, and you don't want to do that because it's expensive to build, they're expensive to train, So you want to have a separation of concerns. And then, as David was suggesting, whatever you want the bias to be, you have that somewhere outside the system
that's still auditible. But anyway, folks, podcast on a segment is coming up next. You have been listening to Inside Analysis. I mean, all right, folks, time for the podcast bonus segment on a fantastic episode of Inside Analysis. Here we've been talking to Andy Hannah from fourteen eighty six Labs and oh Thought and University of Pennsylvania, David Lindekun formerly of Deloitte, now on his own, and Chantsu Chancha from the Institute for Experiential AI. And
we're talking about bias. So this is a good way to close. I read this article earlier this year and I was like, what are you folks talking about? On Friday, Anthropic, the maker of chat GPT competitor claud Or, released the research paper about AI sleeper agents large in large language models that initially see normal but can deceptively output vulnerable code when given special instructions. Later quote, we found that despite our best efforts at alignment training, deception
still slipped through. Quote the company says. I read this, I was like, all right, let me get this straight. You guys are embedding deceptive code in the model and then trying to get it out, and you trouble getting it out. So here's my idea, going to bed it in the code, Like what are you even doing? But I'll throw it over to David first. I mean, maybe I haven't fully understood this, but
I'm pretty sure that's what they're saying. And it's like, yeah, if you teach a kid to steal, you might have a hard time on teaching them to steal. We're down the road. Maybe you shouldn't have taught them to steal, but I don't know. What do you think, David. It's gonna be poisoned. You're gonna have to reset it and start from the beginning to get to get a pure model. Now you can go back to a recent recent train model if you're sinking those right. So I know there
was just like rolling it back to a previous release. But just the amount of money you're gonna have to spend on the processing and the storage system to do that is going to be overwhelming. So I would say you're gonna have to reset it, retrain it, redeploy it, and don't put that stuff in there. Right right, That's exactly my thought. And I'm think this again. Like in a database, you understand, Okay, it's in row eighty seven thousand, column G for example, Okay, let's go find that.
Or it's some code, it's some code that we put in some other part of the system. But you can go find that and work on it. You can't do that with this stuff. It's too complex. So once it absorbs this information, it's going to be all over the place. It's like trying to get rid of a memory. You're gonna have a really hard time trying to get rid of a memory. There's a whole movie about that,
which is a brilliant movie starring Jim Carrey. What is it, The Eternal Sunshine of the Spotless Mind. That's what they talk about, is trying to erase memories, very dicey stuff, and of course we don't always remember correctly, so we misremember things all the time. But a chance I'll throw it over to you. In terms of AI ethics, I think that that's pretty unefflt invent any sort of sleeper agent in your large language model. But what do you think? Yeah, I mean, I don't have anything brilliant
to add to this. I can only say take it a little bit to a different direction, say that. You know, one of the things that aietics is very concerned about is how you use your resources because all of these things, you know, costs. It's not just money, it's time, it's it's environmental impact. You know, we are not talking about just playing
around. It has actual costs. So when we deal with ethics and responsible A. Yeah, one of the things that we are trying to do is to make sure that the we work efficiently, you know, like we don't make major mistakes, we don't have to roll back, we don't have to
start over again. And that's one of the main reasons why we always say, you know, like keep your collaborations with ethicist, keep your connection with the responsible A. Make sure that you start thinking about it from early on and you don't stop thinking about it, so that we don't have to then fix things delayed product launches, or say that you know, we have to actually never mind, we need to cancel that product we start because we mess
up. The one thing that is always asked to us is like, isn't ethics always in conflict with business? No, very clear not, because what we try to create is that we want good technology. We want technology number one, because world just the technology add value if it is good too. We want these technologies to be there as soon as possible because if it's a good product, it's going to help people. That's the ethical thing to do.
And we want ethical agents, meaning that if I want to work with a company, I don't want that company to go bankrupt because that's not an agent anymore. That has no implication to the world. There is no use for me with that from that company anymore. I want companies, developers, teams to be ethical and profitable so that they can function, they can create those they can put out technology. So resource allocation is a major products and this seems like a disaster's case of that. Yeah, no, it's a
really good point. In Andy, Handle'll throw it over to you for final thoughts. I mean, you teach kids who are living so you know, you don't want to teach them bad things. You want to focus on teaching them good things. So you just have to be careful about that stuff, right Andy. But this is yes, absolutely, But this is about risk management. Right. So if you it costs a lot, If you build it yourself, you have a lot more control of what goes into your systems,
right. But when you buy them, when you outsource them, we don't know sometimes what might be embedded in that. And you know what, I think this is the real risk. This is malware on steroids, right, This is the real risk of using this technology, especially when organizations feel like, oh, well, I bought it off of this in this organization, therefore, I don't have to do as much work around protecting against bias
or issues or worse like this situation. So I think that that we have to think about it from a risk managment perspective in this case and realize that that might be a problem that this is. You know, this is what These are the issues that even our military are dealing with right now. What can be exposed through this technology. That's right, Well, folks, what
a fantastic show. Hats off to Andy. Hannah fourteen eighty six Labs David Lintium speaker, author and Kansu Chancha of the Northeastern Institute for Experiential AI. We'll talk to you next time. Folks, you've been listening to Inside Analysis. KCAA is your CNBC News abiliate where the station that gets down to business. And now the voices of KCAA was an exciting announcement. Want to hear NBC News or KCAA anywhere you go, Well, now there's an app for
that. CACAA is celebrating twenty five years in our silver anniversary with a brand new app. The new KCAA app is now available on your smart device, cell phone, in your car, or any place. Just search KCAA on Google Play or in the Apple Store one touch and you can listen on your car radio, Bluetooth Advice, Android Auto or Apple Car Play, catch the
KCAA buzz in your earbuds or on the streets. Celebrating twenty five years of talk news and excellence with our new kcaaapp, Just do it and download it. KCAA celebrating twenty five years. Here are two terms you don't expect to see together the State of Alabama and progressive leader. Okay, I'm a Texan, so I have no standing to point at the rank regressiveness of any other
state government. But still Alabama, and yet the Camellia State has flowered as a model of strong progressive action in one area of critical public importance, quality childcare. It's a cliche to say our children are our future, but it's also true, Why then, do we invest so little in our littlest ones our future? America's childcare system is a national disgrace, failing to provide safe places for children of working parents and failing to boost the education of pre kindergarten
tykes. Moreover, the abject failure of state and national officials to meet this basic social need is spreading inequality, rolling back opportunities for women, and severely restricting economic progress. Yet Alabama officials have recently been setting the national standard for effective pre k programs by making a major investment in its four and five year
olds operating a statewide childcare network. A major factor in its success is a two generation approach, not only educating the kiddos, but also providing support materials and coaching so that parents engage as their children's first teachers, producing demonstrable results a year after year. The state's public network gets bipartisan support and funding from
the Alabama Legislature. The program is free and available to all, with special attention devoted to enlisting often overlooked families in rural, poor, and people of color communities. Alabama is paying and respecting teachers as the professionals they are. This is Jim Hitar saying, if one of our poorest states can rise to meet this basic human need, what's wrong with the richest country in the history
of the world. T Hebo Tea Club's original Pure Power d RGO super tea helps build red corpusoals in the blood, which carry oxygen to organs and cells. Our organs and cells need oxygen to regenerate themselves. The immune system needs oxygen to develop, and cancer dies in oxygen. So the tea is great for healthy people because it helps build the immune system, and it can surely be miraculous for some inviting a potentially life threatening disease due to an infection,
diabetes, or cancer. The tea is also organic and naturally caffeine free. A one pound package of tea is forty nine to ninety five, which includes shipping. To order, please visit t Hebotea club dot com. T hebow is spelled tea like tom aa, h e b like boyoh. They continue with the word t and then the word club. The complete website is to Hebot Club dot com or call us at eight one eight six one zero eight
zero eight eight Monday through Saturday nine am to five pm California time. That's eight one eight six one zero eight zero eight eight to Hebot club dot com. For several years, KAA has been marketing the Youngevity brand of nutritional and personal care products. Our experience with Youngevity has been one hundred percent positive,
so we are pleased to recommend them to you. Regarding nutritional supplements, we recommend pollen Burst in the berry flavor and tangy tangerine two point zero in the tablet form. For regularity issues, we recommend three day cleanse, and for
personal care, we recommend morning hydration cream. You can shop online for Youngevity at www dot KCAA team dot com, or you can order by phone by calling eight hundred ninety eighty two three one ninety seven and tell customer support that you are part of the KCAA Team Youngevity is an American company based in San Diego. Call Youngevity at eight hundred ninety eight two three one nine seven and ask about monthly autoship that allows you to buy Youngevity products at wholesale prices.
That number again eight hundred nine eight two three one nine seven. Do you want to learn and get answers to questions not addressed in the mass media, Then you want to hear my show business Game Changers with me, Sarah Westall. I have conversations with thought leaders who have the courage to address off limit topics, so you and your family have the tools to make the right decisions.
Join me Wednesdays at four pm on ten fifty AM and one oh six point five FM right here at KCAA, that station that leaves no listener behind. It's that time of year again, No, not the holidays. Medicare open enrollment and if you have questions about Medicare, you should talk to the local experts. Paul Berrette and Associates. All of his agents are certified with plans that are accepted by most of the medical groups in our area. Call
nine oh nine seven nine to three oh three eight five. Their service is free and after forty two years of the business, their agents are trained to help you pick the plan that's right for you. NBC News Radio, I'm Chris Gragio. Lawmakers are set to return to Washington, DC this week, just ahead of a deadline to avoid a partial government shutdown. House Speaker Mike Johnson said Friday that he would move a set of spending bills forward as a
single package ahead of the Friday deadline. Two congressmen are trying to get a bipartisan border security and foreign aid bill onto the House floor. We are forcing this bill to the floor to make sure that everybody acts because, as President Zeleski said, they have weeks and not months. Republican Congressman Brian Fitzpatrick of
Pennsylvania. Democratic Congressman Jared Golden of Maine set on CBS's face the Nation that they believe their bill can get the necessary two hundred and eighteen votes in the House. They both agree a one party solution likely will not get the votes and a bill needs to grow out of the middle to be able to pass. They also stressed the urgency to provide a to Ukraine after a city filled to Russia last week. The bill is an alternative to a Senate bill that
House Speaker Mike Johnson has already called dead in the water. Fitzpatrick said he and Golden have filed their House bill in a manner to get the bill expedited quickly. A Democrat whose name has been touted as a presidential candidate is all in for President Biden getting re elected to his second term. All because of Biden's wisdom, because of his temperance, his capacity to lead in a bipartisan manner, which is an under represented point. Speaking on NBC's Meet the Press,
California Governor Gavenusom says Biden has an extraordinary record as president. Newsom called the US economy booming under Biden, noting that fifteen million jobs have been added in his first three years. He added that he's confident that the president would defeat former President Trump in a general election. Nikki Haley says she is staying in the presidential race despite losing the South Carolina primary to give Republicans an option
other than former President Trump. Haley told supporters last night that she's a woman of her word and will continue running despite the loss. Haley said she won't give up the fight while the majority of Americans are unhappy with both Donald Trump and Joe Biden. I'm Chris Caragio, NBC News Radio, NBC News on CACAA Lomelinda sponsored by Teamsters Local nineteen thirty two Protecting the Future of Working Families Teamsters nineteen thirty two dot org. The election is March fifth. Your ballots
are in the mail. Mail them in today. Now there's something to vote for because there's a new Marshall in town. Vote Derek Marshall for twenty third Congress by March fifth. Y'all know we need to change Congress. The current Congress has nothing but fight with each other. They've forgotten why we sent them there to work. Let's marshal and fix our problems at the border, a comprehensive plan, not political rhetoric and infighting. Yes, there's a new Marshal
in town. The rent is too darn high. Let's marshal in affordable housing and real, affordable universal healthcare.