EP 292: Is AI Transformative or Have We Not Fully Grasped it Yet?

This is the Everyday AI show, the everyday podcast where we simplify AI and bring its power to your fingertips. Listen daily for practical advice to boost your career, business, and everyday life.

When we talk about artificial intelligence, it's always people talking about, like, oh, it's, it's real, it's fake, it's, it's hype, right? But is it actually real? Like, have we realized what generative AI can do, or have we not even begun to understand what it's capable of? So this is something I always think about a lot. And, you know, I'm excited for today's conversation because that's what we're going to be tackling, whether AI is actually a transformative technology or if we haven't even really grasped it yet. So before we get started, just as a reminder, if you are listening on the podcast, thank you as always. If you're on the live stream, get your questions in. You know, we'd love to engage with you all and make sure if you are listening on the podcast, check out your show notes and go to your everydayai.com Sign up for our free daily newsletter where we will be recapping today's show and sharing a whole lot more. All right, so before we get into today's conversation, let's first start, as we always do, by going over what's happening in the world of AI news. So the US Is considering some AI restrictions of sorts on China. So the US is considering further restrictions on China's access to advanced chip technology used in AI, according to reports. So the potential restrictions would target a new transistor architecture known as gate all around, which can improve chip performance and lower power consumption. There were previous export controls on AI chips to China that have been tightened, and the potential GAA restrict are still being determined. So it is something to keep an eye on, you know, AI and just the GPU chips that kind of power AI globally are being treated as a resource. Right? And, you know, the US Government is really cracking down on exports there. All right, our next piece of AI news. OpenAI has dodged a lawsuit, but one that wasn't very serious, if we're being honest. So Elon Musk has dropped his lawsuit against OpenAI, the artificial intelligence startup he co founded in 2015. So the lawsuit alleged the company had abandoned its original nonprofit mission and reserved some of its advanced AI technology for private clients. So Musk's decision to drop the lawsuit comes after a series of critical posts he made on social media against OpenAI after Apple announced their partnership. Musk's lawsuit against OpenAI has been dismissed, ending that months long legal battle between the two parties. So Musk has had accused OpenAI of pursuing profit instead of its original nonprofit mission. But the company claimed. But the company has denied those claims obviously. And we talked about on the show a couple months ago, we went over it point by point and how that this lawsuit wasn't very serious and there wasn't merit in it. Last but not least, large language model maker Mistral just made a big funding splash. So the French AI startup Mistral AI has raised a funding round of 600 million euros, valuing the company at 5.8 billion. They are competing with other large AI companies such as OpenAI to become Europe's AI champion. So Mistral has raised that significant amount of funding in a short period of time and is not only obviously competing with OpenAI, but Google, Gemini and Propit, Claude, etc. Competing on a global scale here. So the French startup is building a large language model and has attracted the attention of major investors including General, General Catalyst and Microsoft. So Mistral's co founder and CEO Arthur Mentz has become a notable figure in the European technology scene. All right, so there's a lot more happening in the world of AI if you didn't know. So make sure if you haven't already, go to your everydayai.com and sign up for the free daily newsletter for more on that. But we are here to talk about AI. Is it real? Is it hype? Have we even begun to realize what it's capable of? So it's not just me talking today. Let's go ahead and bring on our guests. There we go. I'm extremely excited to welcome to the show Jamal Khan, the head of Helix center for Applied AI and the Chief Growth and Innovation and Innovation Officer at Connection Incorporated. Jamal, thank you so much for joining the show.

Good morning, Jonah. It's good to be on the show.

Hey, it's great to have you especially man technical difficulties today. But regardless, Jamal, tell us a little bit about your role at Connection and what Connection does for those that maybe aren't aware.

Sure. Connection is a Fortune 1000 public company that's essentially a global solution provider. We have north of 36,000 customers primarily based out of the U.S. but some of those customers have a global footprint as well.

Recording in progress.

Connection essentially provides a whole broad set of services solutions to our clients and that's essentially what we do. In terms of my role as the head of the Helix center for Applied AI, we've been as a company going down this Journey now for almost five years, a journey that in large measure began almost as a data transformation journey, which was our own internal efforts on how we become more data oriented. And out of that process we built a capacity that we're now scaling to sort of help address the challenge that our customers are bringing forth for us, which is how do they navigate their own AI and data journey. And that's the Helix center for Applied AI. And there are a whole broad set of other areas within connection. The global business is still part of my mandate as well as the marketing organization. But the one part that really takes most of my time these days is this whole thing called artificial intelligence that, that's really pulling me in for a whole broad set of reasons.

Yeah, yeah. And Jamal, you just dipped, you know, just dipped your toe on, you know, what connection is and you know, what you all do, but as a public company that's been working in this space for a very long time and you know, as we think about artificial intelligence and you know, the hype cycle, is it real? Is it transformative? Maybe can you take us a little bit and even just in your own personal experience and in your career. Right, because AI is not new.

Right.

Technically, AI has been around for like 50 plus years. But as this generative AI kind of wave started to form, you know, three or four years ago, how do you think kind of the climate changed in terms of people now who maybe didn't have a use case for, you know, neural networks and deep learning, you know, traditional AI, you know, know, how has that just changed over the last couple of years with this resurgence or surgence of generative AI, I guess it's become tangible.

Right. It's in some ways democratize access to this notion or this vague notion that was very ethereal or very academic historically to think about artificial intelligence. It's sort of democr. Democratized access to everybody through a very conversational way of interacting with, you know, large language models. LLM, I remember, you know, I, you know, Jensen, the CEO of Nvidia, used to have or perhaps still does these fireside chats where he would bring a small, select, you know, small group of execs and, and sort of talk about what's happening in the ecosystem. And I used to consistently go, go to those Fireside chats and about three years ago, or two and a half, three years ago, used to say, guys, you've got to really focus on this thing called Gen AI. And we were like, what is Jensen saying? And obviously we didn't know Jensen was providing the first, you know, DGX1 hundreds to OpenAI. And he knew all, obviously, the insight, and he would say, guys, you've got to focus on this gen AI thing that's coming right around the corner. And he was spot on. And so I think the reason why we've seen this amplitude, and I don't want to use the term hype, is in large measure because it's democratized, it's become much easier now. It's no longer the domain of, you know, mathematicians, you know, toiling on deep neural networks and underlying algorithms. It's now you can, in a very simple way, through a very simple prompt, just ask LLMs. And it sort of tends to give you a cognitive response, and that's made it very easy for people to understand.

So let's, let's just go ahead and skip to the end here, Jamal, and let's just answer the question. So when we're talking about AI and, you know, for those that aren't, you know, familiar with, you know, technology, hype cycles, right? But it essentially says that, that, you know, all technology can essentially be plotted, you know, from when it comes out and people get super excited and, oh, you know, we've kind of reached the apex and now we're gonna, you know, go into this, you know, I guess, disillusionment or whatever it's called, you know, kind of on the downswing. Yeah, yeah. But you know, Jamal, like, is. Is AI, is it actually transformative technology that's gonna, you know, continue to be more and more impactful as we go on? Or can we just say, hey, it's just another dot on the hype cycle?

So it's a really, really good question, Jordan. I think it's a good question because it impacts and is likely to impact a whole broad set of considerations, from policy to investments to company strategies. Well, for someone who's worked within this space now for 20 years, I will not argue that it is not transformative, because if you've been at something for 20 years, obviously there's an implicit bias. And I've always believed in artificial intelligence, but I think it's a very difficult question to answer because artificial intelligence is not necessarily a ubiquitous technology, Right? It is very diverse. It encompasses elements such as computer vision, it encompasses elements like large language models. There are different approaches, there are different applications and use cases and new studies. So I think in sort of the arc of AI's diverse ecosystem, there's certain parts of artificial intelligence that are far further along and more tangible and perhaps moved beyond that trough of disillusionment into sort of providing some level of productivity as opposed to others. And so I think you're going to see this variance in how AI as a collective field of study evolves over a period of time. So I absolutely believe, I think it's transformative. I think there's certain parts of AI that perhaps are still in the early stages of that hype cycle, but are likely to go through their process of evolution. But there's a fundamental reason why we find that this AI boom is not a bust, where historically we've had these multiple booms and bust cycles in AI. And that's something to your point, started almost 50 plus years ago. But when you look at processing capability, you're looking at this explosion of data, you're looking at a democratization of tool sets and algorithms, you're looking at this explosion of an underlying need to become more efficient and bring in automation. All of those things drive in large measure how artificial intelligence is becoming more real and practical. So I think there's a reason behind why I believe this time around is not necessarily a bust cycle. But again, the key point here is it's a diverse ecosystem. We're not going to see everything evolve at the same time. You're going to see certain parts of AI evolve much faster than others.

Yeah, and Jamal, you bring up a great point there, you know, and you're even talking about your background, you know, been in this space for, you know, 20 plus years and you know, I love going back and you know, hearing this story of, you know, Jensen kind of telling everyone, you know, hey, pay attention to this generative AI thing. You know, my, my thought is because, you know, all the studies, all the research says that, you know, you know, especially public companies, Fortune 500 etc, they're all saying that generative AI is one of their highest priorities yet. I, I think some of the most recent studies say that only 4% have actually implemented it company wide. Why do you think there is this disconnect between everyone saying yes, we know it is a top priority versus hardly no one has implemented it from top to bottom. Why do you think that is?

I think there are a multitude of reasons. There's no one single reason. And we see that cycle of customer maturity, right? So you've got some customers that are on the upper end of maturity. They've been working in artificial intelligence now for 10 plus years. They made those investments very early on and it was in some measure either driven through some level of computer vision for quality control functions, or data analytics and driving More insights through data. But those organizations are far more mature in terms of where they are. And then on the other end of the spectrum, you have customers who are still scratching their head. They're asking us that basic question to the question that we're trying to address here as well. Is this real, is this hype? And then within my specific vertical, what are the use cases that can truly provide the roi? So you've got this broad spectrum of users in different levels of maturity. And then there's another interesting factor that we're seeing. We're seeing almost this. Let's develop our skill sets, let's develop our knowledge base, let's develop this understanding of how this can truly impact and apply us, and then let's get our budgets ready. We're seeing a lot of companies and in general trend that we're seeing where there's almost this oxygen that's being grabbed out of other projects and everything else is being slowed down because everyone's preparing for this massive spend that they expect in building out this infrastructure's ecosystem and acquiring the skill sets. There's this massive skill set challenge that exists as well in terms of resources that can help address that. So again, I think that the short answer is it's not a ubiquitous pathway to deployment. You've got different levels of maturity and then you've got customers on both spectrums of that maturity. You know, litigating the issue.

The, the phrase there, get your budgets ready, I think is, is an overlooked portion like, like piece of this, right? Because I think the, the narrative, and maybe it's just because things like this are easy to sensationalize, but seemingly this, this narrative is, oh, generative AI, it's going to save you so much time, so much money. And you see these studies from McKinsey saying, oh, up to 80% of knowledge workers tasks could be automated by generative AI. So people, I think maybe just look at generative AI as a cost savings, as a shortcut. But you said something very important there. You talked about budgets and skill set challenges. So Jamal, I'm curious, as you've, I'm sure talked to many people about generative AI implementation. Implementation. But you know, is that just a common misconception or are people just not wanting to fully invest into, to reskill and to upskill and to kind of relearn how they work with this skill set and budget kind of challenge to implementation?

No, I think the intent is there. We see that there's absolutely the intent to either upscale or sort of acquire those skills. But Jordan, to be honest with you, it's not an easy transition even within connection as we're going through the process of building a more robust organization. By the way, the skill set up skilling challenge is not just in acquiring resources that can deliver artificial intelligence solutions. It's also sort of upskilling even the users of AI. How do you actually upskill your knowledge workers to become more effective users on some of these systems? So I think that's going to take its natural cycle of time in developing that and that is a multi year journey of upskilling. So the intent is there, the desire is there and the effort is ongoing. The real skill shortage lies in those real critical resources. These are your engineering resources. These are your resources that can help you build your rag frameworks that can assist in your gen AI systems. These are resources that can help you arbitrate the different models you could leverage theoretically to fine tune that. That skill set is few and far between and, and, and that's I think where you have a skill set shortage. So hopefully that gives you an answer. I know it's, it's, it's a little bit of a vague proposition because I often find that there's no one single way to explain many of these, these issues that we talk about. There's a whole broad spectrum of different varying situations that we need.

Are you still running in circles trying to figure out how to actually grow.

Your business with AI?

Maybe your company has been tinkering with large language models for a year or more, but can't really get traction to find ROI on gen AI. Hey, this is Jordan Wilson, host of this very podcast. Companies like Adobe, Microsoft and Nvidia have partnered with us because they trust our expertise in educating the masses around generative AI to get ahead of and some of the most innovative companies in the country hire us to help with their AI strategy and to train hundreds of their employees on how to use Gen AI. So whether you're looking for ChatGPT training for thousands or just need help building your front end AI strategy, you can partner with us too.

Just like some of the biggest companies in the world do.

Go to your everydayai.com partner to get in contact with our team or you can just click on the partner section of our website will help you stop running in those AI circles and help get your team ahead and build a straight path to ROI on gen AI.

And you know, hey, as a reminder to our live audience joining us, thank you as always. But if you do have a question for Jamal, you know, specifically on Gen Gen AI implementation and your thoughts on if it's transformative or not, please get them in now. Jamal, one thing I always think about is this issue of trust, right? Both trusting the models, there's this black box of ambiguity of what are large language models, what is generative AI? How should companies address that trust issue and maybe just trust in the process from a change management perspective? How can companies face that?

So I think that is the seminal challenge with AI and I like to sort of unpack that just a little bit. You know, I'm going to age myself. I remember, you know, in the early days of the Web, you know, mid-90s, late-90s, you know, I was working for a startup in those days called Verisign. And for those of you who don't remember, Verisign essentially built root certificate authorities that enabled browsers to essentially undertake SSL transactions and also provide some level of bidirectional authentication. You could have an end user certificate and then you could have a server certificate, and then based upon that, you could establish trust in an HTTP transaction which is otherwise a stateless, sessionless protocol. I remember our CEO Stratton Sclavos in those days used to say, unless we solve the underlying trust issue that exists within the web, we will never get meaningful applications at scale on the web. And you know, as a, as a 20 something year old person, I remember going into Fortune 50 companies in their boardrooms and trying to teach them on public key infrastructure and why trust is a cornerstone of them, moving some of their applications, whether these were trading applications, financial applications, banking applications. In a very analogous way, when I talk to leadership in companies around their desire to build out artificial intelligence, we are encountering exactly the same challenge, which is, what is that trust framework upon which they can truly build these AI systems? Because AI has a unique set of challenges as it relates to trust. There's a transparency challenge, there's an explainability challenge, there's an accountability or a fairness or bias mitigation challenge. You don't want to be in the business of a bank that's built loan accreditation or loan issuance systems based upon AI algorithms that may have implicit bias around a protected category of people because you're opening yourself to downstream litigation at some point. You've got reliability and safety around these AI systems as they become more kinetic. You've got data and privacy concerns, you've got intellectual. So there's a whole broad set of these underlying issues that need to be litigated, that need to be addressed. You've got legislation that needs to catch up, you've got case law that needs to catch up. So I think there's this natural organic arc of time that with some of these areas becoming more mature, you're going to see AI solve that or the industry solve that underlying trust challenge. But I think that is a precursor if there's anything that's pulling us back or slowing this process. Because we're seeing buyer's remorse, we're seeing companies out there that have deployed personal assistance day one and by day three they're switching them off because all of a sudden they're not sure who's got access to what data. What's happening is data egressing. What's the data that I'm bringing in? Should my graphics team have access to finance data? So we're seeing all of that complexity bear itself out as companies truly try and implement practical projects around AI.

It's, it's fascinating, right? Normally when you think of buyer's remorse in a technology project, you know, you're, you're thinking a little further down the line, right? Maybe, maybe weeks or quarters after, you know, you've implemented something. So, Jamal, one thing, you know, I'm even personally curious about, right, so we, we talked about a couple of minutes ago how this process, it's, it's a multi year, right, Upskilling the investment, etc, and you even mention, even for connection, right, this is, it's a transition. So, you know, I'm curious even internally there at connection, you know, kind of when you look at this, this hype versus transformative technology, how did you all personally face this transition? And maybe what are some of the key takeaways or findings that, that you all kind of went through that you think may be helpful to share with others?

Sure. I remember vividly the conversation I had with my boss almost five years ago where I remember expressing to him, I said, Tim, if we can be a data and AI company that is looked upon by our customers as the company to help them in their journey for AI and data transformation, we will be in a good place. And that was a conversation we had five years ago. And I think that sort of spun off in his mind, in my mind, and then collectively the board support as well, which was, you've got to start looking in house first. How do you become the consumers of a democratized data plane? How do you enable your internal resources to drive greater insights? I know that sounds cliche, but it's a pretty significant challenge that required a complete retrofitting. And by the way, it's an ongoing journey, it never stops. There was almost this underlying cultural shift. And if you ask me, one of the biggest challenges that one can have with respect to artificial intelligence is what I would call change management. It's a mindset shift within organizations. That one, it's okay. It's okay to democratize data. It's okay to break those silos and have access and give access to everyone in some measure. And by the way, in that process, there may be some scar tissue as well, and there's nothing wrong with that. So I think that was a cultural mindset shift that needed to be constantly driven. And by the way, Jordan, that was not an easy undertaking. These are well entrenched silos. These are well entrenched cultural ethos that serve the company well. But you've got to try and go and shake that a little bit. And then I think there's a consistent commitment from upper management to sort of support that. And again, all the waste travels up to the board. And then the last thing is just the avoidance of trying to boil the ocean. Right? You don't want to sort of go into these projects saying, you know, AI or data transformation is suddenly going to radically shift the business. You've got to really identify small wins, and you've got a almost. It's a flywheel effect where these small wins give you that amplitude and acceleration that you need for eventual adherence to these, these transformative shifts. And then the last thing, which probably will irk a lot of people. I am fundamentally of the view that a square root applies here. That initial group of individuals that you want to really try and support your efforts in these journeys is usually a square root of the total that exists. So if you've got a company of 900 people, at the end of the day, there are only going to be 30 people in that company that can truly initially go on this journey with you. And with them, they're willing to sort of, you know, swing with the punches and try and really push the envelope and then be the best evangelist. But if you don't have that initial 30 core group in a company of 90, let's say again, applying that square root rule, you're going to have folks who are not really committed to this process. So I think those were the principles that we adopted. Consistent support from top down all the way up to the board, small wins, a square root principle, and then just the ability to say, it's fine to fail, it's fine to sort of get a little bit of scar tissue, and there's nothing wrong with that. I think that were some of the elements that we drove internally that I think gets us to where we are today.

You know, speaking of the, this square root principle and getting some of those early adopters that can champion generative AI throughout the organization. You know, a lot of it, what we said, it goes back to trust, right? So Cecilia's question here. Would love to get your, your thoughts on this. Jamal so she's asking, does the natural arc of time for the Trust Change Challenge include the complexity of communication paired with lack of time? Have you seen companies create good models for communication of the AI transition that are inclusive of all stakeholders and issues? Nothing like just getting a fastball, you know, super specific question, you know, so early in the morning. But what are your thoughts on that? Jamal I think that's a really good question.

And I think Cecilia is identifying, again, a fundamental challenge as well, which is how do you have an effective communication strategy internally, you know, given the compression of time that exists? CECILIA I think the point I'll make is that, you know, there is no such thing as a compression of time per se, because I think there is a natural arc of, of learning that happens over the organic arc of time. So I think let's, I'll give you certain examples of what we've done. You know, we started very early on these, these training curriculum which we made available to our resources. And this was through LinkedIn learning and other systems. We were very open in having our resources take courses. We continued down the pathway what we call fireside chats. These were consistent fireside chats that we ran every month. And what all of these environments gave all of our employees, the ability was in a very disarmed way, in a very sort of environment where everyone can be vulnerable, where we said, you can ask the basic question, we want you to come along this ride with us. And I think that sort of communication strategy, knowledge learning strategy, and again, that ability for folks to say, hey, we're willing to sort of slow down so that you can come along, gave us all this culture and environment within connection. Connection, by the way, is a very familial sort of culture to begin with. We're very much, as a company more driven by how we deal with our employees and treat our employees. And I think that was an important aspect on our journey. And I think that that is an important aspect to sort of address this complexity of our technology and its inherent communication with the lack of time. But I think from a timing perspective, Cecilia, I think it's just an organic arc. It just takes the time it's going to take and there's not much we can do about that.

Love that. I think we have one more here from the audience I'd like to get to. So, yeah, Jamal, when we, it seems like for the last year we've been asking this question, is AI transformative or do we not even fully understand it? So to Monica's question here asking how much longer do you think at least here in the US we will be talking about AI implementation versus full blown adoption? So what's your thoughts on that?

So, so Monica, I think you, you'll be surprised to know that there, that AI adoption is already well underway and, and quite often in a way where, you know, it's behind the scenes. You know, every time you go to Netflix, every time you're going to Amazon, you know, from a very consumer oriented perspective, it's using predictive models. The likelihood is that wearing a Walmart or in a Target, you've got, you know, computer vision based models that is looking at challenges such as shrinkage. So I think you'll be surprised how pervasive artificial intelligence already is in its adoption and utilization within sort of the enterprise corporate structure. You straddle into the other end of the spectrum, let's say you move into the vertical, like the DOD vertical. You know, there's, there's already considerable amount of work around what's called the C6 standard. And the C6 standard is about bringing artificial intelligence to the edge combat system, where combat systems have the ability in a very automated way to assess what's the best way to actually conduct a kinetic action. So you'll be surprised how pervasive AI already is. So I think almost the Hollywood esque approach to AI is something that in some ways I think is a departure from reality that we're not likely to see the Bicentennial man, or we're not likely to see sort of the Terminator walking around for a while. But I think my argument would be that AI is quite pervasive. It's subtle, it's sometimes behind the scenes and we're not even aware of it.

So, Jamal, kind of as we wrap up here, because we've gone through a lot, we've hit on a lot of important points from the trust and explainability of generative AI, you know, needing budgetary supports around initiatives and kind of this square root principle that you talked about of identifying, you know, internal champions and stakeholders that can push AI forward. But what's your kind of your takeaway message for business leaders out there that are still, you know, fully asking this question, is AI transformative or do we not even understand it? What's that takeaway?

I think we don't under. AI absolutely is transformative. What we don't understand is how it's going to impact all of us. And I think it's almost. Andrew Ng uses the terminology AI is like electricity, always uses that analogy. And then when you unpack, what he means by that is if you think about when electricity became mass scale and was distributed at mass scale, it had a massive impact on everything. It had an impact on certain industries that just disappeared overnight. For example, the candle lighting industries, the street lighting industries, the whaling industries, ice manufacturing industries. They literally overnight, in a very short, compressed amount of time, just disappeared. But electricity, when it first came out, gave rise to a whole broad set of new industries. You had entire towns shift historically. When you're looking at the industrial base in factories and industries were built near towns. I mean, I'm from the Northeast. You go into any sort of one of these northeast towns, you'll see, unfortunately, the decay over decades of how those mill towns that were built around rivers and streams just didn't need to because they were no longer using water as a means of driving their machines. They could sort of move into more urban centers and they started consolidating around industrial areas, around cities. So I think that's a very similar situation we're likely to find in AI. It's likely to fundamentally impact a lot of businesses in, in sometimes unfortunately negative ways, but it's likely to open up a whole broad set of new categories of businesses that we haven't even conceived of. So again, I believe that AI absolutely is transformative and it's happening due to those convergence that I initially talked about. But what its impact is going to be is something that we will unpack over the course of the next five to 10 years, if not longer.

Just a lot to think about, you know, to start your morning here. So, so much good information and takeaways. I can't wait to go re listen to this conversation now and write our newsletter for it. So Jamal, thank you so much for joining the Everyday AI Show. We really appreciate it.

It sure, absolutely. John, great to be on.

And hey, as a reminder everyone, yeah Jamal just dropped a ton of knowledge on our heads. Don't worry, we're gonna help you make make sense of it and give you some practical takeaways from a super informative conversation. So if you haven't already, make sure to go to your everydayai.com it's going to be in the show Notes. Sign up for our free daily newsletter and if this was helpful, please tell someone about it. You know, repost this if you're listening on social media if because if your company still hasn't been able to make this transition, I think today's show is really going to help with that. So thank you for tuning in today, and we hope to see you back tomorrow and every day for more Everyday AI. Thanks, y'.

All.

And that's a wrap for today's edition of Everyday AI. Thanks for joining us. If you enjoyed this episode, please subscribe and leave us a rating. It helps keep us going for a little more AI Magic. Visit your everydayai. Com and sign up to our daily newsletter so you don't get left behind. Go break some barriers and we'll see you next time.