EP 150: Navigating AI's Tsunami - Strategies for Recruitment, Retention + Growth

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.

Using AI and even just keeping up with it can seem like a tsunami at times, right? Especially what We've seen with OpenAI the past couple of days. But seems like there's always new tools and tips and technologies and how do we make use of all of that? How can we build strategies to recruit better retain better talent and to grow, Right? To grow our careers and to grow our companies. That's what we're going to be talking about today on Everyday AI. Welcome. If you're new here, my name is Jordan, I'm your host. And Everyday AI is a daily live stream, podcast and free daily newsletter where we help everyday people like you and me not just understand AI, but how we can all actually use it to grow our companies, grow our careers. I'm extremely excited today to talk about those things, but before we do, as always, let's first take a quick dive into the AI news. And actually, you know what, if you're joining live, thank you, but let me know what you think of all this OpenAI drama, everything that's been going on. We're going to be talking about that here in a second. But also if you're listening on the podcast, thank you for that as well. Always make sure to check your show notes. You can always come back in and join the live conversation after the fact. So let's talk about what's going on in the world of AI news first. Yeah, there's other companies out there in the AI news right now aside from OpenAI. But anthropic yesterday announced Claude 2.1. So Claude is their large language model from Anthropic. Kind of the big piece of news here is a new extended context window of 200,000 thousand tokens. What that means if you're not a dork or care about tokens or do token tests like I do, essentially that gives you the ability to relay over 500 pages of information to Anthropics Claude, if you are on their pro paid plan. So the company Anthropic also said that this new update in 2.1 cuts down or gives a 2x decrease in hallucinations. So apparently fewer hallucinations will be testing that out, as well as improvements in their API which could finally bring web access to Claude. That's why I tell people don't use it right now. Well, maybe that'll change in the near future, especially bringing Web access could be huge. Also, they announced a new workbench which is similar to OpenAI's playground. And this is very recent, right, because Google just a couple of weeks ago announced a $2 billion investment. A couple of months ago, Amazon announced a $4 billion investment. So if you want to know who's competing with OpenAI, it is anthropic cloud for now. Speaking of OpenAI. Yeah, we're gonna talk about it, y'. All. We're gonna talk about it. So OpenAI has officially, officially, Officially, right. Because could you, could you even keep up with this? It was, it was a tennis match, you know, back and forth watching this. But OpenAI has officially brought back Sam Altman as their CEO. So original founder was fired on Friday. Still so much drama back and forth, you know, non stop. But it is official, official now after much negotiation to bring him back. So OpenAI in a statement says, we have reached an agreement in principle for Sam Altman to return to OpenAI as CEO with a new initial board as of Brett Taylor, who's the chair, Larry Summers, and keeping on Adam d'. Angelo. Interesting. And then saying we are collaborating to figure out the details. Thank you so much for your patience through this. Some responses. Sam Altman put out a statement on Twitter saying, I love OpenAI and everything I've done over the past few days has been in service of keeping this team and its mission together. Also important to know because originally Microsoft CEO announced that Sam Altman was going to be joining the Microsoft Team. So about that, he said, when I decided to join Microsoft on Sunday evening, it was clear that was the best path for me and the team. With the new board and with Sadia's support, I'm looking forward to returning to OpenAI and building our strong partnership with Microsoft. So Microsoft CEO Sadia Nadella also put out a statement saying that he supports Sam Altman's move away from Microsoft and back to OpenAI. A lot going on. And then also apparently it looks like Greg Brockman, who's the president, is back as he shared a selfie in the wee hours of the morning, looking like he's back with the OpenAI team. So, so much going on in the world of AI news. We're going to be dissecting this probably a little bit more next week. And also Emmett Shear, interim CEO, who lasted 72 hours, presumably he's done OpenAI really didn't say that. He put out a statement essentially saying, you know, he's glad to have been a part of the solution. So presumably his time as interim CEO is is up. So, yes, that's a lot. We're going to have more news as we always do. So make sure if you haven't already, go to your everyday AI dot com, sign up for the free daily newsletter. But that was a lot. Thank you for your patience, y', all, because I'm excited. I'm excited for today's guests. So please help me welcome to the show. Let's bring her on. There we go. We have Usha Jagannathan, who is a responsible AI leader. Ex McKenzie, Usha, thank you so much for joining the show.

Thank you. Thank you for having me, Jordan. You know, it's a pleasure to be here and especially right before Thanksgiving. And I'm hoping many people would be unwinding if they are on the show. Wonderful to see all of them here.

Yes, yes. Always, always great. You know, if you're a podcast listener, you got to join the live stream. There's so many everyday people, even some AI professionals, people from big companies. But hey, shout out to Dr. Harvey Castro saying, great to be here. Brian saying, good morning from Minnesota. Woozy joining us from Kansas City. Thank you all. What do you want to know? What would you like to know from Usha when we talk about navigating this AI tsunami? But let's start there. Usha, the last couple of days have been insane. What's, what's your take on everything that has unfolded? Because it's literally been a tsunami the past five days.

It is, it is. So, you know, we are overwhelmed by this AA news and what since last Friday. Right. So we are struggling to understand the benefits, risks and, and also, you know, how a board can take us forward because it started all with the executive board leadership. So what you see here is the takeaway for all of us is like the diversity matters and how the leadership sets, you know, it speaks volumes like how the leadership sets the strategy and how they create the vision. So those are all the things that I look at as the takeaways and I'm hoping like it's dust is settling down where, you know, Sam Altman is going to be expected to return to OpenAI as a CEO. And we did see the news on Microsoft CEO talking about it like wherever he is, you know, we are going to support. So those are the things or the takeaways. You know, how it was a win, win situation at one point, what Microsoft was suggesting. And then later on, right now we know Sam Altman being expected to return back to see open OpenAI. So it has been a Whirlwind and the takeaways for us is like how are we going to. Because we are, it's AI is disrupting the world every minute, every second of the day today and how we are going to navigate this and how we can build responsibility products and also make sure like we can be able to you know, stick to the mission what they are talking about as benefit all of humanity. Right, so the diversity then definitely matters.

Yeah, absolutely. And real quick, you know, let's, let's maybe even Usha just, just hit rewind. So maybe could you just explain a little bit even about your kind of personal and professional experience in AI? Because I believe at, at McKinsey you were helping to develop and deploy some AI applications there. So maybe just give us a very brief overview of kind of your professional background as it comes to AI.

Yeah, sure. Yeah. So I started off as in the software engineering background and then I pivoted when I was starting to build a lot of faculty student AI projects at Arizona State University. And that's when I said like okay, I would like to come out of the zone comfort zone and move into the corporate side. And I went into corporate working at Marsh McLennan earlier I have worked for Washington Post as a consultant and then I moved into McKinsey. And the recent experience with McKinsey was like trying to build and deploy a applications and also making sure that how we can deploy it in the cloud, like a lot of cloud migration that's happening across all organizations and making sure that I am a passionate, more responsibility advocate to make sure we are able to build with fairness, transparency, accountability. Because all these things matter. Especially if it's a management consulting. You look for the accountability everything in every organization because every, you know, the regulatory standards, how they are scrutinized, the auditing for all of this, the AA models, what the responses it spits out, you know, what it gives. We want to make sure that it is ethical and fair manner.

Yeah. You know what? I love that you had a background of the Washington Post. You know that always hits home. I love having people who used to work at newspapers. I used to work at newspapers way back in the day. But let's, let's talk a little bit Usha just about present day because it seems like, like companies are always struggling, especially smaller companies, you know, not the size of, of your McKinsey's in the world. But how can you know, maybe what are some guys like high level strategies, you know, for companies on how they can start to use AI and to start to navigate this AI tsunami, maybe specifically let's, let's talk recruitment. So maybe what are some strategies for companies and how they can best use AI for better recruitment?

Okay, sure, sure. Yeah. So it's basically like starting with the recruitment. If you talk about today, what I see is we have seen the tech layoffs have become the new normal, right. In the entire past, like last nine to 10 months, right from since December 2022, November or December 2022, right when Twitter dismantled the media communications team, the responsibility team. Since then, up until now you have seen like even very recently we saw meta dismantling the responsibility team. So not necessarily I'm talking about responsibility, but many teams that got collapsed. What you see today is like the recruitment has started but the recruitment, what they are doing is they are putting the feeler out. I feel personally that so they are taking their own sweet time in order to hire people. And there could be, but at the same time today it comes with, it's a need of the above. Where many companies, what I notice is like they want to close their budget for the end of the year. So there might be people who would be hiring. And so we need to watch out what they can close out before the 15th of December when everyone go for holidays. So that you could be able to, you know, apply and get the job. What you are looking for because you want to land well in the next role, whether you're transitioning or you're basically coming out of as a graduate and trying to seek for new freshers for all of them. I would suggest that you know, look into the media and look into what they are looking for. Study the, you know, do your homework with the company and see whether you could be able to reach out and apply. And it's not easy said than done. So it's like how we make it work that, that takes a lot of time and you know, proper planning and organizing.

So yeah, that was great. You kind of tackled it from both sides. But maybe even from you know, people who are trying to grow their careers with, with AI, which I think is just as important because sometimes from what I've seen and from what we've talked about on this show, sometimes it's, it's the individual employees, you know, in these, you know, small and medium sized organizations that are actually taking the bottom up approach, right? And they're saying, hey, you know, the company hasn't said anything, let's try this. So maybe how can you know those out there listening, how can they maybe upskill or reskill, you Know, you know, let's say they're, they're mid career and maybe they, they, they don't have, you know, AI governance in place or AI initiatives in place. How can they upskill and also help to, to bring that to their organizations?

Absolutely. Yeah. So there was a recent article that came in Harvard Business Review where they say, like I've noted it down here, so the average half life of skills is now less than five years and in some tech fields it's as low as 2 and a half years. Which is quite true. Right. So the companies, what we see, all the organizations upskilling is alone won't be enough. But in order to upskill they are like they all are in this where they want to see how they can upskill their employees because they don't want to, want to lose their talent because once again using the resources to bring back new talent, it's going to cost them. So the need for a reskilling revolution is quite apparent. Right. So one thing is reskilling. It's, it's, you can see in the Harvard Business Review what's the return is, it's a strategic imperative because in this labor market what we see today, we want to make sure how we can reskill based on what the company it's based on. Like you need to develop skills that are company specific. So let's say it's a supply chain or it's a beauty brand company based on that. What do they are looking for? If it's specific to AI, data and AI, then what do they are really what tools that they are using, what they are looking for. And you could be able to see in YouTube and many medias that where they would talk about, oh, we are leveraging AWS and we are leveraging LLM integrating this particular tool, Gena Tool. So that gives us a hint. Okay. This is the cloud technology that they are using. So let's upskill on that. If you're applying for an architect position. Okay. Something that they might need in order to show it doesn't have to be a certificate culture but having that credential companies look for.

And you know, we will be sharing this in the newsletter, this you know, kind of article from the Harvard Business Review that Usha just that just mentioned. But I'm curious because I'm looking at it now and it says a generation ago the half life of the value of a skill was approximately 26 years. Now the half life is often less than five years. Right. So I believe half life is kind of your ability to you know, retain the skill set that you kind of learned or have acquired Usha maybe. I mean, do you know why it's, you know, why it's down like that? Like, is AI changing how employees are able to learn and properly leverage skills? Or is it just we need to learn and leverage way too many skills and we need to upskill now, maybe monthly instead of yearly? I mean, why, why do you think the shift. 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 Genai. 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 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 Gene.

Why, why do we think the shift. See, it's. See earlier we used to call it as a digital era, right? But today we call it as everything. It's an ever evolving AI landscape. Every day some tool is coming. We cannot keep oppressed with each and every tool. And that's why I mentioned as the HBR review is talking about, we need to be company specific. Let's say that you are working for a management consulting but they might have a technology arm behind it and if they are working on certain tools then make sure if you are not, you might be working on one particular tool, just don't get hooked to that particular. For instance, you might be very good in Python and Java, but if there are some skills that it is better to learn. If I'm back in engineer having that holistic experience, something else that you would like to learn to have that experience that might help, that might help you. I always tell whether it's students, community or the junior colleagues that I mentor or my peers that I tell, like if we don't keep up with the technology, we will become Technologically obsolete. That is so, so true today. And I'm sure you, you agree. Yeah, it is hard. It is hard. It is hard. So I generally tell them, you know, like for instance, you know, Dr. Andrew Ng or like, or like all these data, all the scientists, the experts that you know, if you are following them, they have the, for instance, deep learning courses. Some of the courses that they have make sure what, what it is you can take it. Like for instance, they have a wonderful course as generative AI for everyone. You do not need to have coding skills in order to have the AI learning. So right now what you see, especially today you see is like everyone is going towards the no low code. No code. So how we can take for instance Party Rock or one tool that they have introduced service that AWS has introduced recently, so it doesn't need coding. So how we can learn that and leverage so that we could be able to apply if your company using aws, how you can apply that? So that's how I look at.

That is such a good point. And it's something that we probably, Usha, don't talk enough about because I think there's maybe rightfully so, but there's this stigma around artificial intelligence because it's not new, right? Artificial intelligence has been used in different sectors for decades. But before, it's like, yeah, you might have had to have a master's degree or a PhD in computer science or something else to take advantage of deep learning and neural networks and building these models. But it's not like that anymore. Like you mentioned, right? No code or extremely low code. Right. So we've seen some recent offerings from even Microsoft 365 Copilot, their new studio that allows you to build some really robust solutions with, with drag and drop. So even with that in mind, maybe let's talk about that a little bit more, you know, because maybe the average everyday person is kind of scared when they hear, oh, AI or generative AI sounds difficult. Sounds like I need, you know, Java or Python or whatever. But it's not, that's not necessarily the case. Right?

It is not. It is not. So for instance, currently I'm like, I'm. I'm associate consultant for Arizona State University for the W.P. carey School of Business. So what we do there is like we, we work on the industry, academia, partnerships. And what I do is it's like we are building technology training programs in communicating with the industry. So let's say it's ServiceNow, for instance, then the we in ServiceNow, you notice like you do not need to know the entire ServiceNow tools, but they help you to. What we do is we build a course and. And that course can be able to be leveraged where the people. If you want to become an IT admin then you can be able to take that course and that course can help you at the end create a. Build an exam virtual. You know, like you can take a certified system admin exam and once that exam is done they might take you for an IT associate technician level course. But what I'm trying to say here, they also have introduced AI in IT and all you need to know is how you navigate the ServiceNow platform and understand it and then how you can be able to leverage their AI applications. Everything you can do it with less code, but you are able to understand how to navigate the platform. You know how you can be able to manage their instances very beautifully. The course is structured and so I'm going to facilitate the course. Just I'm giving an example. So people are. The companies are trying to see how they can partner with universities and bring these programs so they can get the talent also talent pipeline. And at the same time they can show like it's open to everyone. Not like. Not like only people who have a certain degree like an undergrad or a grad degree. But anyone can be able to use their skills. So that is something like, you know, I feel like there is no necessity for degree in order to. Because you see like how they are going paperless which I want to bring this up in a minute. I would like to take on this. Where there are many organizations have apprenticeship programs, work study programs where they do not look for. I have spirited these programs at McKinsey Marsh McLennan where you do not need to have a certificate showing like you have a degree. But we bring through vendor partners and then they. They might have a background and there are some people who had background in carpentry in something, you know, janitor supplies and it's totally different. But they come with the passion of developing software and that's where they start. And you don't expect that they need to know in and out of AI skills or in and out of software development skills. But we trained them, groom them, that was one of the part of my role and mentor them. And after six months, 18 months, we assess, reassess them. If they are good then we, you know, good in their tech talent. Then we hired them. So I don't think that everyone needs to come with a certain degree requirement as long as you have that passion I love that. Yeah.

Especially in a field that is growing so quickly. And this is a conversation for another day, Right. Because half of the, you know, universities out there are banning generative AI yet, you know, so many companies are needing to hire people with that experience. So Usha, great point that sometimes all it takes is if you work at a company, have the passion, upskill, reskill. Such a great point. Gonna take a question, great one for Mike here. So just pretty. Pretty simple here, but just saying, Usha, please tell us about a solution you have built. Presumably, you know, in AI. But yeah, maybe let's talk, you know, specifically what's. What's an example of. Of a solution that you've helped build around AI and maybe what was the. The impact that it had?

Oh, sure. Yeah, definitely. Yeah. So. So I would like to step back like at Marsh, like one company before, like that I work for Marsh McLennan. So the products we build is for customer facing or it could be internal facing. So thank you Mike, for asking that question. So it was like for, let's say for client claims approval for insurance code, the application is like, you want to make sure that aa. What it builds the a. Like, sorry, what the product that. That has been built. You want to make sure, like let's say you are applying for claims. You have submitted a claim and I have submitted a claim, but yours got accepted and mine got rejected. Then I need to know why it disapproved that. Why the model, you know, rejected my claim. Then it would say like probably it rejected. It gives you the counterfactual. So I created an explainability with my team to add an explainability component where using lime or shap, where it provides you a local or a global interpretation, but at the same time it provides the counterfactual saying like, okay, if it's a claim or an insurance.

Both.

We did both claim approvals model and also insurance cross selling. Okay, if this the insurance code. So if for instance, like the last three months of Usha's utility bill was not paid on time, and for these reasons this, you know, the insurance code was rejected. And if she were, she had paid it on time. And if she can come back after three months and does everything on time, probably it would, you know, the. Her insurance code can be approved. So it gives you or alone can be approved. These are like three different sets of products I'm talking about. I'm not trying to go in tangents, but what I'm trying to say is giving that counterfactuals that helps to Understand the customer. Oh, okay. This is the reason mine was rejected or mine was disapproved. And it is not something of gender bias or anything. So that is that explainability component is extensively. I've worked with Mike and where we try to create the added counterfactuals and say this is the reason that this was denied or this was disapproved and if this can be changed in the course of action when you reapply, yours will get reapproved. And that is that model that I built, I felt like that was. Felt like an accomplishment. The reason is you are trying to build something not only creating a seamless digital experience for the customers, but you are also making sure that the customers are able to understand what exactly the model is providing the response. Because there are so many innumerable chatbots out there today. And you want to make sure the response that you get is not only today human like content because of the LLM that we leverage, but we also want to make sure that it is fair and it provides you the right information. Right. And it, you know, you need to feel comfortable about. Okay, this is what that I expected. Oh, and is this the reason that they need that explanation? Is this the reason that my claim was denied and they want to know. They have every right to know that answer.

Yeah, the explainability pieces is always huge, right. Because both internally you have to kind of be able to interpret the whole black box. But the explainability also helps on the back end as well. You know, maybe in this example, why a customer's claim maybe was accepted or why it was denied. That's. That's a great point. So this is another, another great question here from Cecilia. Cecilia, thanks for joining us. Because this one hits on kind of strategies, retention and growth. So Cecilia asking, have you seen how AI can help identify capacity for scale skills and candidates when specific skills are not identified? I love that because we just, you know, kind of had the, you know, the janitor kind of example, you know, kind of transitioning or upskilling. But how can maybe organizations use AI to help identify skills in, you know, candidates or maybe even employees that aren't easily identifiable. Is. Is. Is there a way to do that?

Yes, yes, I believe. Yeah, I'm sure like a couple of years ago, you know, the. A hiring algorithm that, you know, every organizations uses this. So even McKinsey uses the hiring algorithm. So making sure that there is because we did face like quite a few years ago, like where, you know, we all saw in the news where Amazon had the hiring algorithm issue and then they fixed it. So the same way here what they try to do is like whether what kind of skills is it matching matching what they are looking for. For instance, like let's say I'm coming with no background at all with some. Some background in history. But I am passionate and curious about. But what cecilia that we look for is that ATS will be able to track like certain keywords that how it is fed and making sure like have they done some projects even if they have not built any applications, have they worked on some projects, some software development projects and and what we look at is like if they have developed some boot camp projects through that vendor and what kind of projects that they have showcased and those are the things that that gets picked up. And then even you know we just. It's always a human being, a human in the loop. They say there is a human behind it and not just discarding the resumes like making sure like okay, if there is something like that they have built the projects, we take that and we look into it. Okay, why don't we bring this candidate and have a look at it and talk to them and why they are interested to come come in here as an apprentice. So you know, starting with an apprentice in a junior level role and trying to understand. So what I would say is like the specific skills if it is not identified, it gets identified in based on how you feed, you know, certain stop words that you give, how you feed in the algorithm and making sure that you don't look for whether they have bachelor's, whether they have masters, but also look for what type of project they have done. And I think even if you are bringing people with graduate level people. What I look for is I have recruited so many people and what I see is you. We look for what type of projects that they have done. How it is going to be relevant when they come here to get into the transition into this role. Because that matters because that hands on skills is what every company looks for. Even when you are coming into a leadership role there is like the technical skills is what they look for so you can be able to steer and mentor the people that you are going to train in your team.

You know Usha, I feel, I feel we have a little, a little bit better idea now of how to kind of navigate this AI tsunami and everything that's happening. But you know, because we did talk a little bit about you know, strategies for recruitment, for retention and for growth and, and a lot more. But maybe what is, what is one takeaway as as we wrap up today's show, what's maybe one takeaway that you really want? People, whether they're trying to grow their companies or grow their careers, what is the best way for them to navigate, you know, this AI tsunami? What's that one piece of advice that you have for people?

Okay, sure, sure. So make sure like be curious, be curious and be open to trying things come out of your comfort zone. Whether you know people, if you have worked completely in like let's say you are a student watching this podcast, make sure that you might have done all projects in the academia setting. Do not worry about, come in and test the skills. Try to explore if you can take internship and even if it's not a paid internship, if you're in the middle of your junior year, before getting into your senior year, make sure that you are able to get some unpaid intern job and volunteer and learn that skills so that you get that industry knowledge and you could be able to easily transition into going for a paid internship and then probably the company will acquire you as an employee. So that's for the students, I'm saying. But for career seekers or transitioning, I would say still, you know, be curious and try to explore. Like even if you have never tried low code, no code, it's not like it's completely no code. If you're going to be a technique, if you have a technical background, then I would suggest like we would be customizing that code a little bit. It's not going to be completely zero code. Right. So in that way, because low code, no code is getting so much popular, how you can leverage that in your work because if you're using AWS or Google or Azure, definitely you would be making sure that not only cloud agnostic tools, you would be playing with what kind of cloud native tools with their services that you can leverage because already the licenses are paid by your company. So how you can be able to leverage those services and bring that in your applications that you are developing and make sure that we develop ethically responsible AI products because so that it serves for the social good, for the company, for the customers and for whomever that you build. That is something that I would say stay curious and come out of your comfort zone to try and explore.

That's perfect because that's what we do every day here, Usha. We always stay curious. We're always trying to learn. So thank you so much for joining us on the Everyday AI show. We very much appreciate, appreciate your time.

Thank you so much for having me. I really appreciate it, Jordan. All right. You all have a wonderful Thanksgiving.

Thank you. Yes, yes. To those out there in the U.S. we hope you have a great Thanksgiving. Yeah, we will be off the show tomorrow, but we'll be back Friday. Don't you worry. So thank you so much for tuning in. Make sure if you haven't already, go to your everydayai.com Sign up for that free daily newsletter. We're actually going to be making a couple tweaks and a couple changes, but we're going to be asking you about them first. So if you're not already signed up, make sure you go sign up and make sure to join us back again not tomorrow, but Friday for for Everyday AI. Thanks, y'.

All. Thank you. 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 everyday AI.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.