Leadership in the Age of AI with IBM Data Scientist, Ivan Portilla

Leadership in the Age of AI with IT Architect and Data Scientist, Ivan Portilla

In this episode of The Leadership Habit, Jenn DeWall welcomes guest Ivan Portilla. Ivan is a Senior IT Architect and Data Scientist with the Cloud and Cognitive Software Group at IBM. He is also a prolific author and speaker, and he is recognized for his innovative work in data science, robotics, and artificial intelligence. Ivan is also a member of the IBM Academy of technology. This episode’s audio is from a webinar that we did recently with Ivan, where he talks about the skills that every organization will need to prepare to embrace AI or artificial intelligence. Enjoy! You can also watch this webinar on YouTube!

Full Transcript Below

Jenn DeWall:

Good morning. Good morning. We are going to get started in a few minutes, talking all about artificial intelligence and closing the skills gap. What we can do as an organization, as a leader, to make sure that we are as prepared as possible for the new requirement of skills that we will have of our employees in the future. I am so excited to have Ivan Portilla with us today. He is going to share valuable insight on what we can do. I know that you’re going to appreciate this conversation, plus you won’t hear me talk the whole time, which is probably great for some of you that have attended ours. So I’m going to go ahead. It is 9:00 AM Denver time, which means that it is time to go. We are starting out by talking about artificial intelligence, and one of the important things to think about as we do this as more and more artificial intelligence is entering the world more and more emotional intelligence must enter into leadership, AI, artificial intelligence, or AI as it’s commonly known is something that we can’t ignore.

And it’s something that we need to be very attentive to as leaders to know how we are skilling our workforce to prepare for these changes that artificial intelligence will bring. It’s not something that’s just resigned to a room where a data scientist reads out your data. It’s something that we all needed to be a part of. So that’s what we’re going to be talking about today with Ivan. For those that don’t know me, I’ll just be your host today. My name is Jenn DeWall, and I am a Leadership Development Strategist and facilitator for Crestcom. You can always email if you email me if you have questions or connect with me on LinkedIn, but I’m so excited to be bringing you, Ivan. For those that are unfamiliar with, Crestcom just a very brief overview. We’re a global leadership development organization that focuses on developing managers into leaders.

And today’s topic is one of the things that’s going to be very important for you as a leader. You need to understand soft skills. So without further ado, let me go ahead and introduce the man that will be hearing from Ivan Portilla, who is a data scientist with IBM edge consulting. He’s an author, a speaker, and is recognized for his innovative work in machine learning, artificial intelligence, and robotics. Ivan is also a member of the IBM Academy of technology. And today, Ivan will be talking about how you can upscale or prepare your workforce with the skills that they need to lead in the age of AI. Ivan, I’m going to go ahead and stop sharing my screen. So then you can go ahead and start, but Oh my gosh. Cause we’ve, you’ve just had so much to share. So go ahead and start your screen share here, Ivan. And again, everyone can throw in where they are from in the chat. I know Ivan would love to see where you’re from as well, but Ivan, the floor is yours.

Ivan Portilla:

Okay, perfect. And hello, everyone. I’m going to share something on the chat. If you’re looking in your Q&A bubble there, I’m going to share a link to the report that we’re going to be using for this talk. So I’m going to share my screen now, and please let me know, Jenn if you can see?

Jenn DeWall:

Yeah, there we go. Perfect.

Meet Ivan Portilla, a Data Scientist Pushing AI to the Edge

Ivan Portilla:

Perfect. Excellent. So before it started IBM legal team asked me to say this, this is my view, my opinions, not of my employers. So this is a typical disclaimer that you seen in many, many talks. So with that, let’s talk about the topic today. We’re going to be talking about closing that skills gap on the enterprise, using artificial intelligence. We’re going to share some strategies and some results from a multi-year study. Again, the link is there, and it’s also on the chat, and here’s my contact information.

So I’d love to hear from you. I’m part of a team of data scientists on the IBM Edge Computing Team. I’m in charge of pushing AI to the edge. So with that, let’s get this started. Perfect. So our agenda for today will be very simple. We’re going to focus on four areas. We’re going to talk about the importance of skills and talent in the organization. We’re going to see some of the challenges people are currently going through with skills acquisition and skills lifetime. And then, most importantly, the impact of intelligent automation. How is artificial intelligence impacting the skills shoulders in your organization? And then we’re going to close with some strategies, some recommendations from the authors of this study to close that skill gap using AI. So a very simple agenda. So with that, let’s get started.

The Importance of Skills and Talent in the Age of AI

So, what would you say is the main differentiation between a $2 trillion company and Average Joe startup? Not the multi-billion circle building or the release of brand new gadgets every year. It is the skills that they are going to have to innovate and thrive in the marketplace. And these skills are feeding the global economy. They are creating new opportunities. When they go into a particular location, they are bringing talent, high talent to that location. They’re paying taxes; they’re creating better infrastructure. Currently, it’s interesting to see with this global pandemic, the migration to other zones. Now that people can work remotely from home. That again, skills are fueling the global economy. And when we talk to these technology leaders, the CEOs from thousands of companies in dozens of industries and multiple regions and countries.

The Need for Soft Skills in the Age of AI

We’ve noted several important key factors here that not only technical skills are important. Now we’re seeing people’s skills or soft skills are gaining importance in these organizations. Particularly this out-performance in these organizations. They place skills as the foundation to grow and to innovate. They’re going to new markets areas, and they’ve put a lot of emphasis on investment in a shortage of new skills in their organizations. However, this skill shortage has been remediated with hiring and training, but that’s not enough.

Unfortunately, they cannot get all the skills they need. So we are coming up with the ways to gain those, particularly the soft skills, the behavioral skills, and organizations are doing two things. One thing is they are tapping into the new career employees. Usually, people in the thirties, forties that have been with the company 10, 20 years, and they’re going to retrain these new career workers to gain the behavioral skills, the soft skills to innovate. However, we also developed a program at IBM, and our former CEO, Ginni Rometty, called it “new-collar” jobs.

And these are jobs that will require more than a high school diploma, but not necessarily a college degree. There are certain skills— like web design or cybersecurity— that a community college or some additional experience path high school can teach those skills. So we created a program that we call Pathways to Technology and early college in high school. With this program, we train students entering high school, which in the US is ninth grade. In six years after ninth grade, which means high school plus two additional years in a community college, they can get an associate degree, and they can get hired by companies like ours.

And we’re currently in 24 countries, and we are working with 2200 school partners and additional 200 college partners to educate these kids on the P-TECH program. That’s something that I’m personally involved in. I’m a volunteer and a mentor here in Colorado, and I teach a weekly class on data in AI to these kids. And it’s interesting to see a brand new perspective of a high school student and grasp the complexity of this topic. And I strongly recommend you to check out that website, ptech.org, and at the mission to train these students. We also open up a new infrastructure of learning, and this is free digital content in areas that we believe are important in the marketplace right now. So you can go ahead and enroll. Here’s the URL (ptech.org). There are classes in cybersecurity and data science and professional skills, even artificial intelligence. And you will earn certificates, or what we call at IBM “badges” that tell us that you have taken the class and completed the requirements successfully. So both of those strategies are being used to help close that skill gap.

Here’s an example this summer, we had a “hackathon” as part of the COVID-19 effort in our company, in my students. And we came up with the idea of connecting students with senior citizens in adult care centers through correspondence. So they will write letters to the senior citizens. But we were able to match them using a personality API in the IBM data and AI portfolio. So we analyzed the way they write and were able to match compatible personalities through this tool. And that was very interesting to see the students come up with the idea, develop the algorithm, and apply it to be able to connect several schools in the area with senior adult centers. So that was a very impressive effort.

The Need for an Agile Mindset

Now let’s talk about what the executives are considering are the soft skills and behavioral skills important to them. So in this survey from 2016 to 2018, we noticed a very interesting trend. If you see in blue, those are the soft skills or behavioral skills. And notice that the most important one that in 2018 was this agility, this willingness to be flexible, to adapt to change. And you can see down nowadays with the pandemic. How can you promote a connection with your team while remote working? Also, they consider time management, not needing to be supervised, to prioritize workability, to work with a multi-diversity environment. And that’s very important because right now our nation, we see all these. It’s happening now with race. Working with people who are different minorities from different cultural backgrounds is very important for them. And also the ability to speak up. I try to, you know, in my class, encourage my students to present, to develop those communication skills, both oral and written. So notice the shift in what executives considered important. So those skills from 2016 to 2018, so a lot of emphasis on those soft skills.

Now, when they look at what is the criteria for successful innovation. Again, soft skills make the top of it on the list. Working as a team, be able to be flexible, how to demonstrate strong leadership. And none of these is usually taught in a formal class. This is something that you gain with experience. And that’s what is so important to us to take these very young students and be able to share with them what a real working environment looks like.

The Half-Life of Learned Skills is Shrinking

Now, let’s talk about the challenges of those skills. This is the second area that I want to share from this study. What do you think are today’s challenges with skills? So the employers shared with us, and these statistics are well documented on the white paper. The interesting thing here is that in typical organizations, about half the employers say they cannot find the skills they need. And if you go to larger organizations, they go almost two thirds. They cannot find the skills they need to be successful in the marketplace. And the reason they say is that the candidates don’t have the required experience or they have to retrain and reskill to be competitive, to be relevant in the marketplace. Some of the other metrics in this slide, we believe— or they believe— by 2030, there will be a shortage of about 85 million people with those skills. So, one of the reasons that we have these challenges with skills is something called the half-life of a skill. And what this means is how long it will take for your skills to become obsolete, or half of your skills to be obsolete.

And this is being noticed that currently, it’s about five years. So think about that. You went to school; you went to college, you went to a training facility. In about five years, your skills will be half relevant. So that’s a huge impact compared to a few years ago when the half-life of your skills would be something more like a decade or 15 years. So that’s a huge impact on being relevant in the workplace. Now, not only that, the half-life for the skill is in jeopardy—also the time to gain the skill. I remember a few years ago, if I needed to gain a new skill, I would go to class for four or five days, and I’ll be proficient in that particular skill. Nowadays, if I want to learn a new skill, it will take you more like 40 days, like two months almost. And this is also because of the depth of the skill. Skills have become very specialized. Like if I wanted to get a data science skill, I need to learn math. I need a math program and need to know communication and so on.

And also because of the skills that are necessary today are also soft skills. Skills that you need to grow in an environment that you can’t learn in a traditional classroom education or virtual education. So this is the impact of that. And in fact, only 41% of organizations say that they have the skills that they need to be successful in the marketplace. So almost half of the organizations have the skills they need to be successful. That’s a huge gap in the skills that they need right now.

How will Intelligent Automation Change the Workforce?

The third topic into something very close to what I do every day at work is what is called intelligent automation. That can be both a good thing, creating an opportunity and a challenge in showing you a picture of what I like robots that I work with. It’s called Pepper. Pepper is a humanoid robot rated by a Japanese company, SoftBank Robotics, that we have taken to different environments like a hotel concierge, or banks, or retail. And let me tell you a little bit about intelligent automation. So the impact of intelligent automation is that it’s going to improve the processes, business processes and provide consumers users with personalized experiences as well and enhancing decision-making.

Unfortunately, there is also an impact from that intelligent automation, and this is not new. We have had automation for thousands of years. We have, this is what it’s called the fourth industrial revolution. We also have the steam-engine, electricity, changing industry, and agriculture that have had a lot of automating throughout history. But the reason that intelligent automation is more critical is that it’s using advances in artificial intelligence where machines can learn from data. They don’t have to be explicitly programmed to produce those predictions or recommendations. It has a positive impact. And here are some of the areas that we have been working in my company.

I have been involved in financial fraud detection. For example, a few years ago, I worked with a large credit card bank. And we were dealing with about 60,000 transactions for a minute. Imagine being able to approve or disapprove a transaction. And if we get a lot of false positive, the merchant is losing money. The credit card company is losing money. And most importantly, you have a frustrated consumer.

I also work with manufacturing on edge computing. We are putting AI closer to where the data is captured. So we’re detecting, for instance, in a manufacturing line, where there is a defective part, we can detect it right there before it goes into a car and you get a bad experience as a consumer. So we’re pushing AI to the edge. We are also working with transportation right now. Particularly with the law of demand in air travel. We’ll help several airlines predict when travel will recover, and then they will be able to be prepared for that. So these are some of the areas that we were working with artificial intelligence.

The Positive Impact of Intelligent Automation

This is why businesses share with us. That is important—the positive impacts of intelligent automation. One of the areas I want to highlight here, one of the metrics, is the increase in insights for data. So every day, we produce about 2.5 quintillion bytes of data. This is about 2.5 exabytes of data. In most of these data, we call it dark data. About 80% of that data is not being processed, right there. It’s locked in silos of unstructured data. So how do you count a thumbs up? Or how do you count videos? How do you analyze all that data? And I also want to point out that it is improving worker productivity. So at IBM, we don’t say artificial intelligence. We switched the acronym. We said it’s “Intelligence, Augmented.” We are freeing up the humans to do higher-level thinking, apply what humans are good at, such as creativity and innovation.

Some of the areas are reporting organizational capabilities expansion and improved productivity, but there is a significant impact on the workforce. About two-thirds of people will require new roles and reskilling that they know exist today. For instance, have you ever heard the term robotic psychology? To be able to understand why your smart speakers recommended you something? That’s something of that nature, and there’s progress in robotics and automation that will require new skills in the next five years. So let me give you some specific metrics, what those mean. So when we did the study, they shared that in the 12th largest economies, about three and a half percent of jobs will be displaced or redeployed. And what that means is about 6 million jobs will be replaced by intelligence automation. So that’s a huge impact on those people losing their jobs. And if we look broader, we believe that I’m about 120 million people in the largest economy would need to have to be reskilled or retrained in the next three years with intelligence automation. And the reason for that, if you think about it, it’s about the economies of the labor force from Brazil and Canada combined, and all of that, because intelligence automation is making easier to personalize processes, improve productivity and give you enhanced decision capabilities. So there are real potential and real challenges here with intelligence automation.

One of my favorite books, AI Superpowers by Kai-Fu Lee, talks about this representation of what type of jobs are in danger, and what kind of jobs will a robot take my job? He shows you a quadrant of how safe a job is compared to other jobs, and it is based on whether that job in a structured or unstructured environment. So he has these safe zones. So jobs that have a higher social interaction like your hairstylist is, or your physical therapist, or your counselor. They live in a very unstructured environment that they will be safe from this intelligence automation. And then he has a danger zone where there is a lot of repetition. There’s very little contact with people like this watch or a truck driver that have the highest risk of being displaced by intelligence automation. And they have these two other zones, the human veneer, where even though you have a very structured, repeatable job, you have a lot of social interaction. That makes your job still safe. And there’s an area he calls slow creep, where even though you have high contact, you have a highly structured environment, which eventually will become the danger zone.

So now, let’s look at what executives are expecting of this impact on events— prepared for events—like intelligent automation. So in this chart, it shows in two colors what they believe is important for the effect of intelligence automation and how nations are preparing for that. So that’s the culture in the state. For instance, they have that developing science, technology, engineering, and math skills, they believe it’s 40%, but only 54% of the countries are preparing for that area. And we have others like promoting connectivity or providing reskilling. So it’s not just the responsibility of the employer. But also the nations and organizations that need to be prepared for these skills shortages.

How Can We Close the AI Skills Gap?

Let’s talk about the main area of my presentation. How do we close that skills gap using artificial intelligence and some of the recommendations of this study? Again, we frame it as a national and regional challenge, helping with the industry or the enterprise, but the individual is also responsible. You are responsible for the up-skilling or reskilling of your skills. So these organizations know that they need to act, that there’s going to be a skill shortage. And unfortunately, half of the organizations are doing nothing to retrain you. Very few are retraining in one or two skills, and that’s a huge gap. So if the organization is not helping you, it is pushed to the individual or the regional coordination to close that skill gap, which has been worrisome.

So the study proposed three areas that we can use artificial intelligence to close that gap in skills at the center of these strategies. And we divided up into three areas. We call the first one, “make it personal.” The second one, “turn up the transparency” of those skills and then reach out outside of your organization. So let me go into every single one of these areas with that example or how we can close the skill gap using artificial intelligence, this, that we didn’t make it personal. So I’m going to start with my own company. In IBM, we opened up the transparency and personalization of your skill, reskilling, and training. And personalization is important to us. Imagine when you go to a movie site, and you want to see a particular movie or particular genre, you want the tools to predict what you’re going to watch or what are you going to buy on a commercial website.

Personalize Your Training and Reskilling

So personalization also works with the companies. They want to personalize what skills are needed in the marketplace, what gaps do they have? So when I log into this website, the internal website for my learning, my employer is providing me a roadmap that is telling me, based on your experience and based on the market needs, these are the skills that you need that will make you successful and will make the company successful. Today, 8 out of 10 employees have the skills they need to be successful compared to 3 out of 10 as of let’s say five years ago. So that’s the transparency and personalization strategy.

The second one is to “turn up the transparency.” So we have an example here, a larger telecommunication company in the US, AT&T. In 2015, they opened up to all the employees and said, this is what we need to be successful in the marketplace. And look at what jobs are in high demand compared to jobs that we believe are going to become obsolete in the next few years. And they crafted a curriculum based on that strategy. They say, for you to be successful in our company, and make our company successful in the marketplace, these are the gaps that we would like you to take on. These are the skills we would like you to learn.

So they open up, they work with massive online companies like Udacity and a local university. And they created this curriculum for the employees, and that was the transparency. Be open about what they need and what they want you to learn based on market needs.

And the third one (look inside and out) is very interesting because every company thinks that they can do this on their own. So we have here a company called CEMEX, a global cement, and heavy construction company. They partnered with us, and we put together— in a university— put together a program for digital transformation. And they established a digital hub in Monterey, Mexico, to grow these skills with industry partners. And they used the universities. We use online learning like Coursera and Udacity, and they create this curriculum. So not a single company can do this work alone. They need to expand both internal and external learning to reach and close these skill gaps. So to recap, to close that skill gap: make it personal, customize the learning to your individual needs, and your company needs. Open up, be transparent, create recognition, tell your employees what the market needs are, and then look inside and outside of your company to find those skills.

So with that, this is the link to the study: The enterprise guide to closing the skills gap. Thank you very much to the authors for bringing this wonderful study. And with this, let’s open it up for questions and answers. Yes?

Live Q & A with Ivan Portilla

Jenn DeWall:

All right. Let’s hear Q & A. So, so much of that white paper for those that maybe joined late, what Ivan just presented was based on findings that IBM found from a multi-year global survey. And there’s a lot of data in there. I would definitely recommend checking it out and just seeing what you can learn because it is pretty glaring. To know that half of the executives surveyed feel like their organizations are not pursuing anything to close the skills gap. So we need to be mindful that what we are doing or what we’re not doing today could essentially pull us back. But go ahead. If you have questions for Ivan, I’m going to start with a few, but if you have questions for Ivan, his role in AI, how he supports IBM, anything regarding what we talked about with the white paper, go ahead and throw that in the chat.

Ivan, if you want to stop sharing your screen, you can absolutely do that. And, but just know that Ivan Ivan’s contact information is there. It’s IvanP@us.ibm.com. And if you want to find him on Twitter, it’s @IPortilla, or you can connect on LinkedIn.

So things to consider is that by 2030, from this white paper, by 2030, the global talent shortage could reach more than 85 million people. So the issue is not necessarily a shortage of workers, but it’s a shortage of workers with the right skill. Ivan, to be specific, what are some of the skills that you see as essential for leaders today?

What Skills are Essential for Leaders in the Age of AI?

Ivan Portilla:

One area that I believe is critical is agility. To be able to adapt to change and particularly what is called agile learning. So we want to create a culture of lifelong learning. Everything that you learn in school, in college, will be obsolete, as we saw, in 5-10 years. So you need to keep yourself current. What are the upcoming trends that I’m going to be good at, or I find interesting? How can I improve my skills in those areas? How can I communicate well with others? So not only the technical skills but also the soft skills, how can I work with people from other cultures? How can I adapt to change? How can I be flexible and react to these, for instance, this pandemic how can I work from home successfully, be my own boss, be successful in that area?

Will Data Replace Oil as the Most Important Natural Resource?

Jenn DeWall:

Yeah. It is connecting with others. You know, I think when we think, or when I think before I kind of got into understanding this subject, Is that artificial intelligence lends itself to more data or analytical skills. Whereas it does lend itself to understanding how are you communicating? Artificial intelligence is going to give you data, but you’re going to share that data. You’re likely also going to transcend cultural borders. You need to understand how to work with different people. You need to, you know, from the analytical perspective, yes, you need to know how to look at the data. However, you also need to understand how to make the right decisions with that data. One interesting thing that I gained from that whitepaper is that data has been referred to as the new natural resource. An article from The Economist goes as far as saying, it’s going to replace oil as the world’s most valuable resource. How do you explain that, Ivan? That is a bold claim, if I’ve ever heard one, that it’s the new natural resource.

Ivan Portilla:

Yes. So none of these AI technologies works without data. So it is all based on the data that you use to train and to train these machine learning models. So no data, no intelligence, no predictions, no decision-making. So being able to control that data is so critical that a lot of companies are giving you services for that data. For instance, they give you an email for free, or they give you social media for free. So you can communicate with your peers and your family and friends, but because the data that you generate is so important to them to predict trends, to find needs in the marketplace, that that data is the differentiator. And this is an important point because when you sign up for a service via social media, movie, anything, remember that terms and conditions document, you are giving away your data to this company.

And you’re deciding what companies will be successful in the marketplace. So if you give it to company A, company B will not get your data, and it will be no deprecating. They will not be able to compete with a larger company. So you are decided each day when you use those social media recommendations, even clicks on our website, you are deciding what companies will succeed in the marketplace by giving them your data. The second thing is that data has a lot of implicit bias, and I believe that can be a topic around follow-up conversation. Then when we look at what is causing all these predictions to go towards prejudice or go to racial profile, and it goes to discrimination, and it all goes back to this data was used to train those models. So we need to be very transparent on the data that we use, the machine learning algorithms that we use to train the data to justify a decision. So that’s important. And the third thing I want to emphasize is that as we solve with the economies of scale. With the data, you will have less and less competition. So you will have one social media, one email company, one search engine because they have a monopoly on the data. So that’s a good point to consider.

How Can Small and Medium Businesses Personalize Learning?

Jenn DeWall:

I think even the awareness that data has become this form of currency. Every company wants it because they can use it to identify and predict your behaviors to help them understand how to offer a service. Change something else that they’re doing—but understanding that they’re always harvesting that data and how it may be used. So we do need to pay attention to how we’re using it. But I also love that you brought up that bias because I think we may put bias into our artificial intelligence and be unaware. And I want to talk about that. And then I want to go to a question that we got, but how do you, we’re actually, I’m going to jump into the question because I think that this will be helpful. So we got the question that came in, which asked, “ what should small and medium businesses in the US focus on to ‘make it personal’ or practice transparency and reach out? Other than working with our organization, like what can small and medium businesses in the US, focus on to make it more personal with their organizations?”

Ivan Portilla:

Yes. And I believe that this has to do with the three areas presented on the slides to make it personal, be transparent, and reach out. So this study came from serving a lot of organizations, some small, some large, but to make it personal, don’t assume that every role in your small organization has the same learning needs. You need to tailor that learning based on people’s skills, regions, geographical location, also opportunities that you will see in the marketplace. So when we helped this large chemical company redesign their learning materials, or in roadmaps, we went through all these different inputs, and we use AI to gather each personal station. We use content filtering, collaborative filtering, and we’re able to use the data that they have in-house and data from similar companies. We’re able to combine that data to have a more personalized experience. Now, being transparent in a small or medium company, I think is easier than in large companies.

They care less about business politics. The CEO can sometimes be seen in shipping or might be helping in the kitchen or whatever. So they are more reachable. There are fewer barriers to talk to your leadership team, and these leaders need to be transparent. They say, look, our industry is in crisis, for instance, right now in the pandemic. But we think we will, and we can go to these other areas. We can be. We need to be transparent. If we do this in other areas, how can your skills be retrained to help us be successful in that area? And then we can look inside and out. There are tons of, of like open-source organizations, chambers of commerce, local meetups that you can tap to gain those skills in that direction that you lack as a small and medium business.

How are Organizations Embracing AI in the Workplace Today?

Jenn DeWall:

Where do you start, if you’re thinking about not only up-skilling your workforce, but maybe even just incorporating artificial intelligence into your organization, what are some places that you see organizations starting to embrace AI?

Ivan Portilla:

I will say that the start is a set of data. Where do you think you are today? Send surveys within your organization and gather that data. And compare those results to some of the areas that we mention in the study. Like what are your soft skills? What do you think is critical for your organization, what are some of the technologies that will impact your business, and then reach out. Reach out to these local organizations, these meetups, these local learning colleges, and high school, like these tech programs. Contact the leaders and say, Hey, I have this baseline. I want to be here. How can I close that gap?

What about Bias in AI?

Jenn DeWall:

Yeah, absolutely. Get out there. Survey people and figure out what you could be missing. So I want to go back to that question about bias, because one example, I talked to another thought leader within the AI space, and her name is Sarah Alt. She spoke about what inspired her to get into AI. And it was when she noticed a problem when the organization that she worked for leveraged an applicant tracking system tool. So the story that she told is this. She was looking to fill a position. HR had given her a stack of resumes. She looked through them and decided they weren’t necessarily a good fit. And so she went back to HR and asked if she could get or repost that job. And then within the hour, HR came back with a stack of resumes, and she said, where did these come from? And HR responded, “well, they came from the ones that were automatically rejected from our applicant tracking system.” And when she asked the question, well, what did you use to determine to kick these out? They didn’t know. And I feel like that’s where we could have a bias or just that our own, I guess, lack of understanding of how AI works can really hurt us or just push us further away. Can you talk about how that bias can show up how they can minimize that bias?

Ivan Portilla:

Yes. And before I answer that particular question, let me give you a little bit of context. Okay. Within artificial intelligence, which is intelligence demonstrated by machines, there’s a field called machine learning. And machine learning, as I mentioned, is using data to train those statistical models and be able to come up with our recommendation: either classification or a number, a prediction. So within machine learning, there’s another subfield. So I think this of this like the babushka, stacking Russian dolls. AI machine learning, and then deep learning and deep learning, is all in favor nowadays. It has achieved significant results in certain areas, such as visual recognition and speech detection. And it’s kind of modeled after how the brain works, that you have these input layer and this output layer. And in between, you have this series of hidden layers that does the deep learning; unfortunately, with deep learning, most of the learning is a black box.

So you have inputs. These are the features, the characteristics that you want to train your model. And this is the prediction, either a classification? A true or false or span or no span? Or a prediction? What is the price of my house? Or do I need to hire this person? Or whatever. So with deep learning where the suffering of bias because most of the algorithm is a black box, and it’s not being exposed to the users. So not only on hiring but also on the legal system, certain algorithms are recommending more jail time for people of certain races. Or even in schools that are denying school applicants to enroll in a particular school based on their race or socioeconomic capabilities. Or even in my job, I have to use a similar tool to determine the pay raises and job evaluations. So if we don’t understand the algorithms, then we are very liable to bias.

So how do we avoid bias there? So we need to be very transparent of what data they use. What algorithms are used to train those models, and what are the expected results. But the data that is being used to train these models can have some implicit bias. And I can give you a good example. For instance, in my class, I say, okay, everyone draw a shoe. So everyone in the presentation here draws a shoe. About 80 to 90% of the people will draw a male shoe like a business shoe. And very few people will draw a tennis shoe or a high-heel shoe. That’s already biased in the input side, implicit bias. And you didn’t think about it, but it’s already present in your data. So if you use that data and I, and I use stat models to show up like a sandal, you will give me a wrong answer because the data that was used to train the model is already biased, and then you have other types of bias.

So one of the efforts that we have in IBM is to open up all these technologies, and we make a couple of open sources. So we have several resources that I will share with you later. And you can share with the attendees where you can see what the variables that are causing my model to reject or approve that loan is? Or what is the impact of real data has on the accuracy of the model three months from now. Because when I train my model, I have this data, but when I put it into production at the end of that account for the pandemic, I didn’t account for all these external forces. So you can retrain that model. So I will share those resources with you.

What is AQ and Why Is It Important?

Jenn DeWall:

That is so interesting. You know, that that’s a very, I think, easy example to understand that if the programmer is thinking of a male business shoe if you will, that it will directly affect the AI because it’s not accounting for the different shoes. And that’s a very easy way to understand that we might see things differently. And based on how we see things, if we’re the one that is the originator, you know, based on how we see it, we’re not incorporating a diversity of thought, or making sure that other people are in that conversation, that’s when bad decisions can happen. Or, you know, even talking about the bias, knowing that there can be bias built into these algorithms that can make consequences worse for someone else, not offer the same opportunities to the next person. And these are really serious things, and leaders need to understand. And I think this is where going back to that initial quote; emotional intelligence is so important to artificial intelligence. You need to be able to think about it from a broader perspective. I remember in the white paper, they talked about something called AQ, which was what does AQ mean again? And I do have another question that I’m going to get to

Ivan Portilla:

It is the ability to adapt to change that Adaptability Quotient. So you have AQ and EQ. Now we think another one important quality that is important is how well do you react to change?

How will AI Affect Leadership Development and Education?

Jenn DeWall:

Yeah. When we’re thinking about hiring people, or we’re thinking about who can be, who has the right skill set, one of the foundational things is, are they going to be able to handle change? Maybe it’s looking at their organization right now and saying, how well did everyone handle this dramatic shift into a remote space or adjustment to our strategic initiatives? So question what, what changes happen to the training and coaching industry due to AI robotics and machine learning? Because I know that those are typically ones where you might see them and sorry if I’m getting this wrong, but they are soft-skill driven. They are generally live. How do you see AI impacting that space? How do you see what changes will happen in that industry of coaching and training due to AI, robotics, and machine learning?

Ivan Portilla:

Yes, those are very good questions. Now, one area that we have seen in professional training is how we can better target learning to a particular individual’s needs. So like I was saying, not all training fits all. Now we know who’s the smarter person in the class, who is the person that needs additional time to complete their assignment. Who is the person I call it, tourist students that just come to the class to mingle with other students or to be away from the office? So we now have more visibility because we can now track individual learning paths. So that’s one impact that we’ve seen. And the other one is what I call the unknown unknowns, things that I don’t know, I don’t know. AI will expose those areas. It will say this is an area that you don’t know that you don’t know, but it will be a good area based on your experience, or the market needs to go into. So those are two places that I see an impact.

Jenn DeWall:

How do you see it? I want to take that one step further because in talking about the new collar. So there’s clearly a disruption to the education system that will happen as a result of AI. I know, I think, you know, Google last week just really announced its formal certificate program where you can just go and get a certificate with their hopes that that certificate will then be the thing that you need. And that employers will recognize that as you know, a sign of completion, a symbol of that skillset, and that’s going to save, I know for me, that’s going to, if I were going back to college today and had the option to do certificates, that would save so much in student loan debt and so much time, but how do you see it really disrupting the way that the education system works?

Ivan Portilla:

It is, it is huge. And like I mentioned, in our P-TECH program, imagine graduating from high school, going to an associate college for two years, and getting an associate degree that makes you employable right away without going through that four-year college. So that’s a huge impact. So we see these pockets of specialized skills and this market gap. So we can have a win-win situation. Shorten the education learning to get you right into the working environment, out of high school. And I was going to mention another area that we’ve seen success is these interactions. Sometimes we have created these kinds of like a chatbot or a self-service tool that will help you, the resources that you need quickly to complete the training. So that will free up the time of the educator to be able to focus on high-value activities, such as one on one conversations or, or, you know, exploring opportunities for growth. So that’s one area.

How Can AI Improve Online and On-Demand Training?

Jenn DeWall:

Yeah. That’s, we’ve got a few more questions that are coming in. How do you see online or on-demand training changing with AI? I mean, I know if you’re, you know, what even incorporating the chatbox feature, how do you see it changing?

Ivan Portilla:

We, we, we will see becoming more engaging, having these AI tools talk with you. We use them, so it’s not like, okay, go and look, what’s this video in and answer these five questions. These AI tools will be able to have a natural language conversation with you. And poke and find the weak areas and then provide an alternative learning path that will help you close those gaps in your knowledge, as well as exposing new things, new technologies that you were not aware of that can have a potential impact on your learning.

Jenn DeWall:

Ivan, do you think— and maybe this is my scared curiosity— but do you think that training with AI could also do something as simple as I’m in a conversation it’s recorded and observed, and then it will tell me what I did well, what I didn’t do well, and then give feedback to me? Do you see that as being something?

Ivan Portilla:

Yes. However, as we mentioned earlier, watch out for that bias. Always have those critical thinking skills and challenge whatever feedback you get from this automated intelligence.

What are the Ethical Implications of AI?

Jenn DeWall:

Well, that is, that is very mind-blowing to me to think that, you know, real, that real-time feedback with AI of how that can be used, how you even go about, you know, harvesting that data. That’s just so interesting to see that we can have that. It is also really great if a system could understand what these skills look like when done well. Assuming there’s not bias. Getting feedback from that system versus input from someone that doesn’t understand what that’s supposed to look like, but has a brief understanding and then provides feedback. So then you might learn bad behaviors as a result of that. That’s again, obviously the question of bias, and then the question of ethics comes into play, and we’re going to be talking about ethics, too, in tomorrow’s webinar.

But ethics, how or what role do you see ethics playing in artificial intelligence? Because I know we talked about soft skills, and there is a question about the skills that leaders need to develop themselves in their managers. What skills are most relevant for AI when you think about ethics and these new challenges that are going to come in, and let’s go back to that first question, what skills do we need to make sure that we have as leaders and that we’re building in our team right now?

Ivan Portilla:

Yes. And this is another critical area on AI because we have all of these points of conversation. We think of things in the future, right? Something that we have seen in movies, you know, we, we talk about the Terminator and all the bad AI. But in fact, today, we are programming ethics into our AI algorithms. Let me give you an example. We work with a large car manufacturer in Europe. We have to program into the self-driving car how to decide when somebody’s running in front of you to either hit the person or make a right turn into an abyss and kill the driver. Or make a left turn and collide with an upcoming car and killing all the people in that car. We have to make that decision. And it was a very tough, ethical decision to make. So this is happening today. Imagine when we have AI-enabled soldiers or drones, dropping bombs in areas based on the pictures you share in social media. So it’s a very important area that a lot of companies are working on it. There’s an organization called Open AI that has funded all these efforts. And these will take us another hour to go. So know the implications of ethics on AI.

Jenn DeWall:

Yes. Well, and even thinking about that, but yes, a self-driving vehicle will have a range of choices that someone will have to choose. What choice is the best, given the situation. And that’s difficult. I don’t know if I’d want to be the person making that decision. That’s pretty heavy. And I love this— Derek commented, “Going back to that question about how AI can give you feedback, not even about how you answer, but think about the expression on your face, your body movements and all of that.” Maybe they’ll say, Jenn, you use way too many hand gestures or you need to smile more because people think that you’re upset. That’s just so interesting. It’s, you know, Ivan, I know that we’re wrapping up, and you did have some resources to share. So if you want to go ahead and share your screen again, you can talk through some of those resources and then I’ll wrap it up. Or do you want to talk about those resources? Whatever works for you, but I know you’ve done your due diligence to help everyone further their learning by providing these resources.

Episode 39: Leadership in the Age of AI with IT Architect and Data Scientist, Ivan Portilla

Leadership in the Age of AI with IT Architect and Data Scientist, Ivan Portilla

Full Transcript Below

Meet Ivan Portilla, a Data Scientist Pushing AI to the Edge

The Importance of Skills and Talent in the Age of AI

The Need for Soft Skills in the Age of AI

The Need for an Agile Mindset

The Half-Life of Learned Skills is Shrinking

How will Intelligent Automation Change the Workforce?

The Positive Impact of Intelligent Automation

How Can We Close the AI Skills Gap?

Personalize Your Training and Reskilling

Live Q & A with Ivan Portilla

What Skills are Essential for Leaders in the Age of AI?

Will Data Replace Oil as the Most Important Natural Resource?

How Can Small and Medium Businesses Personalize Learning?

How are Organizations Embracing AI in the Workplace Today?

What about Bias in AI?

What is AQ and Why Is It Important?

How will AI Affect Leadership Development and Education?

How Can AI Improve Online and On-Demand Training?

What are the Ethical Implications of AI?

Recommended Reading to Learn More about AI

Interested in a free Leadership Skills Workshop with your team?

Categories

Latest Posts

Stay Updated

Browse by Topic