The GenAI Revolution, Ep 3: How our Alumni are Wielding AI in Entrepreneurship and Big Tech

Listen on all major podcast platforms

Episode Transcript

Publishing Date: October 8, 2024

[00:00:00] Sam Altman: I think this is probably the greatest time, at least since the internet, to be a young person. If you're interested in entering the technology industry, this is a very special opportunity.

[00:00:14] Dr. Burge: We expose the students to design thinking, creating IP, and the use of patents and trademarks, starting as early as the intro to engineering class.

[00:00:23] Coretta: We are the makers and creators that drive those businesses forward. Why couldn't we actually be the true owners of those products?

[00:00:34] Kweli: The next part of our story leads us beyond campus to Howard alum who are wielding AI, both in big tech and as entrepreneurs, and how parts of Howard's tech and entrepreneurship ecosystem can support the creation of more like them, despite the shadow of the persistent black wealth gap and lack of black technology professionals and leaders, currently.

Let's dig into it.

Welcome to HU2U, the podcast where we bring today's important topics and stories from Howard university to you. I'm Dr. Kweli Zukeri, a Howard alum currently serving as the Assistant Vice President of Web Strategy and Innovation at Howard. And I'll be your host for this third and final episode of our three-part mini-series that dives into how Gen AI is affecting the globe with the focus on the university community.

We've noted throughout the mini-series that a major part of the origin of the bias problems in tech and technologies, such as Gen AI, is the lack of black people involved in product development and all levels of leadership within the companies that create them. So, I want to ground us just briefly with an understanding of the extent of the lack of African American participation in the technology industry, including tech entrepreneurship.

A 2022 report from the NAACP and Kapor Center entitled The State of Tech Diversity, the Black Tech Ecosystem, which was co-authored by Dr. Ivory Toldson, a professor and researcher at Howard School of Education, provides a useful overview. As it notes, while black people are 13% of the U.S. population, we only represent 7% of the computing workforce, and a disproportionate amount of even that group consists of lower level IT support jobs. In other words, by and large, we are not playing a significant role in the development and direction of this industry, which includes the AI space, although that space itself continues to rapidly grow and thrive.

As of the writing of that report, the tech sector employed 12.4 million jobs, added an additional 200,000 jobs annually, and was projected to expand four to five times the national rate of employment growth.

One big corporate of our lack of sufficient inclusion in the space stems from a lack of computer science foundational education. There's a digital divide, an opportunity to get that proceeds the so-called academic achievement gap in K-12 and higher education. The result of the lack of inclusion is the bias built into digital products and their applications that we've been discussing throughout this mini-series. This has very real-life implications for black people, which is being documented more and more. If you haven't already, I suggest you check out the groundbreaking documentary called Coded Bias on Netflix, just to name one illuminating source.

The state of black tech entrepreneurship is similarly bleak. According to the same report, despite all of the hollow commitments of corporations during the pandemic to invest more in black businesses in response to BLM, black founders still receive an astronomically low amount. Between February of 2020 and 2021, only about 1% of the almost $300 billion in venture capital investments was for black-owned businesses. So, needless to say, when it comes to equity in big tech and tech entrepreneurship, this society has a lot of work to do.

Okay, now that we've summarized the sobering state of things, let's move on to the positive stories of folks who are addressing these issues via their own personal success and by providing opportunities for others.

AfroTech is an annual conference that brings together black technologists, the tech industry, and black tech entrepreneurs around the goal of exploring ways to empower black people to succeed in both big tech and entrepreneurship. It's owned by Blavity Inc, a black-owned media company that reaches over 100 million black and multicultural consumers each month. Simone Tyler, an alumna of Howard school of communications, is the conference’s primary leader as Blavity’s SVP of Revenue in AfroTech. She shared some of her Howard story with me.

[00:04:58] Simone: Like most people that graduated from Howard, it’s the best experience of my life. I decided to go to Howard after having been at primarily white middle school and high school and feeling really excluded and also being really intrigued by the idea of going to an HBCU. I'm from California. We don't have any HBCUs. Well, we have had one, but we don't really have an HBCU community. And I'm like, “I got to get away.” Was really integral in me being able to create a life that made sense for me, but also to be prepared. So, I think a lot of the skills and things that I was taught at Howard, specifically in the school of C, and I look around at my peers and I'm like, “Oh, you needed that.”

[00:05:43] Kweli: After graduating and working in news and media for some time, Simone unexplained how she started working for Blavity in 2016.

[00:05:50] Simone: A friend of mine was like, "Hey, Blavity, who's the parent company of AfroTech is hiring. And they're building a team. You should take a look at it." And at the time, I'm like, “I don't know. Like, oh, the blog, like, that sounds really interesting, right?” But I'm not sure. Had a conversation with the CEO and then COO. I was sold. I was like, “This sounds like a place where I can match both my interest in promoting the black experience and celebrating black people and providing opportunities for diverse communities, and also, like, the knowledge I've built around how to also build business. Since then, I've probably had, I don't know, six, seven different roles within the organization. Came on the team, I think, as hire number five or six. And we've done a lot since then.

[00:06:39] Kweli: They really have done a lot. I went to my first effort to conference in 2016 in San Francisco when its attendance was roughly 1,300 people. Even then, the black folks in tech I met were so hyped to be in a space where everyone understood them and how it feels to be the only black person in the room at their company at times. And I'm always sitting there thinking to myself like, “Man, I'm so fortunate to not even know how that feels because of where I work.”

Last year, there were well over 20,000 people in attendance. I was surprised to link with friends and frat brothers there that are not even in tech. And I'm like, “What are you doing here?” Simone described this feeling well.

[00:07:23] Simone: AfroTech is, if they go to an HBCU, I often say it's, it's almost like a tech homecoming, right? So, it is a space that is designed originally to make sure that black people have access to the technology ecosystem. It is expanded to be not just getting jobs within tech spaces, but also to make sure that entrepreneurs are getting equal funding, or funding at all, frankly, to make sure that students are aware of different job paths that they can take and new opportunities that can take to also make sure that we're putting dollars in the hands of people that can make impactful decisions about what happens across the tech space and make sure that we have thought leaders that are celebrated.

So, for example, for years, people had no idea, like, who, right, was actually working in these companies? They're like, “Are there other black people in this space?” I'm like, “Yes,” right? And so, AfroTech was able to very quickly say, “Here are hundreds of people, if not thousands of people, that are either working in this space and can get you a job, working in this space and can connect you with investors and VCs and PE firms. For me, this is very much so legacy work. So, when people talk to me about why I'm still working with the AfroTech teams and building products at Blavity, a lot of it is because I know that, in 10, 15, 20 years, this is going to be a foundation for something that did not previously exist.

[00:08:44] Kweli: I expect that attendance will increase again at this year's conference in November in Houston, Texas. I asked Simone to tell me one reason why she's excited for this year, in particular.

[00:08:55] Simone: I think, too, really getting back to some of the roots of the work that we've done. So, really, really doubling down on founders and entrepreneurs, there is a huge opportunity for us to better connect people with money, frankly. A lot of resources and information are out there, but how can we do connect people with people that can actually give them money to run their business? So, that's one, that we've, kind of, like, we've always done, but really, like, that is the bread and butter of the work that we were originally doing when we launched AfroTech.

[00:09:23] Kweli: As Simone alluded to, AfroTech has a dual mission — to support advancement in both black tech entrepreneurship and in the technology industry, generally. During our discussion, I mentioned to Simone that, while I see a need for both, I think that, as much as possible, our black technologists should learn from the established spaces and then leverage their experience and knowledge to establish their own companies. She went on to explain the value of learning from within big tech as a starting point.

[00:09:51] Simone: How they build out their processes, their technology, how their teams are built, how they're building and scaling really quickly. Sitting on a team that's five people and watching it go to 20 to build a product actually teaches you an incredible amount about how to be an entrepreneur.

At the end of the day, I agree with you, which is, figure out how it works, take that, and build your own thing. And building your own thing, to me, does not mean that you have to build your only thing only for black people, right? There's no reason why, you know, I think Instagram and social networks are a huge example of that. We are the makers and creators that drive those businesses forward. Why couldn't we actually be the true owners of those products, right? There's a huge opportunity to go into the tech companies and, like, learn and also go get the big fat check, like, part of what I learned and truly believe is that a job at one of these organizations actually can, for many families, like, change the trajectory of their family and their wealth within their family.

And so, getting equity in one of these kinds of companies and being able to sell is incredibly valuable. And so, I don't want to say don't do it. I think that go in, get what you can, and then if you're made for it, because not everybody's made for it, like, go ahead and build your own thing, right? Or, or join a company. That's building.

[00:11:14] Kweli: One of the mainstays of AfroTech is the giant exhibition hall, which has very large activations and representation from all of the big tech companies and many others whose product portfolio includes or depends upon tech. This gives AfroTech the opportunity to hold them accountable for doing genuine equity work that they may espouse to.

[00:11:33] Simone: Oh, yeah, we have lots of fun closed-door conversations, building out a more formal advisory board for those kinds of conversations to happen. We expect our partners and our clients and companies that come to AfroTech to be true to the word, which is either come and hire, right, and prove that you're hiring people and/or tell us what your challenges are. So, a lot of people, it's a rough market. So, it's totally fine that an organization isn't necessarily hiring hundreds of thousands of roles, but then what are you doing that you can actually show up and provide value? And we have those conversations all the time.

And the truth is there's a lot of, like, a lot of the people that we're talking to are black people. They're black people that work in these organizations that are in DE&I roles, that are in roles that are designed to hire diverse talent. And so, they are, oftentimes… like, the work they do is incredibly hard because they have to go inside of their organization and convince people, yes, spend X hundreds of thousands of dollars to hire two to three black people, you know.

[00:12:34] Kweli: This is even more important, now that DEI initiatives are on the decline across corporate America, which, of course, includes the sex space.

[00:12:41] Simone: There's a huge, what I am calling an attack on DE&I space, right? And so, a lot of our existing organizations that we partner with are dealing with that. And so, we want to be really loud this year around our support of DE&I, even in a state that doesn't isn't as supportive, right, and has passed laws that are actually against DE&I programs at tech companies.

[00:13:08] Kweli: During last year's AfroTech, I moderated a panel with two folks at Howard that I really admire for their dedication to black tech entrepreneurship — Dr. Legand Burge, the professor you heard from episode two of this mini-series, and Erin Horne McKinney, the inaugural executive director of the HUPNC Center for Entrepreneurship. It was entitled, How HBCUs are Prepping Tech Entrepreneurs in AI and Beyond. During the panel, they should advice for students and entrepreneurs. Dr. Burge is the first show here.

[00:13:41] Dr. Burge: You guys that are tech students, you guys need to be participating in the ecosystem. The barrier to access to that stuff is so low with the communities like Hugging Face. Anybody familiar with Hugging Face? If you're not, go look it up. And different APIs that allow you to access some of these models like LangChain and OpenAI. You should be practicing. You should be trying to build your own solution. There's many problems out there that you probably have identified. Okay, try it out, you know.

The algorithmic justice league is one instance of this, but the HBCUs could come together as a center of, of AI fairness and, maybe, create like a Hugging Face where we take all of the models that are being put out in the open domain and we actually test them to test to see if they have biases, and rank them and make it public domain, make it public knowledge.

You know, I think that's one thing that we can do to ensure that, applications don't get put out there that’d be recommending folks if you say, you know, if you're searching for a black person, it brings up a picture of an ape or something like that. We won't have those kinds of incidents because we're part of regulating.

[00:14:45] Erin McKinney: There is really one thing I want to underline, in particular, is that there's no such thing as failure. There are learning experiences, right? And those are the things that make you grow.

And I think one of the great things about being at HBCUs, right, we are the culture creators at our colleges and, and universities. And so, people are coming to us for the culture. But one of the things that you get as a minor, no matter what HBCU you go to, is you get that grit, that hustle, learning to overcome obstacles, learning how to say, you know, learning when somebody says no to you, it's not really no. It's, you need to become a better problem-solver, right? So, learning how to problem-solve.

But I want to remind folks, especially for students in the room or even thinking about becoming an entrepreneur, one of the best times to do it is while you're a student or even in grad school, right? People don't think about going back to grad school. I have seen student entrepreneurs use their student loan money and grants. There's so many grants and things that are available to students that you don't get access to, or resources. Having opportunities to come to AfroTech on a scholarship, things like that, that you cannot get unless you're a student.

It was a humbling experience, right, to be in my early 30s, a mom of two kids, and divorced to be a graduate assistant, but had an opportunity to get into what were now the early days of tech policy was telecom policy at the time, that shifted into tech policy, but being able to work and listen to people and be able to do that in different ways. So, I think it's really important to be able to team-build at a university, college, university, you're also able to meet somebody in the school of engineering, meet somebody in communications, meet somebody in the school of business, and build the right team, right, from those… the people that you meet.

I just think that there is such an opportunity for HBCUs. We've been the thought leaders when it comes to the solutions for the black community since forever and the reason why our institutions were created.

And so, we can't stop that now, right? We need to continue to be the ones to be the change-makers, to be the innovators, to be the, not just entrepreneurs but intrapreneurs. So, even in the organizations that you're in, you have an opportunity. Each of us has an opportunity to say, “What are the black-owned businesses that we are using as our vendors? Are we creating supplier diversity?”

[00:17:03] Kweli: Both Erin and Dr. Burge are pivotal part of the ecosystem at Howard that nourishes tech entrepreneurship. Erin's HUPNC Center supports HBCU entrepreneurship on multiple levels — and not just at Howard. Well, it's a national center based on our campus. They have five regional centers, which all serve HBCUs in their respective areas. During our panel, she told us about their current work and some plans for the future.

What's your big vision of how HBCUs should be preparing resilient tech entrepreneurs?

[00:17:34] Erin Mckinney: Well, thank you for asking that because that's exactly what we're doing with the center, the HUPNC Center. This is a brand-new initiative. So, we just launched in April of this year. This is a historic, I said, grant from PNC. And it's the largest grant that they've ever given, but we're not fully funded, right?

What we're doing is, essentially, one, doing research, because black entrepreneurship itself is still very under-researched. And so, we're… we have something called faculty fellows that we fund faculty at HBCUs who are willing to research black entrepreneurship in every aspect of it, right? And we're actually looking to develop a scholarly journal around that, too, so that people can submit, because one of the issues that we, we've heard from our conference that we held, our Black Commerce Conference, from researchers is that, even when they do research, they have a hard time getting that published. We're using that research and that data to then inform, what are the gaps in the marketplace, and making sure that those, whether it's the resources, the access to capital, or things that they are culturally competent.

There are a number of programs that exist, but they're not really meant for us and intended for us. We also launched something called the HBCU Startup Scholars, which is a $10,000 scholarship for student entrepreneurs. We also have something called our Education Innovation Entrepreneurial Grant that we give to HBCUs, $25,000. So, if they're starting their own center, if they're starting a course, if they're creating a pitch competition, whatever they're trying to do annually, each of these grants and awards go out to support the ecosystem, right?

[00:19:07] Kweli: As I noted in episode two, Dr. Burge has been in Howard for 25 years as a professor and former department chair. During his tenure, he's helped shape that department into supporting multiple pathways for its students.

[00:19:18] Dr. Burge: I advocate all pathways for students. So, some students want to go and become professors. They want to come into academia. So, there's, there's a pathway for them. Some of them want to go into work for a company in the big tech company. And then some of them want to go into entrepreneurship.

So, I advocate for all of them. I do suggest that those that do want to go into the big tech, that they do seek out internships and get exposure to companies so that they can learn processes that actually occur, and they can get connections, network connections. One of my sayings is I'm tired of, of creating sharecroppers, you know.

You know, I advocate for every pathway, but I'm pushing that everybody tries to start something. They have the same capacity and capability that a Mark Zuckerberg or Bill Gates has. Matter of fact, they probably have more because they went school, they got training. Those two individuals did not even actually graduate from school. So, the students are very well-prepared.

Now, in terms of resources, they may not have… you know, that infrastructure may not be in place. That's a whole nother conversation.

[00:20:21] Kweli: And Dr. Burge has been helping push that conversation forward. In 2017, he helped establish Howard West, a pioneering tech exchange program that places Howard computer science students and Google's workplace for a semester. He explained the origin of the lack of inclusion in tech and how the program helps address it.

[00:20:40] Dr. Burge: There's many gaps. So, one is just making sure that black students, and particularly, those that come from HBCUs, have the core knowledge in the space. So, that's one. The other is just experience. And then the third is, like, into Silicon Valley or into big tech.

I think those are the, the main issues. There's different approaches that have been utilized in engineering to close these gaps. Apprenticeships is one way, co-ops, internships. At Howard, we actually did immersive learning. So, we created Howard West. We said, well, we did have companies, practitioners coming into our classroom. Well, in this case, we flipped it. We sent the students to the company, and when they were given their project, they were given real projects to work with on a real team of engineers. And so, we immersed them into the culture. So, those are the things that we, kind of, focused on.

Students have gone through the program. They have gotten job offers from, not only Google who has stepped up as the lead on this in terms of the main benefactor. But these students also go on to some of the top companies. And we've seen offers in terms of first time coming out of college, the offers have actually doubled, you know. So, we're getting students coming out making 280, in terms of salary. I'm like, “I think I need to go, I need to go get back out in the market and change my job when I see some of these offers.” But it's because of that experience. It's also because we're connected with, obviously, the brand of a company like Google that's made that possible.

[00:22:08] Kweli: In addition to creating pathways for success in big tech, about a decade ago, Dr. Burge incorporated design thinking and innovation into the department's curriculum to ensure that students possess the skill set that supports entrepreneurship.

[00:22:20] Dr. Burge: We've been involved with entrepreneurship since 2012. We need to give these skills and impart these skills to our students and have them have a intrapreneur mindset. Even if they don't go off and be an entrepreneur, they can have an intrapreneurial mindset so they can bring skills like design thinking into the workplace.

So, we expose the students to design thinking, we expose them to IP, creating IP, and the use of patents and trademarks. We do it starting as early as the intro to engineering class. And then we've actually integrated it into our senior capstones. We actually created two courses, one called Bison Start-up and Bison Accelerate. And those courses have garnered attention from incubators like YC, Y Combinator, which is responsible for launching Airbnb and Dropbox and all these big, big tech names. We worked with them to create Bison Accelerate. We would fund students to actually build the first version of their product or service and actually hit the button and launch. You know, a lot of times, people are just scared to hit the button. You know, put it out there.

[00:23:25] Kweli: The computer science department has also been helping students acquire some AI experience. Riana Battick graduated this year, and she told me a bit about class project she completed that involved AI models.

[00:23:37] Riana: The CS professors were very excited about AI. So, they encouraged us to try and use AI as much as we could. So, for example, for my senior project class, we had used this AI website to create our logo for our app that we were building. And another example, we had to create a system that was similar to Bison Web, but it would, kind of, be like an advising assistant. So, it would use the student's class history to help them achieve their academic goals while taking into account their preferences for scheduling classes or class recommendations. And our requirement for the project was to utilize AI in some way. And we did this by using a chatbot. So, when the user accesses the chatbot, it enters a loop to continuously tag inputs. And then each query is sent to the language processing model.

[00:24:30] Kweli: Rihana noticed that many of her peers were not as AI savvy, but she was definitely willing to help them out.

[00:24:35] Riana: My friends that were in other majors, even friends that I had in engineering, not, not CS, but engineering, they didn't really know how to use AI or they weren't using it to help them with anything in their classes, nothing like that. So, I would often, like, be putting them on to ChatGPT.

I have a friend that's in journalism. She doesn't know anything in AI. So, I was telling her that she could use it to come up with different interview questions. So, she could just tell it, like, what the interview was about, who she was interviewing, what she's trying to get out of the interview. And she was able to get, like, different interview questions for each of the interviews that she had coming up. So, that was good that it was able to help help.

[00:25:20] Kweli: Now, according to OpenAI founder, Sam Altman. Riana's generation is graduating in an ideal moment to take advantage of the major industry changes occurring due to Gen AI, which he noted when he visited Howard earlier this year.

[00:25:33] Sam Altman: I think this is probably the greatest time, at least since the internet, to be graduating, to be, you know, a young person. If you're interested in entering the technology industry, this is a very special opportunity that probably won't come along again for a while. You all got very, very lucky. It's at the birth of a new industry and a time of tremendous change when young people have the most advantage and the most opportunity. There's a reason that I think young people drive a lot of the technological revolutions. You all are way more familiar with AI tools than people older than you. You bring a new set of fresh perspectives. What can be created now that just wasn't possible before this?

[00:26:14] Kweli: One of the projects that Rihana is working on now is building an AI-supported chatbot to advise content managers for university websites. I asked her to explain the underlying model.

[00:26:24] Riana: The model that we're using is called the RAG model. And basically, it's retrieving large amounts of data from a data set. It's basically utilizing a retrieval and generation model. So, in our case, the chatbot is hosted on Microsoft Teams. So, the RAG model helps the Teams chatbot provide accurate answers by first looking up the information in our data set. So, we have all of the data that's, like, relevant to content managers stored in different files in GitHub. And these are, kind of, just files that have information in a text format. We call it JSON, but JSON is just a fancy way for the model to retrieve the data in a standard way. So, the RAG model retrieves data from this large data set, and then it combines this information with its own language skills to generate a response using the OpenAI API that's also used. And then, this way, the user can ask a question and the bot will respond, again, in a conversational way that’s simplified and easy to understand.

[00:27:35] Kweli: The RAG model, as she described, contains a large language model for natural language processing and a separate data set for the chatbot's knowledge base. This has become a common approach to utilizing Gen AI without having to fine-tune a customed model.

Another product of the computer science department is Antonio McMichael, the principal software engineer at Microsoft who talked to us in episode one about what machine learning is and how it's related to Gen AI. He told me a bit of his Howard story.

[00:28:09] Antonio: During my time as a computer engineering student at RC, I was pretty much top of the class or close to top of the class in all of the CS classes I was taking. And I started to really enjoy it. Like, when I first took my first coding class, I took the book. It was, like, a, I don't know, 400-page C++ book. And whenever I went home with my family, we would be at IHOP. I got the book. I'm reading the book while we're waiting for the food to come. Like, I'm at the barbershop getting my hair cut. While I'm waiting for my time, I'm reading the book. So, I just really fell in love with the idea of being able to, to build things and create things. And that's always been something I've been passionate about, even since, like, kindergarten, elementary school, all the way up. So, when I got to Howard, it all, kind of, meshed together, and it was like the accumulation of all those different things throughout my life. And I really enjoyed and had a great time learning and being a part of that environment.

[00:28:57] Kweli: Antonio explained how the computer science department supported his ambitions to become a successful software engineer.

[00:29:02] Antonio: Having that opportunity to work with experienced professors and doctors in the industry really made a big difference early on. And then as I started to matriculate and get higher up into higher, harder classes, like, the professors, they were really supportive and knowledgeable. So, I think that was one of the big things, is that everyone had a background of research. We didn't type anything. Everything, all the code was handwritten. And so, that makes a big difference, because now you're not necessarily going off of the autocomplete or trying to figure out if you're doing the right thing, the right spelling. Like, everything, you have to internalize it. You have to understand the concepts. You have to really make sure you know what you're doing. And so, when you get to the point where you're getting that type of instruction, now, when you go and you're doing interviews and you're on a whiteboard and you're writing the code, you're confident. You know what you're doing. It makes it so that you're not in a situation where you, you may not be able to meet the expectations of those interviewers and things like that. So, having that support, having that rigor early on made a difference.

[00:30:00] Kweli: One of Antonio's professors, Dr. Burge, also became a mentor.

[00:30:04] Antonio: He was the department chair at the time. I took to him because he was just so passionate about everything technology. He still is. Like, he does so much with so many people. He's truly an inspiration to everyone at the department and even outside of the department. Like, when I run into alumni, we talk about Dr. Burge because he, he has an impact on everyone. I'll never forget, he, he made me a TA as a sophomore in the mobile app development class, which I mean, I don't know how often sophomores get to be TA, but I definitely enjoyed that process. And then he also took me on a trip to South Africa, Johannesburg, for the ICAT Hackathon, International Conference on Appropriate Technology.

[00:30:42] Kweli: Antonio explained the nature of the focus of his work as a principal software engineer at Microsoft.

[00:30:48] Antonio: Almost 10 years in. I think it'll be 10 years at Microsoft. Now, I'm at principal level, which is where a lot of people, they say this, if you get to this point, you get to principal level, that's a very successful career at Microsoft. And so, I'm still relatively young. I think there's a lot more I can do, and I'm looking forward to, to what can happen in the future.

I work on Windows servicing and delivery. It means anytime you get a new version of Windows, like Windows 10, the different iterations of Windows 10, Windows 11, whatever may be coming in the future, my team is the team that delivers that to all the billions of devices in the world. In addition to that, we're also responsible for keeping devices secure.

[00:31:24] Kweli: Antonio's current projects involve development of some of Microsoft's cutting-edge AI-powered products.

[00:31:30] Antonio: We're working directly with OpenAI, which is the, the company that created ChatGPT and exposed it to the industry and to the folks in the world. They're actually using our services to host that experience for their customers. Because of that, we've been able to, to work closely with them to provide our own products to folks that are interested in generating their own AI solutions.

[00:31:51] Kweli: Before he explains just what the new product line is, I want to give you some context to help make clear the significance. As I noted in episode one, one of the major drawbacks of the Gen AI revolution, thus far, is the unsustainably massive cloud computing power necessary to support it. In addition, even with a major amount of resources already thrown at this technology, it still runs up against bottlenecks sometimes when there's a heavier demand for its resources, i.e., a large amount of people that are using ChatGPT or other Gen AI products simultaneously. And therefore, one of those services can slow down or become temporarily unavailable.

[00:32:31] Antonio: Microsoft announced that they are starting to generate these new AI laptops. With these devices, we're, kind of, starting to shift from services to now having on-device models that can compute responses locally without having to deal with the, the network bottlenecks that may be there.

So, let me, let me back up. You have something called an NPU, which is a neural processing unit, which is a device or a processor that's dedicated to processing large language model requests or any ML model request that comes onto your device. There is a high number of resources required to generate these responses, to build these models, and to do things related to AI to generate these experiences that we're starting to become more accustomed to. And with the amount of processing needed, we're now starting to develop devices that have a specific processor that can actually offload the responsibility from the CPU so that it can work on regular global interactions with the operating system to now its own device that's dedicated to processing.

[00:33:27] Kweli: In other words, on these devices, the work of the Gen AI tools are not processed in the cloud. They are processed on the laptop via the NPU, or neural processing unit. Meaning, it can potentially work faster and not rely on network connections and, thus, avoid the issues I mentioned earlier.

[00:33:44] Antonio: And so, I'm working on a team that's actually working on preparing those devices for the market and making sure that, when they come to market, that they are optimized for the user to have the experiences that will be good for them and good for us as a company as well.

[00:33:58] Kweli: The NPU laptops were released in June of this year. And other companies are beginning to release similar products that localize Gen AI capabilities. I asked Antonio to share some of the lessons he’s learned on his career path with current computer science students that may be tuning in.

[00:34:13] Antonio: I think one of the best things I did as a student was not just do my schoolwork. You get a lot of foundation from the classes that you're taking that will teach you the what and the how, the academic side of things, but you have to be able to take the academic side and turn it into something practical. Like, you have to be able to actually build things.

And so, for me, I did a lot of side projects. I can't even count the number of side projects that we worked on. I was working with text-to-voice translations, haptic feedback sensors, guidance and vectorization of Google Maps API responses. So, I was doing all of that as a junior. So, that's another thing that I think Howard really does. If you identify someone that is really passionate and wants to do something, they find a way to help you find a position or a way to contribute and continue to expand your knowledge. And so, I think I'm, I’m grateful for those opportunities that were presented.

[00:35:04] Kweli: Now, to close this episode, we'll learn about the exciting and innovative work of one alumni entrepreneur who is wielding Gen AI to power her company's product. I met Coretta Martin at an ed tech conference at Howard last year. When she told me about her company and its tool, IEP&Me, as both a parent and psychologist whose research focuses on education, I immediately recognized how useful it must be right now for their customers, as well as the huge potential it has to grow further. Before getting into that, let's begin with Coretta's Howard story.

[00:35:38] Coretta: Howard University was the best decision I've ever made in my life. I absolutely loved all the people I met. I was very social. I've met so many of my lifelong friends. And when I was at the school of business, you know, one of our deans would always say something I'm sure you've heard, don't leave your groceries. You don't go to the grocery store, you get caught up in the samples and the aisles and all the things and leave your groceries. And the same thing would apply to Howard University. While here, be here, be in the moment, but don't leave without what you came for. And I feel like, being at Howard, I came here for an experience, a black experience to meet peers, to network, to socialize, and to increase my education. And I felt like I really left with those things, with those skills, and have been able to apply those all throughout my career in different ways.

[00:36:24] Kweli: After graduating from Howard, Coretta begin a career in education.

[00:36:28] Coretta: I'm one of those odd teachers who taught elementary school, middle school, and high school. Loved all three for different reasons, but pretty quickly pivoted into school leadership. I became an instructional coach, a director of special education, and eventually worked my way to being a chief of staff and a chief academic officer for charter school networks.

And throughout my experience as an educator, a teacher, a leader, there were a lot of things that came up frequently in the education space. One was equity, and another was just massive challenges around special education. So, that's something that lived with me when I was an educational aid, when I was a teacher. And I always thought about it.

So, that, as like a pause, while I was teaching, while I was in education, I also really loved entrepreneurship

[00:37:16] Kweli: During her time in education, working in various roles, as she noted, she realized the huge need related to IEPs. IEPs are individualized educational plans, which are created for students who are identified as needing extra learning assistance or accommodations, or what most people refer to as special education. They are legal documents that are co-created by a school, a child's parents, and a school district. The problem is that they're often very lengthy, static documents that are often hard to sift through and understand, let alone, be used to guide and collaborate on a child's education.

[00:37:51] Coretta: My co-founder was very passionate about students and student advocacy. So, the original idea was to think of, how can we help students in the special education process by giving them access to their IEP, by supporting them with a process called student-led IEP meetings, almost like a student-led conference, where the students are talking about themselves, their needs, versus a bunch of adults sitting in a room and telling them what they need? Oftentimes, when people say special education, their brains immediately go to certain classifications, more severe classifications, thinking that special education is actually a very large umbrella of disabilities and differences. And you can have an IEP for a lisp. You can have an IEP for a stutter. You can have an IEP for anxiety, for asthma, or for Down Syndrome, or for cerebral palsy. There's a very, very wide spectrum. And I think, a lot of times, particularly in communities of color, we worry about labeling because it's hurt us. We've been victims of labeling in so many cases, and labels have been used as a weapon against us. So, we sometimes fear labels that could actually support us, support our students, support our learners, and just getting them what they need due to the fears around the labels that go along with it.

And it's a real fear, but it's still something that we should have more conversations of, how do we inform people and educate people on the ins and outs of special education? As she and I collaborated, we thought more about other stakeholders, including the teachers, the parents, the outside service providers, kind of, like, the speech pathologists and the audiologists. And also, particularly in a lot of the communities that we serve, the entire village of adults that support a student, because it's not always mom, it's not always dad. Sometimes, it's a big sister, sometimes it's a football coach, sometimes it is the woman who checks you in at the rec center who really needs access to this information. And how can we use technology to build those relationships and truly collaborate with all the advocates for a given student to ensure their success?

[00:40:01] Kweli: So, in order to address what is largely an issue of collaboration and consensus building, IEP&Me was born.

[00:40:08] Coretta: We solved the problem by creating a platform that, one, you can share. You can share it with families. You can share it with general education teachers in a way that still keeps student data secure. And in addition to that, we add features that enhance collaboration, so that we can keep track of how a student's doing throughout the year and not only during an annual meeting to check in on this document. And we are in a place of just advancement where we can use technology to support them with being successful.

So, there's a lot of tools that support with some of these efficiencies, including just progress reporting — so, being able to communicate back with families, how a student's doing. This happens anyway in the special education process, but again, all of these processes are happening in different places. Word docs, paper, emails. We're taking all the different processes that a special educator needs to do in a day, in a week, in a quarter, and putting them all on one platform.

[00:43:21] Kweli: The first thing an individual user or someone from a school's administration does is input an IEP in PDF form. Then, the tool processes the information and populate the student's profile. Coretta gave me a quick demo of IEP&Me to help me understand firsthand just what the experience is like for one of the teachers, parents, students, or other stakeholders that use it.

[00:41:28] Coretta: So, I'll start with the teacher, but then also show you how parents and students are able to access the platform. So, when you log in as a teacher, you get to your dashboard where you see all the different features that you can use with IEP&Me. Some simple features like messaging and calendar here.

But really looking at what makes IEP&Me special for special education is, first, our student binders page. So, a lot of times, currently, educators receive IEPs in literal binders, three-ringed books. So, we wanted to play, pay tribute and almost an ode to the traditional binders that you receive by having a binder page where each student has a binder. When you click on the student's binder, and we'll click on Katie's, you get a really easy to read version of the IEP. So, we start with the About page giving you all the key information you need about the student, a strengths forward paragraph about who the student is, outside of all the other things we're going to about.

We also give access to all the binder members. This links with school's student information systems. We have a list that's really easy to find of all the adults that work with a particular student. This feature is really helpful because teacher turnover is real. Staffing is very challenging right now for a lot of schools. The speech pathologist yesterday may not be the speech pathologist next week. The school psychologist may change. Or even, sometimes, the student's schedule may change and they may have a different teacher.

When that happens, it can be difficult to know, who do I reach out to, who do I send the invitation to? So, this auto-population of binder members really helps teachers, as well as parents, know, who can I reach out to about this student in different areas?

And just to quickly show you what it looks like for the student, very similar, but they're able to quickly get access to the information they need to advocate for themselves. So, looks the same way, but thinking about this user, thinking, “I'm a high schooler, iPhone's in my pocket anyway. I have ten accommodations. I want to advocate, and I have a substitute teacher today. How can I communicate that? Let me pull my IEP out of my pocket and show them, ‘Hey, I have an IEP. Can I advocate for this accommodation right now?’”

And just creating those systems so that, when we get to a place where there's more open conversations around accommodations, when students want to be that advocate for themselves, they have the tool to do so. Same with families, same with teachers.

[00:44:01] Kweli: Users of IEP&Me likely don't even realize that Gen AI is powering its functionality and ability to take a large file and generate the content that they interact with in the interface. She explained further how that's working.

[00:44:15] Coretta: So, that’s the primary way we're using artificial intelligence. We are able to take IEP documents in any format that they may come. And our model is able to extract the key information around goals, present levels, and then put them into this easy-to-read format. So, even if someone is transitioning to IEP&Me, we can pull the data without the manual entry that's currently required when most individuals have to switch systems, or if a parent were to try to sign up for a system on their own.

I think, along the lines of artificial intelligence, I can also show you here another way we use that is with our change language feature. So, again, we're looking at the goals. Parents are able to leave comments. Teachers are able to leave comments. But here in the United States, many, many of our families are speaking multiple languages at home. So, if I want to translate this to Vietnamese, I can very quickly do this.

[00:45:11] Kweli: Wow.

[00:45:11] Coretta: And our conversation almost automatically switches to Vietnamese. And what you'll see here is that it's not just the title, it's not just the goal itself, but even the conversation is translated into the language of the families, so that teachers can translate, parents can translate, so that you can have an actual discourse going back and forth without having to wait for a meeting with a translator. And we have the option to do up to 199 different languages.

[00:45:46] Kweli: In this episode, we got a small window into both Howard's tech and entrepreneurship ecosystems, learned about how the AfroTech Conference is being leveraged to support black people in those spaces more broadly, and learn about the work that Howard alum are doing with AI and Gen AI in big tech and within their business products.

Creating this mini-series was a deeply fulfilling experience for me because I got to speak with some really dynamic Howard alum, faculty, and other leaders helping shape aspects of the AI and Gen AI space, and then share their stories and insights with you. I send my deepest appreciation to them for being a part of this.

I also want to thank Caitlin Whyte, who was a huge support along the way, and the rest of the team from University FM, our production partner.

As must be clear by now, I'm excited about the potential Gen AI and other AI systems have to offer, which we really only have just begun to witness. I've heard others note that the technology's abilities are worse now than they'll ever be. Meaning, it's just getting started and there's still so much space for additional development and innovation.

And I want to remind us that the technology's bias, which we inevitably touched on throughout this series, as well as the many other negative aspects, some of which I summarized in episode one, such as human and environmental exploitation, are major problems that humanity must take seriously now.

The global momentum behind the technology's advancement, won't be stopped. And the extent to which societies can get ahead of addressing the present and foreseen issues that it presents will determine how things land in the near future. Will we allow it to support the worst of humanity, such as perpetuating unfair treatment of marginalized peoples and to power systems and hardware of war, genocide, and destruction? Or would humanity connect with the urgency and opportunity of the present moment and push for it to be leveraged for equity and sustainable prosperity for those communities who need it most? Let's keep watching and be catalysts to bring about outcomes that represent the most righteous of our collective intentions.

One last thought. As digital technologies advancement grows and takes more and more of the space in which we communicate, create, and focus our consciousness, it's so important that we nourish real-life connections to each other and the rest of the natural world, as well as intentionally create and maintain practices and habits that continue to engage and strengthen our own critical consciousness.

Humans have an essence that machines cannot replace. Cultural historian and Howard alumni, Anthony Browder, reminds us that people of African descent can ultimately pull from our own deep well of African heritage for inspiration and guidance, which is what he calls the real AI — or ancestral intelligence.

Thanks for tuning in to HU2U. In this third and final installment of our three-part mini-series, examining the impact of the emergent Gen AI revolution on people of African descent to Howard lens, I'm sad it's over. Feel free to send me your feedback. My contact info is in the show’s notes, and I'd love to hear from you. Also, please, subscribe to this podcast wherever you're listening, so you can catch future episodes. I'm Dr. Kweli Zukeri. See you next time. Peace!