I'm with Lucas and this is black tech, Green money. Doctor Nashalie Cephas is the Applied Science Manager for the Amazon AI team, focusing on fairness and AI. Based in Atlanta. She formerly led the Amazon A nine visual search and AR teams, which launched part Finder, the visual search tool for replacement parts on the Amazon Shopping app, in twenty eighteen. The team for Partfinder came from the Atlanta based startup part Pick. In twenty sixteen. Shout out to Jewel Burks.
Nashally was CTO. Doctor Sephus received her PhD from the School of Electrical and Computer Engineering at the Georgia Institute of Technology in twenty fourteen. There's a lot of conversations about AI happening in our general social discourse, and I wonder what's being overlooked.
Yeah, So, I think we often talk about the limitations
and the fear of technology. We often talk about how we don't feel like the technology was created for us and people in our communities, and to some degree, you know, I can definitely relate to that, but I do still stress the importance of not letting that be an excuse to not learn about the technology and not letting it be an excuse for us to just dig in, get your hands dirty, and see what you can do to make it work for you, because all the while, if
we are continuously shine away from it, then we're just getting further and further behind. And I literally see it every day. I've been in the tech industry, in the corporate and the startup industry world since twenty thirteen or so.
I started down this path when I started undergrad at Mississippi State in two thousand and three, and Georgia Tech I started in two thousand and eight, and so I literally have seen it and lived it all of my internships and all the companies you can imagine all of the research projects over the past twenty years or so, we are still left out of the conversation. We're still left behind. We're still not taking advantage of the technology and how it can work for us in our everyday lives.
So that's that's the message that I try to always tell people about when I'm speaking about it.
So when I when listening to you talk there, it gives me this idea that you know, when even when I have conversations on this podcast, it's about there, there's a great fear. You know, with AI, you know we hear the news talks about this is going to take jobs and not even go to restaurants now you know, you go to a drive through when you can hear AI talking to you.
These auto.
Order takers are taking orders now because it was for a long time it was hard to get people to fill those roles. So they true to a certain extent, these are going to take jobs. And so when you have communities of people that you deal with everyday, communities that we live in, who do see this happening? What is the offense we can play instead of just defense? And how do we play that offense strategically and just tactically, like what should we be doing?
Yeah, so it can be as simple as reading a blog on AI. Letting that be a part of your regular routine, staying in the know about the news and what's new, what's to come. We recently saw that some companies have released texts to video AI and generative AI. They have now released instead of just in putting texts into a chat boy, you can now input texts and images to be able to create outputs. So they're becoming what we call multimodal. There are also new business segments
that are coming out. I also talk about because people have a concern about the technology and the data that it's being trained on. We now need our and will soon be mandated by policy and government to have special companies come in and test your technology and do what we call red teaming, intentionally try to break it and see what the issues are. All these are new business opportunities and who better to come up with business opportunities to test for biases than those who are in the minority.
And so we have even at my role and working as a principal AI scientist at Amazon, we have done contrasts with several minority minority businesses to help get testing data to help test our algorithms for bias, biases and bias evaluation and other areas of responsible AI. On the other hand, you look at the work that I've done in the community, for example with bean Path in Jackson, Mississippi, my hometown. We have and very intentional about going straight
to the community. This is literally what you can start doing today. You can start using this for your business. Here are the tools. We even have workshops set up you can come in for free. You could be a mom and pop business. You can be a startup company, you can be a mid sized business, real estate company that's been in operation for over you know, two three decades.
We have something for you. We also have a Senior Citizen program where we teach senior systems how to come in, get on the computer, not just check for their adoptor's appointments and log in and get their lab results. They are now able to engage in social media that now are able to engage in how to build a website for UH you know, a group like a Sunday school class or something like that. We also impact the youth.
We have summer camps, we have UH spring camps, fill camps, we have stem Saturdays where they come in and engage in our maker space and able to use not just AI at a computer, but tangible things you can touch, like use AI to plan a guarden and use AI to design a T shirt. And so we try to make it as engaging and as culturally relevant as possible to let people know that AI is for everyone and
this technology is for you. You belong here, even if you're not going to major in this field and get a computer science degree, which I want you to do, but I understand everybody's not going to do that. There's still a place for you.
I love that, and so I feel like you know more and more as we have this AI conversation. You know, when people say AI, it's like saying Africa, Like we don't realize there's one hundred one hundred countries inside Africa and people just say I'm going to Africa, Like okay,
where in Africa? It's like it's a whole continent. And so if you can talk about different AI use cases, different AI technologies so that we can start to part or what is happening in the world today because you've got chat botes, you got agi people are really afraid of. But can you break down what these things are?
Right? So, what we like to do when just describing this, we like to put it into two different categories. So there's what we call traditional AI, excuse me, traditional AI, which is what we normally would have thought about AI before this last couple of years, where you're training models like a computer to act, think, talk, respond, et cetera, like a human based on previous data that it has
learned from. So now we can make decisions about things in the future that we haven't seen before because of the things that we've learned from in the past. In this in this what we call an AI model, So using it for predictions, using it for data analytics, using it for it could be predicting what I'm saying in speech in terms of natural language processing. It can be in terms of programming or robot to move arms a certain way according to the control system that for every
action is an opposite reaction. There's also we saw in terms of recognizing things in images, recognizing things in video. That's the traditional AI that we were talking about now today. Most recently the last couple of years, we've seen generative AI become really popular, and that basically means AI framework that allows you to generate keyword genitive content based on some previous data. So we can be generating a conversation like in a chatbot, we can be generating images or
videos based on some prompt that you've given it. And so this is the era of genitive AI, which a lot of people use for things like marketing. If you're trying to create pictures for your website, if you're trying to get an understanding of something that you're trying to search for it can respond in a way that's very similar to a human by putting together the different uh words and a sentence that most likely, according to statistics, match the style that they think that you're looking for
based on the prompt that you gave it. And so that's what we call genitive AI. And again, the use cases are pretty much endless anything that you neique content for, which is a lot of different things, Like I said, marketing, website development, coming up with ideas on how to do pretty much anything, create a vacation planner, or create a middle plan, or even understand how can you go about, uh,
you know, building a dy project. Uh. Those are some of the use cases that we've seen for generative AI.
So what's so funny about this is, you know, during it's not funny, I shouldn't say that. It's so interesting about this is like during COVID there was this human among iss like freelancer movement, because people say, why am I going to this job when I can just go start my own thing. I can go start my own marketing company, my own photography business, et cetera, et cetera.
And then you have these technologies that are coming out and to your point, I can do my whole content calendar for social media by type and a few prompts in on chat GPT you know, four oh just came out, you know for Omni just dropped in this by this recording yesterday. And so I think about we kind of got boxed into this, you know, new world that we didn't even know, many of us didn't know we were
going to. So we've taken this leap in entrepreneurship and now you have this thing coming to each at lunch. And so how how do people who have taken that leap of entrepreneurship make sure that they're, if not building, leveraging it.
Absolutely. So I'll have to say this to my my MENTI. I have some mentees, you know, people that I mentored, and you know, this gen Z I'm a millennial, This gen Z generation they're different, you know, and even after them, the pandemic babies, they're different too. But I have to explain to them, you know, you have to make yourself marketable. I understand you not if you don't want to go to college. I understand you don't think it's worth it
monetary wise, especially if you don't have financial aid. You know, there are ways to make money. There are ways. They're very successful people who didn't go to college, But you have to figure out how to make yourself competitive in today's world. The same is true for any large tech company. The largest most successful tech companies are successful because they figured out how to constantly adapt. They figure out how to constantly move with the times and stay relevant and
stay competitive. And that's what you as an individual have to do with your business too, regardless of what it is. And so in terms of using AI, use it again, find a way to use it to your advantage. Find ways that nobody else is using it because we now have this interface on top of this, like you said,
this technology. People may not have seen it coming, but for example, we were using genati of AI in the actual AI industry, you know, like years ago, and so now many people can use it because these companies have figured out, okay, let's make it easier for everybody to interface to this, which I think was a great idea because it helps level of playing field. The one thing technology does is it's cold, right, it's computer programming code. Once it's released into the world, you can't really get
it back. That could be a good thing that could be a bad thing. But that's the one thing that you can't take away from us and people in our community. Once it's out there, is out there. So now we have access to it. Now we have these interfaces, we
can now use it to our advantage. And like I said, there are so many ways people are innovating every day with these these new technologies and interfacing figuring out how to use it for their business, whether it's on the internal operation side, like helping make you more efficient, or even on the external side, like generating content for other people that are part of your freelance business. And so
it really just takes some sitting down. You have the tools, you just have to let your mind take you to that place of innovation so that you can make a difference and you could be that competitive person in the market. I would definitely start by just educating yourself. There's no shortage of information out there, especially in today's information age. There are so many tutorials out there, a lot of them are free, and so if you're disciplined enough you
want to do that, then great. If you like to enroll in a program or course or an academy of some sort. I encourage you to do that too, but just don't stop learning, and remember you have the tool now to innovate.
It's particular the video on AI Generative AI. You know, it's happening so fast and so one of the critiques people have is, you know how you can tell you can look at an image and tell, okay now because the finger maybe it's eight fingers on one hand, like it's something because there's crazy stuff. But a year ago it couldn't even produce half of what is producing now. But it's happening so fast and so and I'm not going to give the negative spendity. I want you to
give the positive side of this. It is what excites you most about the future we're walking into, specific to how AI is going to impact our lives.
So I think it is really excited for me to have a conversation like this with just about anybody, even my grandmother in Jackson, Mississippi, because I tried for years to have this conversation with people, and you know, nobody knew what I was talking about. Nobody you know, like, oh yeah, what's these are smart girls, that's not for me. So finally we can sit at the table, we can talk about these things. At least even if it's on
a high level, it's at least something. And so what we've been trying to tell people, uh for the longest, now that they're believing it, they're seeing it, and hopefully they're able to take advantage of it. I think for me in particular, I always say some of the most primitive areas and industries that I think AI can impact
our healthcare and transportation. I love the applications that I'm seeing come down the line, especially in terms of look at the black community, especially black women, black women's health, mental health, you know, physical health, even even amongst the black male community, Like, there's so much innovation we can
do there. That's probably one of the most untapped areas that if you have the data, you can do a lot with AI and data science to be able to help people, you know, feel better, help people you operate better in their everyday lives. And I mentioned also transportation. I'm still waiting on my my flying car so I can fly with Tavvy in Atlanta, and so I just hope that you know, someday, real soon, I could do that.
So how do we get more Nashally's in places like you know, Jackson, Mississippi, because I think about what you're doing, and there's others there's people who you know, I won't mention names right now, but there's other people I know who are gone back home, back to their original hometowns and have recognized that it's not just about being in Silicon Valley, it's not just about being in Miami, but
they're back there. The people back home need what the future that they're being exposed to, And so how do we get more people to recognize this opportunity? Like what did you see or what called you? Pulled you back home and said, you know what I'm doing big in Atlanta and wherever else I'm at, but there's still work for me to do here in Jackson.
Yeah. Well, first of all, all my family still in Jackson, majority of all my family in Jackson and Mississippi area. So never will forget home, never will forget my people, my friends, my family. And then also it had so
much potential. I remember coming back home after being in Silicon Valley, in New York City and Atlanta overseas even as close as you know, Memphis and New Orleans, and you're seeing so much progression happen all around you, But in my hometown of Jacksonssissippi, I would see, you know, just I didn't see the same rate of change that I was seeing in other areas. And so I wanted to be a part of that and at least contribute in terms of STEM education, exposure and tech because that
was something I had been successful in. And so I knew that I contributed that, you know, started the nonprofit being Path and working on the real estate development to do a lot more there in Jackson. And so I knew that, you know, why not Jackson because it had definitely has the most potential probably out of any city I've seen this, and I think if we can do it there, we can do it anywhere.
So a big conversation that people who are concerned about AI have is where it's getting its data from. And you mentioned this earlier, like, you know, it learns from
historical you know, facts, historical deta sets, et cetera. And so I was listening to an episode of some podcasts two weeks ago maybe, and it was talking about how, you know, particularly like open AI and others have downloaded like entire Hollywood movie, you know, libraries and entire music catalogs by record labels and entire you know libraries from you know, random housing, et cetera, to feed the model.
Some of that on some of that illegally because those are copywritten you know materials and can you So there they are deep ethical concerns. The point here and so what concerns you when you think about how we're building this so that you can work in your own way to ensure that we're doing this in a way that's still fair to people and creators.
Even yeah, it we I mean, I can't say it enough. But so in the tech industry there's three percent. There's a stat that says three percent of the tech industry are black females, and that number has been the same for I think they said last twenty or thirty years. So that is a problem and I said black women, but the stats are similar for black men and look
at any other BIPOD group. I think that until we have diversity on the tech development teams, we will continue to see a lot of these issues because that's where a lot of this is flagged that you know, upper level management. I think we're seeing a lot more black and BIPOD CEOs of tech companies which is great. I'm hoping that they can help steer, you know, more inclusion
in these emerging technologies. But the person who's actually doing the coding, the person who's actually training the model, the person who's actually gathering the data and making sure that the data is good enough to go into the model. That's where we need that extra set of eyes to make sure that we're seeing the level of inclusion that we want to see and also thinking of the right bias testing that we should be adequately thinking about. Also a lot of the data. I mean, in this country,
we have no policies in place. I mean they're a lot closer than what we used to be just three years ago, twenty nineteen, but we're very inching along in terms of you know, what is considered in compliance, what is not in compliance, what are how do we hold companies accountable? How do we even put some standards in place? And it's very difficult. So it's not a trivial thing, even if everyone's hearts and minds were in the right place.
So it is something that we're constantly going to have to work at, and it's a what we call a shared responsibility model. Everybody has to play a part in this, But ultimately, I would say again it boils down to, you know, getting more diversity in the tech development roles.
What skills are going to be most valuable in a market that's driven by largely AI models, Like what skills when you talked about those kids who were like questioning going to college, you know, what skills should they be learning to ensure that they can still find a place in the market.
They they should definitely be very hands on, uh and at least somewhat tech savvy when it comes to these emergent technologies. I again, it's so much information out there, it's really no excuse to not learn. I mean, even look at social media. We've been exposed. Look at how
fast something can go viral. Look at how fast a trend can catch on across the entire world, or a dance on TikTok and take off that in that same vein these technical courses on how to learn how to use your to AI for marketing, how to learn how to use basic programming skills to help you solve problems, and using different products on the market that you know,
learning how to evaluate products. Even like all these skill sets are necessary and unfortunately a lot of our youth are you know, not engaging with these things, you know, at as they should be in order to be prepared. I do whole hardly believe that people can be prepared for this next wave. And just think about this is only the beginning, Like have we barely scratched the surface of because the technologies just now get into the hands of people. It's just no getting into the hands of
the black community. So we barely scratched it. Serve there's so much potential here and so I just hope that they, you know, take the bull by the horns and just run with them.
This is quote on your ex I wont say Twitter, but your x account where it's just maybe a couple of weeks ago, where you said tech is not black or white, it's green as in money. That's why we can't risk losing DEI diversity, equality inclusion in tech, furthering the financial power gaps in communities, financial empower gaps in the communities, from entrepreneurship to big tech. Find a way to make tech work for you, your company, or your community.
So with that what you said there, you know, as you and I know, there's only so few nationally, doctor doctor, you know, walking around and you know, I'm sure you walk into a lot of rooms where nobody looks like you or is your gender, you know, especially in the work that you do. And so to that last point you said, find a way to make it work for you, your company or your community. What was the way you found to make this work for you where you added value?
I saw the need to, uh, you know, you know, it's the saying, the biblical saying, teach a man how to give him a man to fish, You feed him for a day, teach him how to fish, feed him for a lifetime. So I wanted to help people feed themselves for a lifetime. So I started the nonprofit bean Path in my hometown of Jackson, Mississippi, which has no shortage of challenges. Are you gotta do is watch the news or google Jackson, Mississippi to see all the challenges.
And but I knew that, uh, if you can teach these people here, they will go on and teach other people, and they will teach other people because that's the cycle. And someone even showed me. I remember my eighth grade teacher sent me to my first engineering camp when I first realized what engineering actually was and if she had not done that, I would have majored in music. I probably was still really fine. But I think, you know, I love music too, but that was a game changer.
So if I can change the game for somebody else, hopefully they keep giving that cycle back and then the whole community benefits from that, you know. I think that's what did it for me. I mean everything else is just kind of you know. I love my role and as being a principal scientist and being able to interact
the cutting edge technology every day. I work with some brilliant people from Amazon to startups that I advise to other you know, techies, but the real passion comes from teaching people how to fish when it comes to technology.
So I'm kind of position this question a little bit ago, but you just brought it back up with you know, the teachers that what you would have been in music and et cetera. There had to be a conversation or a moment or et cetera where somebody pointed out, you know, this math and computer science area to you and it's like it finally registered, like this is the path of me or I don't know what your story is, but there's had to be a moment where it just you
could have went left and you went right. What was that? Can you tell more about that conversation that actually made the switch flip, because what we're struggling enough already is just having math principles in our community to be able to take these paths. So can you talk about what got your interest to the level where you're like, I'm in that's for me.
Yeah, I knew that, I mean the way this particular. So I've done math and science classes. Uh, you know, I was I was excelling in math and science as a kid. That was just kind of my thing. I wasn't really big on on history and all these other subjects, but that math and science was my thing. And so I remember that eighth that camp after eighth grade, they they just showed us, Hey, you type these letters and numbers in the computer, you can control things around you.
And I was just like, oh, what hold on, let me try this. And it was so fascinating to me. And I was like, wow, I one day I'm gonna be able to control the whole world. And that's really what it really and so and little did I know because this was you know, I mean, I graduated high school and like in three and so back then, like
computers didn't control everything like it does now. Like you know, I was, we went through the Y two K together and all that, and so we didn't know it was gonna be, you know, all this, and so I didn't even know that and then let let alone. How I ended up getting into A and knowing how that would take,
I never knew. I just was interested in it. And I think for people, we just got to find a way to show people in our community the possibilities that's out there and how it relates to them according to what they're interested in. I got into AI because I was interested in the Shazam app and I thought that was so cool. Because I was a musician, I was like, well, I want to hear this song. I was like, how is this thing working? This is so cool? And that
was the same techniques in AI and music. At the time it was even called AI, it was machine learning or pattern recognition or digital signal processing, all of that. Those are the same techniques were using today in AI and national language processing and jeergy of AI. And so you know, you never know where things will take. You just start with what you're interested in. So we present it in a way that you know, shows people. You know, hey, you now some people on the street, they've been doing
math for a long time. They're involved in some legal activities, but they know math. You know, we just got to make it relate to the things that they're used to, the things that you know, excite them.
You know, you started being path you know, your nonprofit to help along this route. And there's a lot of nonprofit people, you know, executive directors and community organization folks who will listen to this episode to hear you. And what role do you see those community organizations and nonprofits playing and making sure that we are educated and aware of what's happening.
Yeah, the the organizations and the community, they play a huge part in the ecosystem what I call the tech ecosystem because I think takings really at the center of it all. And people and so whereas you have the colleges and universities, you know, they're going to educate the people who are going to college. You have government, you know they're going to take care of their constituents. You
have the large corporations, you have the startup companies. But what about everybody else, which is the majority of people. Those who are not in school, those who don't have a startup company are working tech, and those who are not in government. That's the majority of people. If we touch those people, then we connect the gap in the ecosystem. And that's what Being Path focuses on, and that's what
a lot of other nonprofits that are similar. They try to focus on that gap in the community and make sure that those people are not forgotten.
Black Tech Green Money is a production and Blavity afro Tech on the Black Effect podcast Networking I Hire Media. It's produced by Morgan Debond and me Well Lucas, with additional production support by Sarah Ergon and Love Beach. Special thank you to Michael Davis and Kate McDonald. Learn more about my guests and other tech THIS'SPEP. It's to innovators afro Tech that the video version of this episode will drop the Black Tech Green Money on YouTube, So tap
in enjoying Black Tech Green Money. Shot up to somebody look at your money.
Peace and love,