Melissa Leone has a lot on her plate these days with three artificial intelligence (AI) ventures she started over the last five years.
Between getting each company off the ground — and trying to leverage one of them that could help tackle the wave of expected disinformation during the upcoming 2024 election — Leone finds herself juggling a lot these days, she tells Tribal Business News.
Leone currently stands at the head of three AI ventures: SpiderV2X, a 2018 start up which aims to build better communication between autonomous vehicles and smart systems; Hu-GPT, a 2023 venture which helps identify AI-developed or otherwise falsified content, and Datagent, another 2023 venture around an in-development AI project meant to provide companies with in-house AI assistants, cutting down on leaks while improving intra company communications and data access.
Hu-GPT stands out as especially important given the upcoming flood of political content ahead of the 2024 election, Leone said. Hu-GPT is meant to root out AI-developed or otherwise falsified content, such as deep fakes, which are convincing facsimiles of real, talking people. After spinning up Hu-GPT in January last year, Leone hopes to commemorate its first anniversary by providing “forensic-level” analysis of where content comes from and how it was made, such as metadata around its original location.
“Our goal is to help create environments where people are comfortable moving forward,” Leone said. “We want to make sure people can be protected.”
Leone, who is a member of the Oglala Lakota Sioux, spoke with Tribal Business News about her companies, what it was like getting into the tech industry, and advice for other Native people hoping to break through existing barriers to that sector.
How did you end up in the tech sector, working with AI?
I had a military and educational background. Education was important in our home, and so I was always growing and learning and always looking for new things to learn. Any time I went to my parents to ask about something, they would say, “I don’t know, look it up.”
I came from that background of, if I don’t know it, I can learn it, and it’s why technology and myself were able to go so far so quickly.
You didn’t run into any issues on your way up the ladder?
Information is information. It really empowered me to be able to break barriers that make it ordinarily difficult for a woman, let alone a minority, to break down. Technology really opened all those doors - the outcome was based solely on what information and technology I put in.
You coached professional soccer at one point in your career. How did that inform the work you do today?
The same skills that I used for team development as a professional soccer coach I have been using to help lead my companies to achieving success. Outside AI, I have led multiple startups, both my own and those I consulted for, to become more nimble, plan and execute faster and with greater success, and most importantly to capitalize on market opportunities others miss.
And that led you to AI. Has being Native affected your experiences at all, or how you’ve approached problems in these businesses?
Being Native, I view the world very differently. And even though science and technology is a whole different environment than the natural world, there are still solutions in the world already that help influence new solutions.
How so?
With SpiderV2X, we’ve built the world’s only bi-directional communication stack between autonomous vehicles and existing infrastructure. That solves a major problem for autonomous vehicles, since they rely on sensors. That’s like relying (only) on your sight — if we had a driver that couldn’t speak or hear, we would consider that to be a handicap.
And bi-directional communication solves that problem?
We've all been in this scenario where we've seen an ambulance (try to) go through an intersection and it can't because it's blocked up. Now EMS vehicles can communicate with the traffic light and avoid lane closures, and can communicate with other vehicles to keep a safe lane open. That’s a problem that other companies haven’t been able to solve.
Why not? And how did you come up with a solution in their stead?
We’ve all asked ourselves if we can make cell networks have the bandwidth to handle requests and information, and they can’t. So I asked myself, where else in the world is this kind of bidirectional information exchange happening?
The answer that came back was biological, and something that we as Native people have understood for a long time: the world has the largest communications network through mycelium, or basically a fungi network. We looked at how that network worked, and what kind of communication it was sending through plants and trees and each other, and I was able to come up with a solution.
You can’t — or won't — discuss that solution in detail yet.
I'm learning to be cautious with how I move forward with exceptional technology and surrounding myself with stronger teams.
Speaking of moving forward, you mentioned that AI-dentifi is preparing to launch a service expansion, is that right?
Currently, we analyze images and photos to determine if they’re fake, or if they’ve been manipulated in any way. We will soon be training the AI to analyze audio or video files, and the final phase will be concluded by the end of the month, which will allow us to do this on a forensic level, instead of just running through basic vectors that most companies run through their analysis with.
Verification seems like a pretty crucial service in this day and age.
No more will you be catfished on the internet. Guys won't have to be bamboozled by the good- looking girl that's not (who she says she is), and the same with a woman who might be unexpectedly targeted by someone using fake photographs.
I think this gives us power to authenticate what's happening in front of us.
How could AI be of use to Native tribes or Native communities?
My grandmother used to tell me about the rainbow around the sun (that) meant it was going to be a colder winter. So that's something really simple, but on a scientific level, AI can use weather patterns and help tribes make appropriate decisions on what crops to grow, when to grow them, migration patterns of animals for sustainability and hunting and those (types of) things.
But that means sharing information with an internal database. Do you think tribes will be on board with that in this era where they’re trying to define data sovereignty?
Part of that is from [training AIs on accurate information.] For tribal entities to have access to an AI that's theirs, that can house their own proprietary information — their history, their language, their ceremonies, their interactions within themselves. That all requires training of an AI that I have to honestly tell you that I'm skeptical about - that they're trained with real information.
I don’t doubt the key to moving forward is in AI, and I think having something like a datagent, or one that's housed as a Native databank would be a good thing. I think maybe as that databank grows, and as we make sure the information going in there is good and protected, we could then start asking questions that affect not only how do we move forward in a positive way but start making correlations like weather patterns from asking about our own history, or making language models so that we can teach our languages to our youth.
So what advice would you have for someone attempting to break into the tech industry?
Information is the best opportunity. It’s the power of the future. It’s not what you know, it’s where you go to get it. Ignorance is a choice, because knowledge these days is free.
Information doesn’t care if you’re sitting in a college class in Harvard or in your house on the rez. It’s just there.
Our biggest struggle is empowering our generation to know they have these tools. If you need help, most tribal entities have some type of educational assistance, or vocational opportunities … that can help remove the barriers (and help) get you viable internet and a computer. To me, those are the only things you need to arm yourself.
That’s how you got started - information, right? You’ve mentioned you don’t write code yourself.
I'm great at finding the process that creates the code I need to produce the outcome. You can teach yourself how to code from home. I created AI-dentify, and probably most people will not understand how I did it. It really is because I took my computer and I found myself a corner with some internet and I did the research to know what I needed to put together for this to happen.
If you're not finding that information, ask a better question. That's how you fine tune getting the results and information you're looking for.