Christian Ortiz: What generally happens is if you ask the question the wrong way, it may trigger or activate some sort of racial trauma, and the conversation spirals, and now the conversation doesn't become safe, and so then they pull away and they're like, well, I don't know how to do this the right way. And so, the AI provides this tool that just says, you can ask me whatever you want, and I'm just going to give you the answer.

Sam Fuqua: That's Christian Ortiz, and this is, Well, That Went Sideways! A podcast that serves as a resource to help people have healthy, respectful communication. We present a diversity of ideas, tools, and techniques to help you transform conflict in relationships of all kinds. In this episode, we talk with Christian Ortiz about using artificial intelligence to help improve communication and understanding around bias. He's an entrepreneur, founder of Mod Atlas Media, and the developer of Justice A.I.

I'm Sam Fuqua, co-host of the program with Alexis Miles. Hello, Alexis.

Alexis Miles: Hello, Sam.

Sam Fuqua: And we're very excited to be joined for this edition of Well, That Went Sideways! by Christian Ortiz. Hello.

Christian Ortiz: Hello, everyone. How are you? And thank you so much for having me today.

Alexis Miles: So Christian, I'd like to ask you to explain for people who are not familiar with AI, what it is and what's different about AI from Google, for example.

Christian Ortiz: Absolutely.

Alexis Miles: From a search engine.

Christian Ortiz: Yes. And that is a fantastic question. From, going back from bare bones to the concept, what a GPT normally does instead of, we have normally a question that we want to resort to Google to, and we will always put in some random question and Google gives us about 10 to 15, maybe hundreds of articles on this topic. And, we as individuals have to spend the time navigating through these articles and determining which sources are, are the best or who are our most trusted sources. And then if, if the information falls through in some sort of way, then we have to kind of backtrack and then do some more research. A GPT, um, pulls information that is not only, uh, as close to accurate as possible, but immediate. And it breaks down the, the solutions immediately for you. You can ask it to provide recipes if you're hosting a party and use it in practical ways of determining, I know that seven of my friends have gluten issues, two are, uh, vegans, I need a recipe that benefits all of them. And so, immediately it'll provide you with recipe ideas and it can give you historical details about places you're trying to travel to, and things of that nature. So, it is an immediate hyper-use of Google per se where it just saves you a lot of time and gathering information that's needed.

Alexis Miles: So, would it be fair to say that the answers are curated as opposed to a search engine that just throws out everything?

Christian Ortiz: That's exactly...

Alexis Miles: This is more curated to what a person is looking for.

Christian Ortiz: That's exactly right. Mm hmm.

Sam Fuqua: Like, most or all major developments in human life and technology, it can be good, it can be bad. Where do you come down on it's, uh, on the good versus evil of AI?

Christian Ortiz: The realities of AI, I always go into the conversation this way. I say, artificial intelligence is neither artificial nor is it intelligent. It is simply a regurgitation of societal norms. And so, the information that it pulls from is simply gathered by the typical biases that it pulls through our society. And, I was introduced to ChatGPT in its early stage and its early release of 2022. And being that I'm an activist and I tend to, um, engage with software and in different ways in terms of helping me excel in terms of providing, um, cultural information, I realized that the GPT was extremely biased. It was just missing the mark. And so, when we talk about these biases, this is a new conversation that is becoming a global conversation. We're not a society that typically, we don't try to recognize our biases and understand where they come from and understand how they're, they impact society in implicit ways. And so, what I wanted to do was create my own GPT, uh, field with BIPOC academia, uh, and years of knowledge from marginalized individuals who have been in the DEI space practically for over three decades, and collaborate with them to put together an AI GPT that completely does its best in removing all of its biases and decolonizing Eurocentric views that have been perpetuated in our systems. And so, to answer your question, I think AI has the potential to be an unbelievable tool. And when I see the way my AI is working in organizations and in personal use, it's revolutionary, and it blows my mind every time.

Sam Fuqua: Can you give us an example?

Christian Ortiz: For sure. So, what would normally happen is if someone uses ChatGPT, and I'll use a use case with a potential client that I have. I won't give their name, of course, but they are in real estate. They're a global real estate company. Uh, when I approached them with my software, they were really proud, uh, with the fact that they were using ChatGPT 4.0, which is the latest ChatGPT. They were like, “We use it all the time and we create all of our listings with it.” And I said, “Well, that's fantastic. And I'm sure it does a great job. It makes it sound professional and wonderful and it's output.” And so, I asked one of the realtors to develop, um, a listing in real time using ChatGPT 4.0, and of course, it did and it pulled out all of this amazing information and it made it the most perfect listing that you could ever think of. And then we ran it through my Justice A.I. software, and we discovered over six biases that were just completely blind to the naked eye. And immediately they were sold. They were like, “Okay, we see the power of this and we see how this can benefit us.” And so, for organizations who adopted, it becomes more of um, um, more than just DEI, it's a compliance tool, right? Well, when we're talking about compliance languages and being compliant all across the board, the conversation has always been DEI, but now it's extended to B and A, which is belonging and accessibility. And so, we want to make sure that these systems are checking the mark and checking the boxes for diversity, equity, inclusion, belonging, and accessibility platform. And so, that's what this tool is, is doing for many, many organizations and, and individuals.

Sam Fuqua: That's really interesting. Can you tell me one or two of the, uh, the six, uh, things that your, uh, AI flagged in that real estate text?

Christian Ortiz: Absolutely. So, um, when we were talking about the, um, the listing, the, one of the first things that it pinpointed was the language around family homes, um, versus disregarding the fact that not everybody has a family or wants a family and, and, you know, reformulating the verbage to something that meets people who have single family homes or perhaps somebody who has fur babies, right? And so, the language was perfectly reformulated to encapsulate all family dynamics. Another one was the, uh, usage of terms describing the area, making it sound like it was extremely, uh, lavish and lifestyle and what happens is it alienates our BIPOC community individuals who also have money, who are also looking for great places, but also want places where they feel safe and welcomed and seen. And so, the verbage then encapsulated, uh, it actually did a real live audit on the neighborhood to explain the diverse areas that surround the block as well, and these areas that they're talking about so that it, it really focuses on culture more than just a lifestyle that maybe pertains to one set of, group of individuals, right? And so, and the last one that, that was really stuck out was the accessibility piece. There was no mention of how accessible is this house for people who may be confined to a wheelchair or what happens if you are in crutches or what happens if you have mobility issues, and so, it created and reformulated this entire section just for that. And so, it was very, very impressive to see how we can fill in the blanks in all of those spaces.

Alexis Miles: You talked a little bit about the, the data that the GPT, is that the right, GPTs are trained on.

Christian Ortiz: Yeah.

Alexis Miles: So, and you talked about the data that your GPT is trained on.

Christian Ortiz: Correct.

Alexis Miles: So, does Chat, for example, or Gemini, do they have access to the data that you're using? And if so, why aren't they using it?

Christian Ortiz: Right. So, OpenAI, um, who’s the developer of ChatGPT, they allow other GPTs to integrate it from the backend. However, I would have to connect them from the backend using what they call APIs. And, they call it OpenAI because it's roughly an open sourced AI, and in, in my hopes in using an ethical GPT, it will teach others to be more ethical, for sure. Um, that is the goal. But when you use like Google Claude or Gemini, um, they're doing their best to also provide unbiased answers. I think it's actually, they're learning more and more. And when, when software like mine become more public, then it raises eyebrows and people who are in the other companies decide to say, uh, we need to do more in making this a more ethical tool. And so, I actually run tests side by side with, just to say I, the one that I developed versus Claude, I put in the same responses, and they try, they try their best but I would say they're probably at about, um, in 70 to 80 percent mark close to Justice. A.I. Justice A.I. just provides a little bit more context, a little bit more historical context, a little bit more of a decolonized context to help you see the bigger picture at play when it comes to really anything you ask it, especially from a societal case.

Alexis Miles: So, is Justice A.I. available to the public?

Christian Ortiz: It is. And so, uh, there are multiple ways that you can use it. There is the, if you pay for ChatGPT Pro, which is $20 a month, you have access to everyone's GPTs that we've developed. And so, it is, uh, to no extra cost. And it is a powerful tool. What I am doing is I am collaborating with organizations and 

corporations who are willing to adopt it into their systems so that, um, I create a custom GPT for them. And so, this allows users, I mean, it's, it's a little more for sure monthly, but what happens is we can fully customize it to understand you as the individual, as a corporation, as a user, it can understand your background, your goals, it could speak to you in ways, um, it will know you better than just anybody else with, the more honest you are about it, with it, and it's completely secure. So, when I create these custom GPTs, or you use Justice A.I., the information stays within the GPT. So, it is not being shared with developers. I own all the rights to it. And so, it becomes a safe space to have also conversations when individuals are trying to deconstruct from their biases as well. So, I always encourage individuals to use it because it is a judgment free zone when you're having those hard conversations.

Alexis Miles: I work in human resources. So, this kind of resource sounds like it would be excellent for people in that field because we do a lot of DEI work. We do have lots of questions about bias, about discrimination, about many, many, many things. So, can you talk about how it might apply in those kinds of situations?

Christian Ortiz: Absolutely. And I'm so glad that you asked this question because I come from a corporate background before I started my business. I worked with one of the largest credit card processors in the South, and I worked for major insurance companies, and I worked for several other corporations, and I've always found myself in my youth pigeonholed and met with the horrors of corporate America, right? And so, I got to see firsthand, um, the strengths and the weaknesses from a lot of HR departments. And, you know, how improperly they handled cases in terms of, um, employee privacy and things of that nature, or discrimination in the workplace, right? And so, and then understanding HR's role, that is to protect the company, not per se even the individual, even though they're, they're trying, there's a lot of stipulations and ways they have to speak to their people. And so, with my HR background as well, initially this tool was developed for HR positions. I wanted it to be the HR DEI tool that they could use because, uh, for all of my HR specialists out there, they know that HR doesn't have HR. And, what happens when you have an employee who comes to you with some outlandish case and, uh, the HR specialist just doesn't know how to navigate or have the cultural competency to be able to even speak to the matter that's at hand and, and that happens so frequently more so than not. And so, I wanted them to have a tool that they can prep themselves with to be able to navigate these complex situations.

And, it becomes not only a tool that, um, helps the individual become seen and heard and understood and, and they feel taken care of, but it also becomes the mitigation tool for the business, right? So, a lot of times, organizations immediately, first question they ask is, well, what's my return on investment on DEI? And, uh, and DEI becomes a very sensitive topic, and I had to realize that the conversation isn't actually DEI, it's compliance. And so, when we look at it from a compliance standpoint, the return on investment, one perspective is to look at the, in 2023 alone, the discriminatory lawsuit, practice lawsuit settlements from Fortune 500 companies was up to $126 billion, right? And so, and one of the top, top cases was a billion-dollar lawsuit on misgendering an individual. And so, what happens when you have an ethical AI that could prevent all of that, like, it just becomes the problem solver. So, to your point, I think HR would, is the number one target audience for this, for sure, in a corporate standpoint.

Sam Fuqua: As I'm understanding, what you're doing is, since the, the AI is a, is a machine, and it's learning from whatever its input is, you're giving it more input from, uh, BIPOC academics, for example. And then, it can help us if we're challenged or want to check text or are having a difficult conversation. But, what is the risk of, you know, depending on the AI rather than really learning and living as someone who is, uh, like all 

humans, biased, uh, learning and living from the education we're getting from, for example, Justice A.I.? It sort of can be a crutch, a very positive crutch, but a crutch nonetheless.

Christian Ortiz: For sure, and that's a very valid point, because when we're talking about using these systems from an ethical lens, part of that ethical usage is to remember that this tool will never become a replacement of us, yet an extension of us. And so, if we enter it with the mindset of, um, this is going to help me learn and navigate difficult situations, but at the end of the day, whatever the result is, I still have the obligation to review these results with my team, my leadership, we have to, we have to really take a look at, to understand the context of this and what the intent of each output is, if that makes sense. And so, um, I think there is a fine balance. This is a tool that I can leave with my, my companies, my HR companies, CEOs, people, like you said, who, who maybe don't have the best etiquette in email, and they're trying to figure out what's the nicest way to submit that. These are all perfect ways to use it, but I always encourage my consulting behind it because I am a DEIBA consultant as well, so there is that additive feature to it, right? And so, there's user guides and things of that nature as well. So, I think, uh, to the greater point of it all, there always needs to be parameters put in place so that we make sure that it is not only being used ethically, but, and responsibly, but that we for sure have that human input.

Alexis Miles: To Sam's point, are you seeing transformation in the humans that are interacting with AI as it relates to DEI and human resources, and just human relations?

Christian Ortiz: Yes, absolutely. One of the strategies that I've implemented for this specifically is I, I actually created a YouTube series called Decoding Bias, where, um, I provide examples of how to use the software in so many different ways. Um, I have TikTok videos showing people how it analyzes, um, uh, snippets either from biased political headlines to actual transcripts from bias videos of problematic dialogue occurring that individuals usually cannot see through or can't identify in the moment. And so, when they see those videos in particular, they're like, well, that's very interesting. I never thought of that, that perspective before. I never understood what's on the other side of what's happening there. Or so, they start seeing things. As an individual case use, I've had many of my, um, I've had conversations with a lot of my white friends in the South who, you know, they grew up in hyper-conservative households, and they just believe what they believe, and they never had a challenging perspective, and so now, um, what happens when they use Justice A.I. is that the AI removes, (a) it removes all of the emotion behind the conversation. This isn't a finger pointing conversation. They have the ability to ask it exactly what they want. And I have a video on this on one of my YouTubes, and one of this was probably my biggest case study, which was just beautiful to watch.

My friend said he told the AI, he says, "I'm a white man. I have a black co-worker woman who I really admire and I appreciate it. But, I told her the other day that she always sounds so direct and so angry when she talks to me. And I told her this to her face and she got extremely upset, and I don't understand what I said wrong." And what the AI did was in a non-taking-sides way, in an unbiased delivery, basically said, here are the factors that may be at play. There is this reality of this angry black woman trope, and it explained all of this. There's a reality that black women in particular experience in the workplace, where they are perceived this way and they don't, misunderstood with the way that they're, so it broke this whole dynamic down. And then it gave him recommendations on how to repair that friendship. And he asked it to expand on it. And then he came up with this apology, um, that he felt comfortable and memorized and literally brought it up to her, and it was probably two sentences, and they're really great friends today. And so, he has then since used the tool to help him understand how to deconstruct from his biases from all of his 

beliefs that he grew up with. And he told me, he says, “Ultimately, what I realized,” he told me, “is that I needed a safe space to be able to do this.”

And we got into this conversation of this reality that not many people, this conversation that I realized that isn't had enough, is that when our white brothers and sisters want to deconstruct from their biases, um, and they try to have these conversations with their Latino friends, their Asian friends, their African American friends, what generally happens is if you ask the question the wrong way, it may trigger or activate some sort of racial trauma and the conversation spirals, and now the conversation doesn't become safe. And so, then they pull away and they're like, well, I don't know how to do this the right way. And so, the AI provides this tool that just says, you can ask me whatever you want, and I'm just going to give you the answer. No emotion. No finger pointing. And so he said, “What I realized was that we need safe spaces for this.” And I said, “More importantly than safe spaces, we need brave spaces.” Because it is the bravery that it requires to even ask the question. And then the bravery it takes to learn.

Sam Fuqua: In the two years you've been deep into this work with AI, what's been the biggest surprise for you? Whether it's working with the machine or working with the people?

Christian Ortiz: That's a very good question. I would say both, for sure. I would say the, finding how biased our systems are has been surprising, and it's an education that's ongoing for me. Uh, when I first started the software, it was able to pick up 180 biases. Now it lists over 350. And so, that, that could all be implicit and used against marginalized individuals systemically and through society and through literature. And so, the education process from that has been surprising for me. I've been in this space probably since 2015, full fledged, uh, activist, deconstructing, constant education, and when you realize that there's so much more work to be done, it is a shocker, but it inspires me to do more. From the personal aspect with people, I think my biggest shock was, um, the hesitation that corporations have to adopt it, where they realize that it can actually reformulate inclusive cultures and build inclusive cultures, and you have people who've been stuck in their way since the 60s. A lot of my connections are still in the South, so I've reached out to Fortune 500 companies there, and I've had their DEI leaders and HR leaders love the tool. I don't have to sell it to anybody because they see the potential. And then one of the, one of the realities that I got met with the other day was when one of the DEI management looked at me and said, “We celebrate diversity probably better than most. But when it comes to making true inclusive change in our policies, that's when all the red flags start.” And so, this new challenge arose for me on how do I introduce this to corporations because the message is clear for everybody. Every company needs this, right? And so, but I'm not in the job of convincing. I'm in the job of inspiring and bringing those who are already doing the work along with me on the ride. And so, um, those are probably been my two biggest shockers.

Sam Fuqua: I get nervous sometimes about the speed at which, uh, the machine learns and how this is evolving so quickly. But, in your field, it might be, that might actually be a real benefit. So, if we're having this conversation two years from now, where do you think we'll be, you'll be, with this work?

Christian Ortiz: In two years, uh, my, my goal is to be in at least, uh, 40 to 50 Fortune 500 corporations doing the work, um, that way, and then having thousands upon thousands more use it through a personal, uh, lens. Educators, um, my goal is also to get this in the hands of all the libraries in the country, um, to have a, a Justice A.I. account that somebody could use for about an hour. Once the word gets out more and more, um, imagine the amount of, um, homeless people or struggling people who have access to technology to help them better themselves, um, in ways that, um, I, I also was homeless for a year, so I, I 

know the community so well, just in the year that I was, that was over 20 years ago. And so, to have resources like this is life changing and powerful. And to your, to your point, Sam, I think the reason why it also evolved so quickly is because the information has been here all along. It's just never been mainstream, and it's never been publicized in, in, um, in ways that traditional education has been publicized. And so I think when, when this becomes, when I've, start partnering up with more and more people, the tool becomes more and more powerful. And yes, to my benefit, I hope it happens overnight.

Alexis Miles: I would imagine people hearing this might feel some excitement, you know, to know that there's a safe place to go to ask all kinds of questions, no matter how those questions sound, no matter how filled with bias, assumptions and all of that. There's a safe place to go and get some feedback.

Christian Ortiz: Absolutely.

Alexis Miles: And you were talking about the feedback. And in my experience, you can specify, "I want some really gentle feedback," or, "Give it to me. Don't be soft. Don't beat around the bush. Just give it to me over the head." So you can ask for the kind of feedback that you want in a safe place and try things out ahead of time, especially in these sensitive, sensitive areas. Let's use, for example, someone who knows how to Google, but they've never used AI at all. Would you just take us through the very first step that person would take?

Christian Ortiz: Absolutely. And, what, normally what happens is they would have access to the website, chatgpt.com or chat.openai.com, and it would take them to their GPT. And they would, uh, hit this little message GPT. What happens is the GPT automatically gives you four examples of how to start a conversation. You hit a button and it tells you what it can do. For example, there's one here that says what to do, five creative things you can do with my kids art, and it breaks it down in five different ways. You can start these conversations, it automatically walks you through the process. Now, once you get familiarized with the system of asking questions in different ways, I always encourage users to look at YouTube videos. I teach GPT courses. I, I always encourage, there's so much knowledge out there for this. And, you know, I will never get upset if they don't sign up for my courses. I may not be easily digestible in the way that I deliver my information. So, I always encourage people to find their creators, find their, um, educators on this, and just, there, there's walkthroughs on how to use these things, right? And so, I always tell people, familiarize with your 3, GPT 3.5, ask as many questions as you could possibly ask it, um, and then pay for the upgrade, the $20 a month. Use the 4.0. Watch the difference that it makes. It's more intelligent. I believe by springtime, they're going to release 5.0, which is supposed to have genius level, um, you know, answers and responses, and no hallucinations. And it becomes an even better software, which is also going to be implemented in my Justice A.I. So it's even going to enhance that. So for the process, I would always say, start slow. Uh, don't beat yourself up when you're entering this. This is a lot of information.

Um, when Google came out in the early 90s, everybody thought it was either a fad or a nightmare or it was going to be the craziest thing and this is where we are with the GPTs. The GPTs are our tools that not only have revolutionized the world today, but they are going, just like the Internet, the GPTs today are going to look completely different in 20 years. And, we're going to get to a point where as business owners, as entrepreneurs, as corporations, they're going to all need a GPT just as much as they need a website today. So, this is where we're headed with this thing. And so, because it's such a large, um, initiative and a large transformative part of our society, to your point, uh, and telling the listeners on how to approach this, I just 

say approach it slowly, but be proactive and learn as much as you can, because this, this is really not going anywhere as long as we move forward.

Alexis Miles: And anybody can learn to use this. It's not reserved for the college educated, the elite.

Christian Ortiz: Absolutely.

Alexis Miles: Anybody can use it. Is that right?

Christian Ortiz: Can I add to that? That is a fantastic point. So I would love, my mother told me to share this story because it was, it was hilarious to me, um, and beautiful and actually emotional for me. My mother, um, came to America from Puerto Rico in the 60s. Third grade education, was never technologically inclined, she figured it out. She was resourceful. She figured out smartphones and laptops and computers, but tech and software was just not her thing. When she was growing up, she clearly grew up in a, in a, in a time where, especially in Puerto Rico and in Latino culture, misogyny runs rampant, homo, you know, homophobia, transphobia, all the things, all the biases run rampant in our, in our community. So, my mother had this upbringing with all of these ideologies that I grew up with and biases that I never adopted, but I, they were clearly present. And, you know, it was part, uh, it contributed to a lot of our head butting, that it does in a lot of our communities and a lot of our families. We have this reality that not a lot of people want to talk about. I'm completely open about it all because we've overcome it as a family unit.

But my mother had an issue with my kids, uh, being queer and we had a whole thing, right, uh, years and years ago. And, uh, I think it was this year she called me and she asked me how my son was. And she said, "Does he have any, uh, is he seeing any boys or seeing any girls?" And she said it so nonchalantly. And of course, I paused and I was like, I didn't know where this was going. And she said, "Did I lose you on the call?" I said, "No, I'm here." I said, uh, I said, "No, he's not seeing anybody." And she said, "Okay. Well, I just want him to be happy. And I want him, you know, I love him and (blah, blah, blah)," which, this whole narrative was completely foreign in our conversations. And she never had a comfort with this. And I asked her, I said, uh, I said, "I appreciate you asking that." And she goes, "I've got a, I've got a confession to make. I've been using your Justice A.I. tool, and I've been asking all of these things." She spent, I think, 16 hours with my tool, asking it all the questions and deconstructing from all of her things. And what I can tell you is that her experience is it made her feel like a new person, and she's in her 80s. And so, for that to be my ultimate test case, to tell you that this thing is, is so real and so powerful for any individual of any age, of any experience to be able to use it personally to deconstruct. It was a beautiful moment for, for our family and for all of us. And so I, I was kind of speechless.

Sam Fuqua: Well, Christian Ortiz, it's really been exciting and interesting to speak with you. Thank you.

Christian Ortiz: Same here. Thank you so much.

Sam Fuqua: Christian Ortiz is an entrepreneur, founder of Mod Atlas Media, and the developer of Justice A.I.

Thanks for listening to Well, That Went Sideways! We produce new episodes twice a month. You can find them wherever you get your podcasts and on our website, sidewayspod.org. We also have information on our guests, interview transcripts, and links to more conflict resolution resources. That's sidewayspod.org. Our production team is Mary Zinn, Jes Rau, Norma Johnson, Alexis Miles, Alia Thobani, and me, Sam Fuqua. Our theme music is by Mike Stewart. We produce these programs in Colorado, on the traditional lands of the Arapahoe, Cheyenne, and Ute nations. To learn more about the importance of land acknowledgement, visit our website, sidewayspod.org. And this podcast is a partnership with The Conflict Center, a Denver-based nonprofit that provides practical skills and training for addressing everyday conflicts. Find out more at conflictcenter.org.