All right. So I wanna, I wanna have a little bit of fun today.I'd like to, um have you use your imagination along with me to think about the future and, you know, whenever we talk about A I or, you know, the future of tech and there's always someone who's worried, oh, you know, um our, our robot robot overlords are gonna, you know, we're gonna get this thinking machine, it's gonna, it's gonna start taking over and we're all gonna have to do exactly what it says.

And, and so today I want to talk about um what I see is the, the real future of tech and A I um as enhanced by um all of this tooling. Um I don't think we need to be afraid necessarily of our robot overlords, but we do need to think about the makers of tech and how they influence the tech in our lives. So I want to start by just setting the stage, you know, uh I know we all know this but just a few locations around the world power, the world's technology and I, I'm from Silicon Valley. Um So there's some pretty interesting uh statistics um in terms of annual gross domestic product, California's tech industry outproduces every nation except Qatar. Um And this output at which is pegged at 275 billion by the Federal Bureau of Economic Analysis, higher than, than some countries in Finland's, for example. So, although Silicon Valley and California's, you know, uh tech industry is home to less than 2 million people, more than half of the world's tech billionaires, the people who are funding these new, these new companies live in Silicon Valley.

Now, let's think about that population and get even more granular um technology makers are predominantly male, nothing wrong with men or male technology makers. But um it's, it's quite the trend. So uh there's a, you know, some, some uh women in tech surveys. Um one of these surveys reveals that 12% of engineers at tech start ups are women. Only 11% of executive positions at tech companies are held by women and we even lost some ground in this regard. Over the last couple of years of pandemic, only 5% of leadership positions uh in the tech sector are held by women and women make up only 7% of the partners at the top 100 venture capital firms. Um These are the firms that are funding these new technologies that are coming uh to the marketplace. Finally, more than 30% of women over the age of 35 are still in junior positions in tech and women are far more likely to be in junior positions than men, regardless of their age. Um, and, and so, you know, I'm an exception to that. Right. I'm in the C suite. I sit on boards of directors, I'm female.

I've got a technical role, but I can tell you that I'm very, very often the only one like me, um, in, in any of the, these halls of technology companies, tech makers are also young and, and that matters a bit too as we start to see this play out. Um 70% of all tech workers are going to be millennial or Gen Z by 2026. Um And so, you know, these, these new generations, they're a little bit different um than the older generations and, and how they show up is kind of interesting. Um you know, millennials come from the age of internet. Um They're oriented to self, they're questioning um they care about experience and the festivals and the travel, that's what their consumption is focused on. Gen Z. They're, they're truth seekers. Um They tend to be comfortable with multiple, these multiple online identities. In fact, they're identity neutral. Um ha very comfortable with um undefined I DS and their dialogue or um their consumption, they're used to a world of public cloud with seemingly unlimited consumption.

And so the things that they build and interact with, they're not worried about how much data or bandwidth is that using um as they're, as they're building these things. Um In fact, um ageism is an important issue um across industries but particularly true in tech. So for tech, um Gen Xers are being hired 33% less than their workforce representation. So Gen X is the generation before for uh the millennial generation, they're, they're hired 30 33% less than their representation in normal population. Millennials are being hired 50% more often than their workforce repre representation in tech and baby boomers. The generation before gen X are six, 50% less likely to be hired than their work force representation in, in these tech fields. So this means that the people creating our technology these days tend to be predominantly from a few locations in the world. They tend to be male and they tend to be young. All right. So let's think about what impact that might have um context, the, the context from which you and I view the world matters. Um Look what I see from where I sit. Um sometimes feels like all there is um and often that there's this um Nobel Prize winning economist Daniel Kahneman who says, when people think they're being creative or thinking outside the box in reality, your box is defined by what you've heard about what you've seen.

And so, um you know, let's let's sort of uh dial that back if I'm a young male from California and I am creating technology to be used by consumers around the world. Um I may think I'm doing something very creative and very useful for those consumers around the world. But I'm thinking about that from my context. Now, let's think about another component here. Um There, there's a uh author who I love called James Clear. And he created, he wrote a book called Atomic Habits. He says, um he talking about learning and how people um adapt while it's important to know as much as possible before you start real learning comes from experience from trial and error, trying different methods to see what works best for you. Um So how do I build a mental picture of myself, my own context who I am. He says, um that, that my experience of who I am happens through doing, through being and doing so every action I take becomes like a little mini vote for the type of person I wanna become.

If I'm helping, you know, people across the street who need it on a regular basis, I'm a helpful person. Um If I want to learn to speak Spanish, um and I study, uh you know, uh Spanish language every Tuesday night for 20 minutes, every session is, is sort of a vote that goes into my head that says, oh, I'm a studious person. I'm learning Spanish. I'm a Spanish speaker. I'm speaking Spanish and I end up viewing my habits as evidence for the type of person that I want to become interestingly that's the exact same way that we train machine learning. Um is, is, is we provide it with, with tons of uh individual data points that sort of create points um that create votes that helped me to understand um how, how to make decisions, what are the best decisions? Um What is it that, that person said the last 20 times, they, they, they uh made an utterance that sounded like that. Um They were talking about something in particular. All right. So let's pause for a moment at sort of building the foundation and I want to talk about this, this, you know, the future.

So in my dystopian nightmares, um I could see us reaching a point where A I and machine learning tooling because it was built by people who are not like me becomes insensitive to me and my needs and the way that I work, um I have this great example, which is, um, my, my mobile phone.

I, I happen to have, um, an apple, an iphone and um I use Siri to interact with my iphone and, uh for the last 10 years or so. Um I have been voice dictating to Siri when I want to do something like send a text message. And I also have a daughter named Holly. It's spelled Holly like the uh plant every time that I say that I voice dictate the word Holli to Siri. Siri spells Holly Hol I like the Hindu holiday. Now, Siri and I have had a relationship for the last 10 years. My daughter's been alive longer than that. But in all this time, Siri hasn't ever learned that when I say Holly, I mean, most likely my daughter's name, not the Hindu holiday. Um Sir, uh you know, database and her interaction design or his interaction design, et cetera are all a reflection of the maker um of the training data of all the votes that went into Siri. And because there aren't interface um for, for Siri to take on board, um my unique utterances and my unique um uh meanings. Siri is no longer fit to purpose for me. I have to type in Holly every time I want to send a text to my daughter um in the future.

If, if these kinds of trends continue, maybe we'll get to the point where my self driving car doesn't understand my voice commands because I'm uh differently abled I speak more slowly than someone else or I have a particular accent that's not understandable. Uh Maybe my my smart home doesn't, doesn't, you know, let me in the door locks me out because um maybe I'm unable to, to do the, the the series of biometric moves that my smart home requires or I can't do them quickly enough or can't the right sequence. Um or maybe ultimately work is so infused with these tools that my skills and my perspective uh can't make their way into the machine. And ultimately, they're devalued to the point where I can't support my family. Of course, there's the better outcome in my utopian daydreams.

Um You know, hopefully my son and daughter will have rich, personal and social lives grounded in reality, but infused by immersed in technology. Hopefully I'll have, you know, AAA variety of work options and I will be able to thrive in this, this gig and traditional and digital modes of, of, of working. Um Maybe we'll also get to the point where there's no longer a tech worker shortage. And um the wisdom that people of my age have amassed is also augmented with um the the delivery capabilities of technology. And ultimately, my hope is that diversity, especially diversity of makers and diversity of product is structural. But how do we make that structural A I reflects the people who build it? Um Kate Crawford of the A I now res research institute says artificial intelligence will because of all these rules that we've mentioned previously reflect the value of its creators. So, inclusivity in A I matters from who designs it to who sits on the company boards and which ethical perspectives are included. Otherwise we risk constructing machine intelligence that mirrors a narrow and privileged vision of society with its old familiar biases and stereotypes. So how can we help?

How can we change this if you are in a position in a company? Um to hire people and to, to buy product. Um You need to seek experience. Um for example, and there's, there's hundreds of these examples, but I I like some of these in particular, there's an organization called Workaround, which, which is a training data annotation platform that takes your data from incomplete to A I ready and it refines your data and builds data sets alongside top universities and companies workaround is peopled is staffed by refugees um who do human intelligence tasks such as translation, image tagging and research.

So these are people who may have escaped. Um you know, some of the conflicts in the Middle East um who are on long term refugee camps who speak different languages who are highly educated but have um different worldview and different um view of technology and they are um a workforce that's available to help create these training data sets.

Um Andela is another um amazing um uh company. Uh it invests in Africa's most talented software engineers. Um They've hired uh the less than 1% of the over 100 and 30,000 applicants that have applied to them um to work as full time distributed team members from their tech campuses in Lagos, Nairobi and Kampala. So here are um the the cream of the crop, um technologists from various Afri African nations who are there to help you build product. Um Another thing that you can do um if you speak differently than sort of that, that mundane average voice. Um if you have an accent, if you're older and speak more slowly, if you're differently abled. Um Mozilla has this very cool project called a Common Voice that allows you to read scripts and, and read those scripts into um the, the, the listening device and your accent, your inflections, uh your phrasing, your pronunciation becomes part of the training data for natural language processing tooling.

And then you can also help to um if you happen to speak a particular language or you know about an area of um uh you know, technology or history, you can also listen to how things are pronounced and, and how the the natural language processing is processing things and help to make corrections so that um natural language processing tooling is more accessible.

Um At the end of the day, it is OK to ask your technology suppliers about the diversity of their own employees. After all, you want to purchase tooling for your company that enables your company to be competitive, that enables your company to deliver products and have employee productivity from all of the workers who are in the workforce and all of the consumers that are in the marketplace, who built that product that's analyzing and shaping and extending your company.

How did they train their systems? Where do they get their data? Um What do they look like and think like and what are we missing if we're thinking inside of um the outside of their box at the end of the day. Um What I'm talking about is a process and procedures. Look, diversity isn't an accident. It's not a nice to have, it's an outcome. Um So I encourage you to just insist on diverse sources for technology innovation and machine learning training, data. And then don't wait, it doesn't matter how much influence you have. It doesn't matter if all you could do was participate in, in um the Common Voice project or if you're purchasing like I do, you know, millions of dollars of product a year. Um The human race is about to fall because the machine is taking control, you think, you know, but they know you more. Um The past is gone, it won't ever return and the future is false. Um The future is already here. The future is now. And so this Millennial band um the rebels, you know, they're saying, look, it's always now like, don't wait for some of you eventual outcome, get involved in, in some way in, in um creating inclusive art, artificial intelligence. And if you have more questions, please reach out.

Um I tweet often I'm on linkedin.