In this episode of TechSperience, Penny Conway has a frank discussion with Jamal Khan, President of Global Digital and eCommerce Marketing for Connection, on AI and how machine learning is infiltrating our everyday lives.
- Fact vs. Fiction and the data explosion.
- Democratization of toolsets: Tremendous advantages for developers to streamline process and development time.
- Microsoft AZURE ML Studio is an amazing working environment for developers. What was historically a difficult undertaking now saves hours and produces more efficient results and reporting. The studio leverages AI technology and simplifies the process.
- Learn the pace at which the AI and machine learning are growing and how it will impact the job market.
Listen to more TechSperience podcasts.
This is a transcription of the TechSperience podcast – Episode 9
Announcer:
Welcome to Connection’s TechSperience podcast, bringing you conversations with employees, partners and customers discussing the latest trends and innovations impacting the modern workforce. Today, AI – Artificial Intelligence. What is fact, what is fiction?
Our host, Penny Conway, Senior Program Manager for Workplace Transformation, sat down recently with Jamal Khan, the President of Connections Global Serve Division, which provides customers global access to IT products and services. E-commerce and marketing also fall under Jamal’s leadership, and it’s his background in software development that has come full circle now as he leads the company’s AI efforts.
Jamal Khan:
My background, my, sort of, education in tradecraft is actually a software developer. And that’s how I started my career, building trading environments on Wall Street.
Penny Conway:
Interesting.
Jamal Khan:
Um, yes. So, um, so that’s gonna age myself- that’s when the internet had just come out, so there’s this, like, 1996, 1997 time frame. And everyone was sort of scrambling to figure out, you know, how to leverage this medium called the web. And if you imagine those days or remember those days in my case.
Uh, so the models were very basic. It was all about, you know, the number of eyeballs that you can get to a site that would equate to banner ads, and that was the revenue model. And there were some companies- in my case, it was Instanet, one of the largest liquidity pools for equity training. And, and they were looking at leveraging the web as a means of migrating some of their proprietary trading platforms.
Penny Conway:
Mm-hmm (affirmative).
Jamal Khan:
And so those were sort of the projects that I started off. And then sort of, one of the instant challenge that we, sort of, had to contend with was cybersecurity, how do you secure those transactions on an anonymous, uh, session less, um, you know, platform? And so that’s where I got involved with a company called Verisign, which is a small company in those days.
Penny Conway:
Oh yes, Verisign!
Jamal Khan:
Yes, yes. And, and then work with them for about five years. And then sort of moved down after leaving Verisign on my own path, and sort of, sort of investing and building companies and Global Serve was one of the companies that I worked with and for. Was their Chief Executive for eight years prior to it being acquired by Connection in late 2016. And now you guys have me.
Penny Conway:
Fun. And now you’re here, on our podcast.
Jamal Khan:
I know. That’s your luck. (laughs)
Penny Conway:
(laughs) All roads led to here, we’re really-
Jamal Khan:
Yes, to the podcast. This was all converging to the podcast.
Penny Conway:
Right, right. Now, we’re so excited to have you here. And I, Artificial Intelligence is obviously not a subject that I am proficient in, but I think all of us have some sort of interest or opinion around Artificial Intelligence and what it is, future, you know, future forward, but also what it is today.
And you mentioned something interesting, you know, you talked about how you’re doing a lot of marketing and the banner ads, and when the web first came out. And that’s kind of how I look at Artificial Intelligence, as really this phone in my hand being an extension of me, and marketers being able to use my data and find out things about me, and learn about me, and things like that.
So why don’t you tell us a little bit about, you know, what the position of Artificial Intelligence is, you know, for us as people? Um, and kind of what it looks like as a, I think, more of a, you know, more of a scientific study or a field of study?
Related: Demystifying Artificial Intelligence
Jamal Khan:
Uh-huh. Right. Sure. So I mean, that’s a really broad question, because I can now go on a monologue for the next hour. Um.
Penny Conway:
We’ll break this up into a series.
Jamal Khan:
All right, so you’ve gotta help me break this down. So I- I think from a- a consumer, sort of, experience, I think in a lot of ways we’re experiencing AI all the time. And it’s become so subtly embedded in some of the day in, day out functions that we are involved in. You know, simple things like your Alexa machine.
Penny Conway:
Mm-hmm (affirmative).
Jamal Khan:
It- so, how it processes language when you’re sending it, sort of vocal commands. Your Amazon purchases. You know, and then it sort of gives you certain recommendations, so that’s- those are recommendation engines, and they’re generally built, again, on, uh, on some level of machine learning.
And, and sort of your Netflix, movie choices, and then their recommendation engines that sort of give you certain documentaries that you have to watch or should watch.
So those are sort of all these subtle examples, that you know, we’re having to sort of work with and deal with on a day in, day out basis, and in a lot of ways it’s sort of ‘out of sight, out of mind’.
Penny Conway:
Right.
Jamal Khan:
We don’t even realize that behind a lot of that is, is some level of machine learning. And then obviously autonomous vehicles and where they come in. So if you’re a proud owner of a Tesla, you know, there’s, there’s a lot of machine learning and, and, and systems within that as well.
Jamal Khan:
So that’s sort of your day in, day out, sort of run of the mill AI, sort of systems being leveraged. And again I use the term AI very loosely, I always make this mistake, and I, you know, when I lecture at some universities on this topic, I often sort of go through a large amount of time explaining that AI, at the end of the day, is not applied technology. And I think a lot of people make that mistake.
Penny Conway:
Yeah, it’s a good point.
Jamal Khan:
Right. So when we talk about AI, we think, well, I’m applying AI. Well, you’re not applying AI because AI’s a field of study.
Penny Conway:
Right.
Jamal Khan:
AI includes within it a whole bunch of different subconstructs that you can apply. But on itself AI’s nothing but a field of study, something we’ve been studying for, for millennia, in some ways. Going as far back as the Greeks, you know, and- and their notion of autonomous systems and mechanical systems in the thirteenth century, and so on and so forth.
Jamal Khan:
So it’s, it’s almost a something that humankind have been thinking of for a whole, you know, as I said, for millennia. And then of late, I think since the fifties, it’s something that, that’s sort of really taken a life on its own.
Penny Conway:
Right.
Jamal Khan:
And that’s been the progression. Now, the underpinning behind AI, at the end of the day is, it’s just advanced algorithms. And in, since I’ve worked in AI systems and built a bunch of AI systems, I’ve, I don’t know at what point was there that seminal moment where, you know, it just converged into this, you know, that AI’s here. Well, AI’s always been here, you know.
Penny Conway:
Right, yeah.
Jamal Khan:
It’s been here in the form of, really simple language translators since the fifties. It’s been in the form of, you know, expert systems in the six, seventies and eighties, or the micro worlds in the seventies and eighties. So it’s been around forever. I think there’s been, there’s now sort of a convergence of certain things that are happening in the ecosystem, that sort of making AI more sort of relatable sort of technology-
Penny Conway:
Mm-hmm (affirmative).
Jamal Khan:
-or a field of study than it has been in the past. But you know, just to your question, coming back to your question. Fundamentally, on a more sort of scientific, sort of core construct level, it’s a lot of computational mathematics that is well-translated through algorithms. And those algorithms have now essentially become democratized and easier to use, and hence we have this sort of convergence where AI’s being applied across the board, or in a lot of different things.
Penny Conway:
Right. And that’s, I think, I think one of those first- fact versus fiction, because that, um, I think now that we see it in our personal lives, and it’s more- more visibility, more access to what’s going on in the field, we think it’s this revolutionary-
Jamal Khan:
Right.
Penny Conway:
-thing that has come into our lives that’s going to take over. It could be dangerous, all of those things, but it’s truly been an evolution of mathematics and science-
Jamal Khan:
Absolutely.
Penny Conway:
-and data, you know, all the data that companies have been collecting, people have been collecting for years-
Jamal Khan:
Right.
Penny Conway:
-now having actual a system to be able to process that data, output that data. But definitely not new.
Jamal Khan:
Yep. Yep.
Penny Conway:
It- you’ve- not even fifty years new.
Jamal Khan:
No, no, not at all. Um, but I think there’s the, the one sort of aspect that I do talk about in terms of ‘why’s AI becoming so pervasive?’ Is that there is truly- there is a slight difference in terms of where we are today, as opposed to in the past. And I often call that the convergence of- of- of the reasons why we’re sort of genning up this hype, or that we- we see this hype around AI.
Jamal Khan:
You know, one is clearly the explosion of data to your point, you know, as, as consumers we’re now constantly, you know, generating information about ourselves, not to mention systems, not to mention machines, not to mention autonomous machines, and so on and so forth. And now as we transition our sort of, sort of global IT or technology ecosystem into sensors, imagine the explosion of data that’s gonna come out of those sensors.
I mean, you’ve, you’re theoretically gonna have- not theoretically, literally gonna have trillions of small sensors all over the place, generating information. So there’s clearly going to be a need of managing and processing that information in some meaningful way. And as machines become smarter at the edge, the requirement’s gonna be how do you process information- large, copious amounts of data- at the edge? And that’s where machine learning comes in.
Um. The second sort of shift or- of late that’s happened, or change, is, is the way, you know, one clearly the more classic processing capability, we’ve got more powerful processors that are able to process information in, in, faster. I remember in 2003, 2004, 2005 when we were doing some rudimentary sort of AI projects, it would take us literally a day to come back, to see- to run our models.
Penny Conway:
Wow.
Jamal Khan:
I mean, even some models today do take a long amount of time. But these are really simple models. And it took us almost a day to come back, to find that our models had failed. So we had to re tweak our models and then run them again for a day, a day and a half.
Penny Conway:
Wow.
Jamal Khan:
So that- that is a very laborious process. But with processors getting far more powerful today it’s made a big difference in terms of our speed of sort of processing data. And then, sort of, within the processor world, the application of GPU, type of processors to the challenge of processing large amounts of data has also helped us quite a bit.
Penny Conway:
Right.
Jamal Khan:
And that’s sort of with the transition from, you know, as GPUs were historically used for gaming systems, you know, somewhere along the line, someone decided to use them for large data constructs. And that had again- I think that was one of the seminal moments where we have- now have the ability to sort of process information much quicker and faster.
And so you, you have that and then I think, one thing that excites me immensely is the democratization of toolsets and toolkits. I mean, for example, in, you know, the Microsoft Azure AML Studio. I mean, just a very simple democratization and how developers can come in, look at that environment, work with what would historically have been very complex undertakings. Because when we were working early on in 2001, 2002, 2003, we didn’t have sophisticated toolsets and toolkits that made all of the underlying handling of data, orchestration of data, running analytics on data… It, it was something we had to sort of write up from scratch-
Related: 4 Ways AI is Improving Cloud Computing
Penny Conway:
Right.
Jamal Khan:
-you know, in some language, um. And now you’ve got these amazing studios. And you know, I’ve leveraged a whole bunch of them, these env- environments, of what we would call sandbox environments, that let us as developers really play with this technology. And Microsoft Azure AML Studio is one of those really amazing systems that really simplifies that.
Jamal Khan:
So the democratization of those toolsets and toolkits again is something that’s driving the adoption or this convergence. And then, last but I suppose not least, sort of, the need to process. I mentioned this, as machines get smarter, the need to process information at the edge is, is really important. And so that also- so you’ve got these four or five sort of underpinnings, or sort of these underlying sort of sort of new changes, they’re all coming together at the same time. And that’s why adoption, utilization, consumption of this broad field called AI has become simpler. And that’s why it’s-
Penny Conway:
Okay.
Jamal Khan:
-it’s becoming more prevalent.
Penny Conway:
Right. And with the- the become- becoming simpler, and having those, those tools at your disposal to now start, you know, doing things quicker than in a day, you can actually run tests for sure in minutes, or hours for big data sets. What is the- I know there’s a lot of back and forth on the concern over the pace-
Jamal Khan:
Uh-huh.
Penny Conway:
-of AI moving forward. There’s a school of thought that AI could be very dangerous, there needs to be regulation, there needs to be rules around, you know, privacy, how we use it, how we use data, how we use AI. And then more of that scientific, like, we’re just moving faster because we’re getting those tools, and then-
Jamal Khan:
Right.
Penny Conway:
-the more, the further along we get with it, the better it’s going to get, and it’s going to improve society. Where do you kind of sit on- between those sides of, you know, this could be really dangerous and detrimental to our society, to this could really bolster our society?
Jamal Khan:
Yes. So Penny, this is one of, probably the most difficult questions you’ve asked me. And the reason for that is that, if I answer this, I’ve enshrined my thinking, and you know how these things change and shift.
Penny Conway:
Right.
Jamal Khan:
But I- what you’re sort of describing is a classic sort of, Elon Musk and Mark Zuckerberg, sort of head to head on this particular issue. You know, so, you know, I’m not necessarily, and this is not a cop-out. So I’m not necessarily in one way or the other, sit in either of those camps, but if you were to ask me which side I’d lean more towards? It would probably be on the Elon Musk side. And that’s m- less sort of driven by the technology, sort of more so driven by human psychology.
Penny Conway:
Right.
Jamal Khan:
That, you know, we often take really powerful technologies and apply them for sometimes insidious purposes, and we’ve done that across our history, right? So, so I think AI will be applied for insidious things, and we’re- we’re seeing that today, in terms of how AI’s been leveraged to establish, you know, surveillance-based societies and, and what that means to privacy.
Jamal Khan:
And, and I think we’ve got to take those things very, very seriously, because those will deconstruct the way we think our societies are to operate. And they’ll happen very subtly, it’s almost sort of the frog in boiling water kind of thing, right?
Penny Conway:
Right.
Jamal Khan:
So, the- that sort of freedom that we sort of take for granted will really erode very slowly and quickly and out of sight and out of mind again. So there will be no discussion, there will be no debating. It will happen very quietly and often behind the scenes.
Penny Conway:
Right.
Jamal Khan:
And before you know it, what we take for granted is something that we don’t have, and we won’t even realize when that happens. So I’m, so I’m a little bit of a pessimist in terms of how we will eventually- there’s, there’s a huge amount of optimism that goes with what AI can bring, in terms of, societal improvement and sort of helping humanity on a whole bunch of different areas- especially healthcare, for example. But its application for certain other purposes, are- are a little disappointing and a little you know… scary, yeah, at times. So, I would sort of fall on, err more on the side of Elon Musk than Mark Zuckerberg, but, we’ll let- we’ll let time sort of flesh that out.
Penny Conway:
Right. I- I was listening to a couple of Elon Musk interviews in prep for this interview, and I- I have to say, like, he would get to a point in an interview where I was like… I gotta turn everything off, like… at one point he was saying that eventually we’ll be able to take all of the data that we- that I’ve created about myself and transfer it to another being, or body, should this one go away. And I was like, I gotta- I actually think that’s when I started to hear the clinking of the, the whiskey- the whiskey glass. (laughs)
Jamal Khan:
Yeah. So- so that, yeah. (laughs) So that’s the neuromorphic computer ring and yeah. So that’s quite interesting. And again, in terms of the arc of AI as a study, or as a field, its evolution- what has really surprised me is that when, whenever we’ve sort of thought about AI and its evolution, neuromorphic computing was something that was sort of the, you know, the end, one of these sort of, the far reaches or the far rungs of, where this field’s gonna go.
Jamal Khan:
And we’ve sort of- adv- advanced language translations, or two-way translators, and sort of advanced chatbots or surgical robots. We’re sort of, again, further downstream but more within hand.
Jamal Khan:
And then you know, the folks at, um, uh, with the- Elon- Elon’s company came out with the neuromorphic computing, where they built a robot that enables them to sort of, um, establish a synaptic access to, you know- or sort of access to synapse- synapses within, within the brain. Where they can you know, sort of and have provided some level of information to us in terms of what those can mean in terms of controlling computers and things like that.
Penny Conway:
Right.
Jamal Khan:
So we’re- so neuromorphic computing was something at the far rungs or the far edges of the evolution of this field. But it’s, it’s there, and you know. What I’m often sort of ex- interested about is, what is the R&D that’s going on right now, somewhere that we don’t even know of?
Penny Conway:
Right. That’s what I always think of.
Jamal Khan:
And what is gonna come out of, out of those efforts? And something we’re just not even aware of. That’s kind of exciting, but kind of scary as well. And so yeah.
So again, back- back to the question, lean more towards the Elon Musk side. I think there’s a lot of positive that can come out of this, but I’m sure we’ll apply this, as humans always do, to certain insidious purposes.
Penny Conway:
Right.
Jamal Khan:
And, and that’s not going to be good for us.
Penny Conway:
Yeah, we just can’t help ourselves all the time.
Jamal Khan:
We can’t help ourselves.
Penny Conway:
It’s like a human psychology piece.
Jamal Khan:
Right.
Penny Conway:
You had mentioned, you know Microsoft and their Azure product, that’s really helping, you know, in the field, and those toolkits and toolsets. What are, what’s the outlook, what’s going on, who are the major players in AI today? We- you said, you know, there’s R&D probably going on behind the scenes-
Jamal Khan:
Right.
Penny Conway:
-with I’m sure many of the players out there.
Jamal Khan:
Yep.
Penny Conway:
Um, but what do you sort of see stand out amongst the big guys?
Jamal Khan:
Um. So, you know, when I look at sort of, from an AI, sort of, vendor, partner landscape, you know, I s- try and break that down into sort of multiple layers, of ‘who are the guys who are- or gals- who are strong in sort of the hardware layer?’ You know, they’re building really great you know, silicon infrastructure for us to sort of be able to process.
Jamal Khan:
And I think that’s where- and- and, you know, I usually give a talk. I always say, you know, I wish I could come up with a different cast of characters, it seems to always be the same cast of characters. So whether it’s- is Intel or Nvidia or Micron. You know, they’re, they’re sort of helping build out, sort of, that core infrastructure, uh, upon which, um, everything rides. And so those are still sort of the primary players within this space.
Penny Conway:
Yeah. Yep.
Jamal Khan:
And then you have, from a sort of, from a software infrastructure perspective. You know, you’ve got, again those, like I’ve mentioned, like you’ve got the Azure AML folks at Microsoft, you’ve got Google with their Tensorflow sandbox environments. So you’ve got, you know, those companies again. And IBM and others that are sort of your classic companies that give us the ability to you know, deliver around the software layer.
And then you’ve got a whole broad set of ecosystems that’s somewhat diverse, with respect to the toolsets and toolkits. And so that also sort of helps within this process.
Jamal Khan:
And then you’ve got the large data companies. And there’s- then the question is, where will AI innovation come from? And that’s where companies like Google again, and Facebook, and Apple and others that have access to large datasets. So you know, whether it’s Apple through its iPhones and other sort of systems, whether it’s Netflix in terms of behavioral metrics around-
Penny Conway:
Mm-hmm (affirmative).
Jamal Khan:
-entertainment consumption, whether it’s Facebook with all of that dataset that is has through its applications, or any other company that has access to these large datasets. They generally have that opportunity to build really sophisticated underpinnings around machine learning and around AI.
So it’s having that opportunity, it’s having the software toolsets and toolsets and toolkits, and it’s the infrastructure that helps ride all this stuff. So you’ve got, you know, different leading players, and these different, sort of, rungs within the overall ecosystem.
Penny Conway:
Mm-hmm (affirmative). So, like, that kind of talks about the- the big, the big picture of AI, the Zuckerbergs, the Elon Musks. But what about, you know, if we were to bring- take it down just a step, and what AI and those applications. How can, how are businesses starting to kind of apply that for, you know, ROI as part of how they do business? Um, cause I think that’s probably more of the relatable thing we talk about, machines and, you know, robots building- I think there’s a YouTube video out there of a robot, like, that can do drywall.
Jamal Khan:
(laughs)
Penny Conway:
And everyone’s, like, ‘oh my God! The construction industry is going to fall!’ You know, that’s all of that sky is falling sort of things. But it- there’s true application for it in everyday business.
Jamal Khan:
So I- I think, um, so that example that you gave with the drywall, um… I wouldn’t be quick to sort of dismiss the potential disruption that is right around the corner, with respect to the workforce.
Penny Conway:
Okay.
Jamal Khan:
And- and I think there are lots of interesting studies, whether it’s from Brookings Institution or from, uh, um, McKinsey and Co., um, that speak to how AI-specific automation is going to be super disruptive to our societies in the next fifteen to twenty years. And there’s a broad spectrum of numbers that they give. Uh, I think, I think at a basic level, I may be wrong on these numbers, but I think global workforce is around, um, 2.1, 2.2 billion folks that are generally employed globally. At the conservative level, we’re looking at a 300,000,000 job disruption in the next fifteen to twenty years.
Penny Conway:
Interesting.
Jamal Khan:
Yeah.
Penny Conway:
Wow.
Jamal Khan:
And it can be as high as 800,000,000. Uh, and- and so there’s social scientists and- and AI practitioners in general have a view that we’re looking at disruptions at the level that we had in the industrial revolution period. With the Luddites and if you can imagine sort of the textile industry being automated-
Penny Conway:
Yeah.
Jamal Khan:
-and, and what that, you know, that had a significant impact. Or the farming industry being automated. We had at one point, I think, within the US, almost 90% of our population was in some way shape or form, associated with the agricultural industry. Uh, well, today it’s about 2%, or less than 2%.
Penny Conway:
Right.
Jamal Khan:
And- and so yes, we’ve sort of transitioned over a period of time, and we have- we put in certain mass programs, like primary education was not something that we had, prior to the agricultural shift. And we had to bring in primary education to give our workforce the skillsets to be able to transition.
Penny Conway:
Right.
Jamal Khan:
And that period of transition was much longer, and so the societal impact, though significant we had significant challenges, in sort of the farming automation, and then the industrial revolution period. But we had a long period of time to sort of adjust ourselves and our society, and how we trained our workforce to make them capable of delivering our new types of services. That might be different this time around.
Penny Conway:
Because it’s moving way faster. Yeah.
Jamal Khan:
The pace of change is much faster.
Penny Conway:
Yep.
Jamal Khan:
The need to transition the workforce, so, and I’m, and I’m going into the weeds on this, but I think it’s a really important topic, which is, you know… You’ve got low-skilled jobs, you’ve got what we call mid-skilled jobs, and you’ve got highly skilled jobs. Let’s say if we look at the workforce and we divide those into those three categories.
Jamal Khan:
So the low-skilled jobs, though easy to automate, don’t necessarily have a financial upside for businesses to automate, because you can always find low-cost workforce and just use them and leverage them for that.
Penny Conway:
Yep.
Jamal Khan:
And they usually provide service, uh, that the mid-skilled folks are leveraging or utilizing, whether it’s food services or other areas, it’s the mid-skilled folks who are consuming those low-skilled services. The low-skill, the mid-skilled jobs are the, what require some basic level of training. And are repeatable in some cases. Those are the jobs that are most likely to be automated out. So whether it’s a lot of information processes, or where it’s paralegal work, whether it’s customer service. Whether in transportation it’s actually- even transportation, driving…
Penny Conway:
Right.
Jamal Khan:
Those are the ones that are likely to be disrupted. Now the challenge for us as social scientists would be, how do we get our society to be prepared for those changes? And by the way, the- the period of change will be compressed, it’s not gonna be over a hundred years. It’s gonna be within fifteen, twenty years.
And so usually you either move that workforce to a low-skilled job, or you move that workforce to a high-skilled job. High skills in this case are gonna be very difficult to sort of train someone who’s sort of in a mid-skilled. These skills are very sophisticated-
Penny Conway:
Right.
Jamal Khan:
-and very difficult to sort of get folks to make that transition. And then the other challenge is, because the- as you’re sort of attenuating the mid-skilled jobs, and they are sort of the the consumers of low-skilled services, the likelihood is your low-skilled jobs are not gonna grow either. So you’re gonna have essentially a workforce for a good- and that’s, you know, maybe one or two generations of folks who are gonna have a very hard time transitioning into one bracket or the other.
So that disruption is what I think a lot of social scientists talk about is going to be something that a lot of western democracies are going to have to contend with. And now Brookings has come with a proposition that that’s going to impact how we govern. So we- we take for granted that we’ve got these, um, you know, democratized societies with civil liberties. And we take it for granted.
Penny Conway:
Right.
Jamal Khan:
But there’s no guarantee that that’s gonna continue to be around, as the societies go through these convulsions. And so that’s something that really concerns me, in terms of AI and its impact on society, but that’s a completely different topic.
Penny Conway:
No, it’s super interesting, because I, as the everyday person that’s out there that- I think the perception is, and that just- my- and that’s the beauty of a podcast, it’s, it can be completely our opinion, our perception. Um, is that it’s the low-skilled being replaced.
Jamal Khan:
Right.
Penny Conway:
It’s the, you know, the cashier with the self-checkouts, it’s, you know, the- the brick and mortar stores, because I now use Amazon to do everything. And so I think we as a, as a body of people, think ‘oh, we don’t need any more cashiers, we don’t need any more restaurant workers’, you know, that I’m fine ordering at Chili’s at my table and paying for it and not dealing with a human. But we don’t think about that mid-level, which is where the majority of us actually sit every single day.
Jamal Khan:
Absolutely.
Penny Conway:
So, what’s your, you know, hypothesis of how we are gonna tackle that as a society, with that large group of people now that have to go one way or the other?
Jamal Khan:
Now, the optimistic view is that, that AI or, sort of, this transition in general will create a whole bunch of new jobs. And that’s where a lot of the workforce will sit. But there’s some early indication that- and data’s showing this- that there is a decoupling of job creation and automation that has begun to happen. And it sort of started happening in the 2000s. So if you go back all the way to 1940s, you’ll- you’ll often see in sort of a parallel sort of growth of productivity enhancement with employment enhancement.
Penny Conway:
Mm-hmm (affirmative).
Jamal Khan:
And that productivity enhancement usually came from automation. So there was- there was a corollary, or you could make a correlation that as you were automating and improving your productivity, you were essentially creating other jobs. And you basically had the ability to sort of enable folks to transition into those jobs.
Jamal Khan:
Somewhere around 2000, that decoupling happened, where those two lines of growth stopped happening, where productivity enhancement continued to happen through automation, but, jobs weren’t being created. And again, so that’s another thing that is different from what was in the past.
Penny Conway:
Right.
Jamal Khan:
So I- I don’t have an answer for this. I mean, I know it’s a question I wish I could tell you there’s gonna be a whole bunch of new jobs that will get created, I don’t have an answer. And I think the reality is I, my gut sense is telling me, and my sense is telling me, that there’s going to be a significant amount of, of disruption. And, and how we contend with it, what we do, um, you know, whether we’ve got to sort of come up with, again, a- a mass project like the primary education project, right?
Jamal Khan:
So how we educated ourselves, but that primary education construct that we’re still consuming, or our kids are consuming today, which is- are the three R’s, right? Reading, writing and arithmetic.
Penny Conway:
Mm-hmm (affirmative).
Jamal Khan:
Those are the three sort of fundamental underpinnings of primary education. And that needs to be changed.
Penny Conway:
Right.
Jamal Khan:
And- and what would that program be? So someone with far greater imagination than myself within these fields, a social scientist, have to think about these problems, or challenges, and, and figure out what- what would that, that process be that enables our, our populations and our workforce to make a smoother transition. But, uh, there should be no illusion that this isn’t going to be significantly disruptive.
Penny Conway:
Right. And I- I think, I mean, we’ve seen in education, and I have a little bit of a background, um, in K-12 education. And we saw the rise up of the STEM- science, um, technology, engineering and math.
Jamal Khan:
Right.
Penny Conway:
And now it’s arts and you know.
Jamal Khan:
Steam. It’s- it’s steam.
Penny Conway:
At a- at another- at another level, stream, steam, STEM… um, but I think that that was kind of the, maybe the introduction?
Jamal Khan:
Perhaps.
Penny Conway:
And trying to prepare. And get, you know, kids thinking less about, you know, going for maybe a general studies or, you know, some sort of degree that doesn’t have a high level of, of technical skill to it? And trying to get kids to go that direction in preparation for what might be coming down the line in fifteen or twenty years. But agreed, people much more imaginative-
Jamal Khan:
Yes.
Penny Conway:
-than us working, with that education system to try to-
Jamal Khan:
Right.
Penny Conway:
-you know, stave off some of-
Jamal Khan:
Yep.
Penny Conway:
-the damaging effects of what might come to the workforce in the next couple of decades.
Jamal Khan:
Yeah. Absolutely. And I think, um, you know, for some of us who are thinking in terms of how to make our societies more competitive, I think there is a legitimate case to be made that we’re going to be in a state or stage- and perhaps we already are- where your national power is directly related to your ability to build these systems.
And- and I think a lot of these folks sometimes even, you know, sort of talk to these as sort of the war of the algorithms, right? So, how nation states are building more sophisticated algorithms that bring in productivity into their organi- into their sort of economies. And so, whether it’s China or Japan, or India, or other sort of large- or Germany, or other large sort of countries that have sophisticated, sort of, tech base. They’re all striving to build more sophisticated AI engines to sort of help them solve and optimize how they operate.
Jamal Khan:
And so it’s almost like this is a war of, um, um, you know, war of the algorithms. And- and where the US sort of eventually falls on this, I mean, I think we- we are in a good position to start off with, but we shouldn’t again take that for granted.
Penny Conway:
Right.
Jamal Khan:
There’s a lot of interesting work that’s coming out of other geographies, and sometimes we’re so geographically isolated that we, we sort of think that the world begins and ends here.
Penny Conway:
Right.
Jamal Khan:
It does not.
Penny Conway:
Right. (laughs)
Jamal Khan:
(laughs) I you know, I’ve traveled the world, I have seen places and I think in some way that maybe part of our curriculum, that STEM curriculum downstream is like, people need to travel.
Penny Conway:
To know- and know what’s going on.
Jamal Khan:
What’s going on around the world. So that does not make us complacent.
Penny Conway:
Right.
Jamal Khan:
And I think that’s one thing that’s a challenge and, and people, you know, just sort of really think that through. Yeah.
Penny Conway:
Yeah. It definitely think that sort of belief that it starts and ends here can sometimes be what blindsides us-
Jamal Khan:
Right, right.
Penny Conway:
-when things come out of nowhere. So, I’ve- like we’ve said, this is a hugely, or huge and broad topic. What I would love to do is have you back to kind of expand on more topics around Artificial Intelligence, specifically how businesses are using it. Because I think that that is really interesting when we go back to sort of that introduction of the web and those banner ads. And now the data that’s being collected to kind of hit consumers and market brands and have companies use data that they’re collecting on a daily basis to actually increase their ROI across the board.
Jamal Khan:
Uh-huh.
Penny Conway:
Give us a little bit of a preview of sort of that topic and now companies are starting to use AI to produce some ROI.
Jamal Khan:
Sure, absolutely. And, so you know, whether it’s Dell Co. leveraging, you know, AI in their service operations, whether it is, you know, automotive, sort of, leveraging sort of AI or robotics or automation in manufacturing. Whether it’s financial industry looking at AI for risk modeling and- or retail for behavioral metrics or analytics or predictive or prescriptive analytics. There’s- the application is really broad.
And- and I, I would be remiss if I didn’t mention that I cannot ima- remember the number of times I would speak to Tim, my boss, and he would always, like, ‘Jamal, what are we doing on AI?’ And so I can almost imagine there are a whole bunch of CEOs out there who are sort of going after their- whether it’s their CIOs or their key, sort of, innovation officers within their companies and saying, ‘what are you doing around AI?’ Because there is that hype.
Penny Conway:
Right.
Jamal Khan:
And it’s almost like, you know, everyone’s trying to figure out. And yes, it- from an enterprise perspective, what they have is their data. And- and I think there- there is sort of a cliché that or sort of, you know, the proverbial sense or expression, which is, you know, data is, is the new currency. And I we really believe that, we- I think we also believe that inherent connection, that data is our currency.
So how you can get your hands around that data and use that to inform you, to use that to provide you insights, to improve your productivity, and perhaps look at opportunities you haven’t even considered. So would love to come back, talk to you in terms of that particular space.
And then there are areas that we haven’t touched, which are sort of the core underpinnings of AI technology, and what is it? You know, what is machine learning? And how- what’s the evolution within machine learning, what are the different neural networks? What how do particular neural networks work for particular tasks?
And then just one area that I find incredibly compelling, and it’s sort of- it’s the surreal stuff in AI, which is, you know, generative adversarial networks. How you can build two networks that go against each other and sort of build images and things of that sort. And what’s coming out of those systems is just mind boggling.
Announcer:
Again, Connection’s Jamal Khan with his first of what we expect to be many conversations about advancements in Artificial Intelligence. Until next time, thanks for tuning in to the TechSperience podcast from Connection.
