EP 540: Solving the AI Productivity Paradox
30:54 Share

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. In theory, the whole AI thing should be fairly straightforward, right?

It helps employees do their jobs faster, which should increase revenue, which employers should be very happy about. It seems so simple.

Yet there's sometimes this paradox attached to it. Right. The AI productivity paradox, because here we are multiple years into this generative AI wave. Right. Artificial intelligence has been around for many decades, yet some companies are still wondering.

Why aren't we maybe more productive if we're using AI? Why isn't revenue soaring? And I think there's probably a lot of answers to those rhetorical questions, but don't worry. We have an expert today to help us figure it out as we try to solve the AI productivity paradox. What's going on, y'all?

Welcome to Everyday AI. My name is Jordan Wilson and I'm the host and this thing, it's for you. This is your daily live stream podcast and free daily newsletter, helping us all not just keep up and learn what's happening in the world of AI, but how we can actually leverage it to grow our company and our career. So if that's what you're trying to do, welcome, you're in the right place. It starts here by learning things on the podcast from very smart guests, but it

ends by going to our website at youreverydayai.com. That's where you're going to leverage what we learned today in our free daily newsletter. We're going to be recapping the highlights of today's conversation, as well as giving you bullet points of everything else that's happening in the world of AI so you can be the smartest person in AI in your company or your department.

So, I'm excited for today's conversation. If you're looking for the daily AI news as we normally do, it's in the newsletter. We got a pre-recorded one coming for you. So, if you want the daily AI news, it's going to be in the newsletter. So, go check that out. All right. Enough chit-chat, y'all. I'm excited for today's show and today's guest. So, please help me welcome to the show Faisal Masood, the president of HP Digital Services. Faisal, thank you so much for joining the Everyday AI Show.

Thanks for having me, Jordan. All right. So for those people that aren't right, like HP is obviously household name, right? But HP Digital Services, for maybe those people that aren't aware, what is HP Digital Services? Yeah, great question. HP Digital Services was an organization that was formed right around my arrival, which

HP's transition from not just being a hardware company that we all know of, supplying industry-leading devices for both consumers and the workplace, we are transitioning into a software business as well. So what we realized along the way was to have the

the future work be the center focus for HP, we needed to not just sell hardware, but also software. So my team, Digital Services, runs all of the commercial software for HP.

Love it. Love it. And, you know, for people that may not know, like, what does that software actually do? What does it look like? Right. Because, yeah, like a lot of people think of HP as one of the biggest hardware companies in the world. Right. So what is the software at HP Digital Services actually do for for clients?

Yeah, so we have a platform called Workforce Experience. This platform basically solves the mystery around employee experience that most large enterprise and smaller companies face managing their fleet of devices. So we supply the fleet of devices.

But how do you know if those devices are working as prescribed? How do you know if the employees are happy with those ways? How do you know if those applications are working as they should? How do you know if they're secure? So our platform provides three core things to our economic buyer, that's the CIO and our users, which is security, reducing any of the ticketing that's needed on the long tail issues that come up on desktop support by using AI to solve those issues.

and also collecting data on employee sentiment. So knowing if there's an issue with the machine, our product is able to collect data on whether those machines are causing negative experiences for employees and how do you improve those, either by refreshing devices, improving the configurations, patching the software, etc. So our goal is to make every employee's life easier and lower the friction.

So, you know, speaking of in a good transition, that's what generative AI is supposed to be doing, right? But here we are a couple years into this generative AI wave, and it still seems like there's this paradox, right? Like we have this extremely powerful technology that seemingly can do, you know, work that used to take hours in minutes, right?

Yet not every single company is printing money, right? It doesn't make sense. So can you tell us a little bit on what is this AI productivity paradox even about? Yeah, you can look at it in different ways. One is just the paradigm shift that happened post-COVID with just employees' relationship with the employer. The 9-to-5, 9-to-6, 9-to-8, whatever it is, used to be in the office has moved over to hybrid.

That introduces a challenge on its own. That's its own paradox. But when you add Gen AI to that, where now the employers don't know where the employees are, so some of them are not exactly thrilled about that. But instead of focusing on the productivity itself, they're focusing on the proximity of the employees, whereas employees are looking at tools to improve their life day to day.

And if I think about how AI is changing and evolving all of this, in most cases, you don't even see it. I mean, our laptops that we ship are AI PCs.

Some of the work that they're doing behind the scenes is just something as simple as your background can be configured to whatever you want without you having to even deal with anything. Because it's running all of the models on the machine. And if you look at any traditional employee that is just doing their daily work,

I would say the vast majority of them today should be using AI to enhance their productivity. Is everybody doing it? We don't know. But I think that shift towards that is happening pretty rapidly. You know, it's funny you bring this up, right? These

culmination of events kind of happening at the same time, right? So, you know, during COVID, you know, at least for our listeners here in the U.S., right, we had a lot of people go straight hybrid or, you know, still might be work, you know, work from home, you know, still many years later and we never fully

as, you know, U.S. economic society transitioned back to five days in the office, many companies have, but still so many are hybrid. So many are still work from home. And then at the same time, we have this generative AI boom, right? So my thought is, are there still maybe thousands,

hundreds of thousands, maybe millions of employees that might just be pocketing some of those time savings. Is that why all these companies aren't booming when Gen AI promises 30, 40, 50, 60% time savings?

Yeah, the answer to that would probably be it depends, which is not the best answer. I would say if employers want to get the most out of their employees, it would be best to have a really high hiring bar and establish that trust. Because if you're hiring the best people, you know they're putting in 150% every single day. Now, if they're using AI to do that and finding some time savings along the way,

How does that affect the employer at all? As long as you're getting exactly what you need, I think there's a fair amount of satisfaction there. I think where the friction occurs is when there is a lack of productivity, there is an issue with the performance. That's when employers immediately think, "Oh, this is because of hybrid."

Well, is it? And that's where I believe as somebody who's run teams and has been on teams that you have to find exactly what makes you most productive, which environment, whether it's in the office or not. And sometimes it's either one or the other. And the tools that make you productive, such as using Gen A, I haven't ever needed.

And deliver what's needed. And if you can't, that's a different issue. There are some pockets of challenges that have happened with this term called over-employment. I don't know if you've heard about it. People may have six, seven, you read about these things on Reddit, six, seven different gigs, and they're excelling at all six, seven. Well, then whose fault is that?

if they're excelling and getting exceeds expectations on all their reviews, then technically they've done their job. So it's an unsolved mystery, but I feel like the answer is somewhere in the middle, which is being hybrid, but at the same time providing the level of productivity that the employees need. Yeah. So is like...

When I think of this scenario, I think it is messy at times, right? And I have both, you know, friends, colleagues, you know, people I've talked to all the time that say, yeah, you know, I'm work from home. I'm hybrid. I've, you know, automated 50, 70% of my work. And, you know, not all people I talk to have six or seven, you know, different gigs and excelling at them, but it seems like it's

almost like the norm for a lot of people, especially maybe around my age, maybe people that grew up with computers, right, but still are kind of like, quote unquote, mid-career, still with something to prove, right? And now all of a sudden, they have a lot of time on their plate, right? So how do business owners, business leaders, right,

solve that, right? Because they don't necessarily want to be known as a micromanaging company, right? Punch in or come back into the office. So how can you still have that freedom for employees to do their job? And maybe they're doing it well with AI in 20% of the time, right? Where do you find the sweet spot there? Yeah. I think you need to separate the two things first, which is

large enterprises typically lag startups. So what you might see in these massive gains in productivity, whether it's through co-pilot and coding and customer service or what have you, where you use Gen AI,

those savings don't quite translate into enterprises at the same scale or size or percentages. So what might be 40% here, when you translate that to an enterprise, it's probably much smaller. A Y, enterprises typically move not as fast and adopt not as quickly. So that's one. Second is, I think you have to raise the bar on the expectations. Why are employers still, if your expectation was to get to X, well now get to X plus.

and see how the employee can catch up. And I think that the task then is then in the hands of the person doing the work to ensure that they can raise the bar themselves too from where it is today. If you expect that, it's going to, back in the day, customer service, which is the most basic use case, I would say, for AI, where you can translate a lot of that to non-human activity.

You would say you can answer a phone call or an email in 60 seconds. Well, why? You can answer it in two seconds because you have AI. So the goalposts have changed. What you thought was the SLA, the service level that you expected,

Well, with Gen AI and what's happening in AI should no longer be the same. The expectations have changed. It's like when you walk into a really good store, you don't really want to go to the one that didn't look quite as good as the previous one if you're shopping because the bar has been raised. I think this even applies to corporations and startups and other companies and business owners that what could be done in a day or could be done in two days a while back now can be done in hours. Well, then the expectation should be hours, not days.

Raising the bar is a good concept to think about when it comes to specifically working in this age of remote hybrid plus AI. Fossil, your background's very impressive, right? So not only now at HP, but I believe what you were at Alphabet, Staples, Groupon, Amazon, right? So you've worked at a lot of big enterprise companies.

I've actually been a little shocked. So I've talked to, you know, mostly off the record, but, you know, people that work at big companies, you know, trillion plus dollar market cap that don't have AI policies, right? Which is weird because it's also some of the companies building AI.

I mean, is that to blame, right? The fact that maybe big enterprise companies maybe just don't have an AI policy and maybe that's why we're living in this paradox? Is that a thing? I think that...

It's a great point because it reminds me a little bit of back in the day when I was hired to be chief digital officer at Staples. And for some reason, everybody thought this one organization is going to make the whole company digital. That's not how this works.

The notion of becoming digital is an endemic concept. It has to be in the veins of the company. Everything you do has to be thought of digital first, right? When back on the transition from desktop to mobile, it was mobile first. Now it's AI first. So,

Hiring a chief AI officer, well, congratulations, that one person and their team is not going to solve the problem. The entire organization has to be moving in that direction. So I'm typically apprehensive of those types of moves because I feel they are not exactly going to entail in the whole or doing exactly what you expect. What you're mentioning is interesting because you could establish any policy you want at a large company.

The employees at home and they've got their own machine. They're going to do whatever they want. How are you going to police that? And why would you do that?

So I think if you want everybody to use your particular tools that you have put together at your enterprise or your company, then give them the world class tools. And so they don't have to look elsewhere. Today, you can go and do whatever you want in any of these options, whether it's OpenAI or Cloud or DeepSeq or Lama, what have you, you get so many options. So if your options are not going to be the best, then

Employees are going to do what they want. And that's why you're seeing that the policy making is a bit loose because it's hard to manage. How are you going to manage that?

Yeah. So how should they, right? Like, again, like, you know, if we want to solve this paradox, right, it's obviously easier said than done. You know, I'm sure, you know, AI policy is somewhere, has to be somewhere in there, right? But for, you know, maybe our C-suite people out there that this conversation hits them in the gut.

And they're like, oh man, this is probably happening a lot more than I realize. Or maybe some are just turning a blind eye because things are going well. Employees are happy. They're sticking around. Revenue's great. So how can you actually solve this to make sure that employer-employee relationship is, quote unquote, how it should be? I think organizations should think about AI technology

in a super native way where every task that you do end to end, if it's enabled through those agentic workflows that make your life easier, then you're avoiding forcing your employees to do that through outside tools. So whether it's filing a ticket, resolving a ticket, responding to like, if you look at just the Microsoft suite, the co-pilot edition inside Microsoft, it has helped people. Has it helped them as much as we thought it would?

I don't know, but it has helped people. Now, Google, of course, coming out with their own versions with Gemini. I think ultimately the answer is unknown today. Why? Because there's so many options available outside that how can you prevent every employee from using those? So what do you do? You find the best solutions internally and enable them with that toolkit so they're not having to look outside.

Very difficult to do. I would say where it has helped a lot is AI is really good at the thought starting process, which is you can use these tools. The problem that employers have is it's their data from their company that's being exposed to these platforms that they don't want happening. So what would you do? You should build versions of whether it's chat GPT or what have you internally to

that are super powerful that you don't have to go outside for your enterprise work. I don't think that's happened yet, though. A lot of dependency is still on the cloud-based versions, whatever LLMs you can get out there. So I have a very random question that just popped into my head. So let's say, you know, HP Digital Services, you're starting a new arm or a new team, right? And you hire 10 new employees.

After a year, all 10 are there, but you find out that maybe five of them, because of the AI tools that you used, right, they've only been working, you know, 10 hours a week. So five, you know, really, you know, got onto AI, not working very much. The other were working their, you know, 40, 50 hour week, whatever, right?

Are you mad at either group? Right? Like, how can we as leaders, as people in management, right? Like without literally looking over someone's shoulder, how do you deal with that? And are you mad at either group in that scenario? I think it comes down to defining what the expectations are from the employees.

If you define the expectation to be, you know, deliver X by Y date, and that is happening, why would I be mad at anybody? And if I expected more, then I should be mad at myself. Why am I getting mad at anybody else? Because the employees are doing exactly what they were asked to do. How they get to that end state is up to them. Obviously, those using AI are going to excel because they'll have an accelerated pace of doing what they're doing. And kudos to them.

But to the earlier point we talked about, you have to raise the bar. When Amazon started shipping initially, it was free super-savvy shipping, which came in seven, eight days. That has now compressed down to next day most of the time pretty much for whatever you order.

So, the bar has been raised constantly and look what it's done to other retailers. They've had to raise their bar. So, ultimately it comes down to the employer and the manager having very clear expectations of what the employees do and encouraging them to use the capabilities of AI wherever possible and then sort of roll the dice after that and see what happens.

So it almost seems like, you know, way more work for people in management, right? People that are managing large remote teams. And, you know, luckily that's not me because I can only imagine the challenges, not just, you know, managing a large remote team at a, you know, fast moving, uh,

enterprise, but also just the rate of technology. Because my two cents, it doesn't move like it's never moved this quickly. In terms of the capabilities, the fact that we have agentic models that can reason like a human and we can dump all our context, it's kind of wild to think that we have technology like that now, but it's like, okay, do even all people in management know and understand that? Maybe not.

Um, what's, what's your advice, I guess, for people that are managing large groups of remote teams, they are giving them AI tools, but it's just like, they're not really sure what they're capable of. 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. 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 chat GPT 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 youreverydayai.com slash partner to get in contact with our team, or you can just click on the partner section of our website. We'll help you stop running in those AI circles and help get your team ahead and build a straight path to ROI on Gen AI.

Yeah, I mean, that's problematic. So if you've got a leadership team that doesn't have a clear understanding of the capabilities of those agentic tools, that's a whole other problem to resolve than probably have the wrong team. Because this is sort of table stakes at this point. You have to know what's available and what can be done. I think in today's environment, especially in what we do in workforce experience, we

You've got enough tools that give you visibility into who's doing what to the degree that you want, where you have sprints every two weeks and you're delivering this and you're pushing this much code and here's what the product looks like. And velocity and quality are the two metrics that are most important in something like this. If you're getting the velocity you want and the bugs are not rising and the customers are happy,

I think you don't have to overdo yourself. Just keep improving that capability as much as possible. I think where you leave everything static is where the problems occur. You start questioning your team. I'm not doing enough. Others are getting ahead. I don't think that's the answer. The answer is keeping up with the times yourself and raising the bar and all those metrics that you track. And ultimately, is the customer experience getting better faster? You know,

One thing I personally see companies make mistakes on is whether they're normally hiring a group of 50 people annually or someone leaves, someone retires and they look for a replacement for that person.

I don't know, this might seem callous, but I don't think we should be hiring for human roles, right? I think we should be hiring for augmented roles, but we're still, I think, most organizations are still hiring, putting the KPIs, job descriptions, everything around what it looked like 10 years ago, not what it looks like in a year or two, right? But

how can you solve for that, right? How can we make sure our future people that we bring on our team are not just well-prepared for the future of work, but how can you even make sure that your organization is nimble and agile enough to adjust to what the future of work in an augmented society looks like? It's a really good question. And I think it also triggers another question, which is,

What got you here isn't going to get you there. So let's just take an example. If you're in customer service and you're hiring customer service reps, right? In the past, those reps were answering calls, answering emails, and perhaps there's some text messaging as well that they were answering.

Fast forward, if you were having departures in that, you probably heard about Klarna where they said they reduced 75% or some, I don't know the exact number, but they said they were able to reduce it by 50 or 75% without having to backfill. In fact, they said something like they've stopped backfilling.

And then there was an article later that, no way, we're actually backfilling now again. So there is no silver bullet to all of it. But I think what's important to know is what body of work can be done through the agents versus what body of work can absolutely not be done with the agents. And I'll give you my personal example. I had a problem with my car and sent, obviously, the message on their chat bot or whatever it was.

And the answers were not satisfactory. So to get the satisfaction, you had to talk to a person on the other side. Now you could argue, what was that person AI? Was it an actual person? Well, I don't actually care. As long as my problem gets resolved, I'm good.

But where we are today, you have to evaluate backfills. They can't be exactly what they were before. So you're right. If somebody's in that role for many years, is it the same role? Don't know. But that's the responsibility of the manager. Like writing the JD, do the hard work. Write the JD, be precise. Have the right expectations and tie them back to customer experience. And if that's not going to happen, then I think the problem is not the employee. It's actually you.

It's a good point. It doesn't sound easy. And yes, I'm teeing this one up in a very cheesy way, right? So if you listen to this show, I always say AI is not an easy button, right? You have to build it. You just can't click the easy button. I have to ask you this, given your background, staples, the easy button.

Can you quickly tell everyone the story just of how that easy button came to be and how you actually use AI? I think it's such a fascinating story. Yeah, it's a wild one, which is back in, I was at Staples joining around 2013. And I realized that Staples had done an unbelievable job at marketing. They had this marketing gimmick.

called the easy button, which obviously had sold millions of units of the actual button itself. That would just say, for those who don't know, say that was easy. And again, it was just a gimmick. It was just a marketing tool that was used in many places. We realized in our team, I ran the e-commerce business there. We realized that time was coming that our buyers on the B2B side

did not really have to place orders every single time going into the website, placing the same orders every single week. Why don't we just make it easy? Just click a button and be done. So in 2016, we launched the AI-based Easy Button, which basically took about a thousand commands. And those commands were something as simple as reorder me pens or take my return or where's my order, the typical questions. And we actually launched it and deployed it to our test customers.

The unfortunate part is it was powered through, back in the day, IBM Watson, which was at that time the leader. But too early, too soon. Great idea. Timing didn't quite work out, but it's fascinating to see them back in the same similar roles now many, many years later. Yeah.

Such a great story. Yeah, I would kick myself, you know, how much I talk about the easy button if I didn't take that easy opportunity to ask you that. But Fossil, we've covered a lot in today's conversation, right? You know, solving this, you know, AI productivity paradox, we poked at it from

many different angles. But as we wrap up today's show, what's your one most important takeaway, whether it's for business leaders and, you know, quote unquote managers that are managing people or whether it's employees, you know, trying to scrap out even more productivity. What's your biggest takeaway? I think my takeaway is a combination of just my experiences in the past. And one I would call out is Fabric, where I was CEO, it was a venture backed startup

that just hiring more people to do the work is usually not the answer. The employees are not happy. The employer is probably going to struggle too. I think the less is more approach is probably quite valuable here. Do more with less. That doesn't mean put more work on the employees, but enable them through a toolkit and all of the capabilities that they can get the job done without having to hire armies of people. The last thing I'll say on this is that

just having a very large team actually introduces a ton of complexity in just getting the work done. So many handshakes along the way, so much bureaucracy. So if you do deploy the less is more approach where I don't know if you saw recently Shopify also said that, you know, we are going to evaluate every single hire and look at if AI can do this. I think it's a right approach to look at because employees want really good work to work on that they can feel productive and employers don't want to

to have a lot of people doing the same thing. So less is more, I would say, is the approach, whether you're setting goals or otherwise. Fantastic parting words for one of these scenarios that I think so many of us are going through and probably will be continuing to go through for a long time. So Faso, thank you so much for joining the Everyday AI Show and sharing your experience and insights. We really appreciate it.

Thank you. All right. As a reminder to y'all, that was a lot. Maybe you missed one of those nuggets in there and you're like, wait, what was that? Don't worry. We're going to be recapping it all in our newsletter today. So if you haven't already, make sure to go to youreverydayai.com. Sign up for the free daily newsletter. If this was helpful, please tell someone about it. If you're sharing on social media,

don't let this be your cheat code, right? We all need help to get through this together and to understand and grow our companies and our careers. So thank you for tuning in. 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 youreverydayai.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.