Paris Talks Marketing

(00:00) Shay: If you read the Financial Times or the New York Times Economic Section, they might've fooled you to believe that the biggest problem of the economy right now is because of AI automation soon enough there's not going to be jobs for anybody. There's going to be this mass unemployment and that automation and AI are going to replace labor in some wholesale way.

The reality couldn't be further from the truth. The problem in the next few years, and when I say next few years, I mean the next three, five, ten, fifteen years. It's pretty much exactly the opposite. The problem, as the data is presented right now, is not that there are not enough jobs, but there are not enough people who have the skills to do those jobs.

In fact, in the U.S. very consistently and to some extent in Europe in the same way, post-Covid, there are pretty steady two open jobs for every skilled laborer. So there's twice as many jobs in the market as people that can do those jobs. That statistic, of course, is a little bit misleading because in the statistics of unemployed people or job seekers, there's a whole slew of people that are not even counted because they're not seeking jobs because they gave up, because they don't have the skills that are requisite to participate in this modern economy.

So it's a two-edged tragedy, right? Or it's a two-sided tragedy. On one hand, businesses cannot find enough people. On the other hand, some segments of the population can't find work at all. They're not unemployed. They're unemployed a bit. That's the skills gap.

(01:30) Intro: Every SaaS company plays for high stakes, but what does it take to dominate the market right now? Welcome to Paris Talks Marketing, the podcast where we dive deep into the latest trends and strategies in SaaS marketing that are really working today. I'm your host, Paris. and Our guests are SaaS CMOs, founders and specialists, and we discuss one trendy topic in the industry per episode.

Ready to unlock the true power of marketing strategy? In this theme, we'll explore the world of cutting edge marketing strategies and tactics that are shaking up the SaaS industry. We'll share insights on testing new tactics and uncover the latest developments from digital landscape giants like Google, Facebook, and LinkedIn.

We'll also explore how AI is revolutionizing the digital landscape and transforming marketing tactics. So grab your headphones and get ready for a marketing strategy masterclass with Paris Talks Marketing.
    
(02:24) Paris: Hi, everyone. Welcome back to another episode of Paris Talks Marketing. Today my guest is Shay David. Shay is the co-founder and CEO of Retrain.ai. Retrain sees a world where people can work alongside machines and not be displaced by them. Where automation and AI help bring out the best in people, rather than threaten their livelihoods. Their mission is to create a reality in which every person has a good job and anybody can be gainfully employed and not subject to technological unemployment. Their strategy is to develop the world's leading AI platform for skills and qualifications assessment and workforce retraining while working with employers and governments to place 1 million people back in jobs in the first 5 years. As a serial entrepreneur, Shay brings to the table, many years of experience in dreaming up products and making them a reality from concept to market, from slideware to hundreds of million dollars in sales. Welcome to the show, Shay.

(03:27) Shay: Thank you, Paris. It's really good to be here today.

(03:30) Paris: It's great to have you. I would like to lead off with a question about what you were doing prior to Retrain, a company called Kaltura. And I believe this is where your entrepreneurial journey began. And you were there for 18 years. Can you walk us through that journey and how that ended up and how that led you to Retrain?

 (03:51) Shay: Yes, for sure. So, Kaltura my previous company, a company I had to start with three other very talented entrepreneurs was the leader in enterprise video and video for leading media and telecom companies. Kaltura started in the world without iPhones. And in a world without apps, if you believe it, but in a world where we understood early on that media is the primary data type that's going to play a growing role in our lives.

And the idea behind culture was that as media becomes prevalent in the world, while we knew how to do things on early internet with text, like how to publish a website and how to measure it and how to track it and potentially how to monetize when video came along, there was a lot of opportunity to be able to learn to do all those things with video. How do you publish video on multiple devices?

How do you make it cross device? How do you measure it? How do you monetize? How do you secure it at motion and at rest? Organizations didn't really know how to do that and Kaltura built a platform that allowed organizations to take all of these very different use cases with video, both internal and external, both for entertainment purposes and for business purposes and allow the world's largest organizations, probably maybe close to 30 of the Fortune 100 organizations, many of the leading media companies to be able to take the power of video to the masses.

(05:13) Paris:  That's great, and you exited there in 2020 and turned right around, looks like without hardly a break at all, and then launched Retrain in April of 2020, which is right around the start of Covid. And so you're going for act two of turning a PowerPoint presentation into a $100 million company.

 (05:36) Shay: Yes, I think so. And I'm hesitant to call it luckily, but luckily in some sense, there was a little hiatus because of Covid, which presented both a challenge and an opportunity. So I took a little break. I stayed on at the board of culture. I still am on that board, but, you know, give away my day to day responsibilities to other talented managers and Kultura, I've always stayed with the notion that people that were buying video, particularly for enterprise use cases were buying it. For as part of something larger when people would spend a quarter million dollars, half a million dollars licensing a video technology primarily for internal use cases or in many cases for internal use cases, I was always left with the question is, what are they trying to achieve?

Why would somebody buy a video player and video technology as part of some larger worker for meaningful amounts of money? And the answer was many organizations were using it to share knowledge internally. And when you deep dive into that, many of those organizations were trying to preserve organizational skills.

They needed to find people, they needed to onboard people, they needed to share knowledge internally. And when I started looking into that, I stumbled upon if you will, or I began understanding that the world of skills is very much top of mind for many organizations and many organizations are undergoing what now we have the language to describe as the SBO journey or the skill-based organizational journey.

They understand many organizations, many CEOs, many CTOs understand that the way forward for an organization in a 21st century knowledge environment is to be able to maintain, retain, grow, develop the skills that they need to get the jobs done. And in that sense, my first interaction with that type of need was through selling video systems.

People were using video because they wanted to preserve organizational knowledge, to make learning programs and onboarding programs more efficient, to save knowledge for employees that were retiring, etc, many use cases like that. But what most of them really wanted to do was to be able to generate, maintain, grow organizational skills.

A joke I like to tell is that a chicken is an egg's way of making another chicken. And an egg is a chicken's way of making another egg. Meaning that employees from many organizations represent the capacity to do work. People go and hire people because they want them to do something.

If I didn't want people in my team to do anything, then I probably wouldn't hire them and nobody would hire me. And skills become the yardsticks or the measures of that capacity to do work. So when you think about it simplistically, and I'm not saying that in any demeaning way, what organizations really are after is making sure that they have the right organizational capacity as measured by skills in the organization to get their jobs done. And the skill-based organizational journey is all about making sure that you can understand which skills you need, hire for those skills and develop those skills internally so you can grow from organizational skilling to organizational capability and organizational excellence.

And in that sense, for me, even though Retrain.ai and Kultura are completely two different domains, there's very common thread between them because both of them help organizations with that very important mission.

(09:05)  Paris: So the skills-based organization of today. I think a lot of these organizations have an identified gap, skills gap when it comes to AI and my company and lots of others that I've talked to have identified this gap, the AI skills gap, and I presume that now a lot of what you're doing is helping these companies to, to bridge that skills gap with AI. What are some of the top AI-based skills that the organizations that you're currently working with are out to try to improve and can can you tell me a little bit more about the skills gap that exists today in AI?

(09:45)  Shay: For sure. But let's take one step back and understand where that skill gap arises. If you read the Financial Times or the New York Times Economic Section, they might've fooled you to believe that the biggest problem of the economy right now is because of AI automation soon enough there's not going to be jobs for anybody. There's going to be this mass unemployment and that automation and AI are going to replace labor in some wholesale way.

The reality couldn't be further from the truth. The problem in the next few years, and when I say next few years, I mean the next three, five, ten, fifteen years. It's pretty much exactly the opposite. The problem, as the data is presented right now, is not that there are not enough jobs, but there are not enough people who have the skills to do those jobs.

In fact, in the U.S. very consistently and to some extent in Europe in the same way, post-Covid, there are pretty steady two open jobs for every skilled laborer. So there's twice as many jobs in the market as people that can do those jobs. That statistic, of course, is a little bit misleading because in the statistics of unemployed people or job seekers, There's a whole slew of people that are not even counted because they're not seeking jobs because they gave up, because they don't have the skills that are requisite to participate in this modern economy.

So it's a two-edged tragedy, right? Or it's a two-sided tragedy. On one hand, businesses cannot find enough people. On the other hand, some segments of the population can't find work at all. They're not unemployed. They're unemployed a bit. That's the skills gap. And those skills are mostly focused on digital skills, digital literacy.

Within that family of digital literacy, digital skills, 21st century skills, call them what you will, AI is accelerating that trend. So what we're looking for are skills that are going to allow people to thrive within that environment of automation. The natural skills whether that would be baseline computer skills, the capability to go and operate a machine, to read instructions off a screen, to be able to communicate with machines and whatnot.

More sophisticated versions of that, that we're starting to see in the market are people that are dedicated to work with AI. One new profession that just came up, or one new job title that recently became prominent, for example, is prompt engineering. Those are people that specialize in activating large language models, as an example.

But I think that the gamut is quite wide. And that gamut spans the entire field of digital literary skills that are becoming very, very important. And you can think about, you know, past professions like warehouse employees or tractor trailer operators or agricultural tractor operators. Those used to be pretty menial jobs.

Today, all of these professions have to operate robots, have to program machines, have to be able to read reports, have to be able to report incidents, etc, etc, etc. There is automation is bringing digital technologies and digital machines into the rest of the economy. And in that sense, I think that we're seeing dramatic acceleration of that skill gaps in those particular areas.

(12:59) Paris:  Let's zoom in a little deeper into marketing. Most of our audience are marketers at tech companies And in particular AI is really doing a lot of the work that a lot of traditional marketing roles have done. Things like producing content or producing creative for campaigns, even some of the strategy work. And I can say being a marketing agency that I see also that there needs to be a pretty massive transition of the skills that our marketing professionals have today and the future skills of working with AI. What advice would you give to me as a head of a marketing agency? And also you can think about this also as a, any marketing leader who's leading a marketing team and seeing potentially a lot of the work that people on that team used to do is being disrupted or maybe displaced entirely by AI. Well, what advice would you give to someone like me on how to transition the current skills of the marketing team to a future where they can upskill with AI?

(14:09) Shay:  For sure. So the first advice I would give you is embrace the change. Don't resist it. Because anybody who is in any area of the business, like my dear friend, one of our advisors, Carl Frey, called Symbolic Analysts. And Carl, by the way, is the author of a book called The Technology Trap, where he analyzes this current industrial revolution in light of the three previous industrial revolutions of the steam engine, electricity, and the communication revolution.

And he's looking at which jobs are more susceptible to automation and what is the likelihood that more jobs are going to be created. So Carl makes this distinction between professions where people analyze symbols for a living, right? You could be a programmer, you could be a marketer, you could be a copywriter.

The commonality among those various three different professions is that the people in them wake up in the morning and manipulate symbols for a living, right? I could write code, I could write copywrite, I could build a marketing automation plan and whatnot. So, digital marketing is a great example of that class of jobs.

So, tip №1, embrace the change, it's coming. Those professions are going to change dramatically because once you get computers, machines, LLMs, AI, automation, NLP, we can throw in more buzzwords into this alphabet soup. Those technologies are not coming, they're already here. And they're changing the way that this profession works.

So that would be tip №1. Tip №2 is figure out what machines are good at and what humans are good at. Machines think statistically, they don't think, they just compute statistically. So, if the job at hand is to summarize large volumes of text, for example, the machine could do that very well.

If the job at hand is to think about 17 different alternatives for your marketing slogan for AB testing, the machine could probably do that. If the job at hand is to be able to find specific pieces of information within large pieces of text because you're doing a competitive analysis, again, a machine could probably do that.

So machines are very good at that. What machines, at least today's machines, are not very good at is being truly creative, is really understanding nuance, is understanding cultural context, is doing fact-checking in the sense of kind of a human eye element to it that makes sense, etc, etc.

And I think that as you're thinking about, in this case, the marketing agency's jobs, one thing you could do that could be a useful exercise is to say, okay, in day of a life of one of the people working for the agency, whether that's a copywriter or marketing automation person or a creative person, whatnot, what are the tasks that they're actually required to do?

Which of those tasks or things that the machine could do better than a human? If the answer is, you know, half of the job, that's great. That's not the reason to fear. That just means that you're going to let the machine do all the things that the machine is good at and keep the human to do what the human is probably good for.

And there could be a situation. And I think that maybe not necessarily in marketing, that some professions are going to disappear and that's okay. It's of course, very scary. If you're in those professions, it means that. You probably have to make a change. And one example for that, I think that we're already seeing is transcription and translation.

So there's a whole big industry of transcription and translation from medical translation to kind of speech translation to transcription and courts and many industries that use it. Machines have already gotten to a point where they do that better than humans. So there's no point in competing with the machine in being able to transcribe and translate the text to multiple languages because machines could already do that.

So that's, for example, something that machines do better and that the machines do that, what a human could do within that industry, they can do quality insurance, they could do creative, they could do other things, but they shouldn't compete with the machine just like a runner wouldn't compete with a car in delivering FedEx packages.

(18:09)  Paris: Got it. That's good advice. Thanks for sharing. And I think that one thing that I would add to what you just said is that what AI empowers our marketing team and other marketing teams to do is to move 10x faster. So one way you can look at it is to say, well, if we move at the speed that we've always moved, then, if you look at it only within that lens, then maybe, yes, AI can replace a lot of the different aspects of the workflow that humans are currently doing today. But if the humans can spend less time, or maybe transition their time more to training and monitoring the AI outputs and on strategy and on creative. Well, how much faster can we move? And I'll give you an example of our own content marketing service that we have.

We used to typically produce between maybe four, six or eight pieces of content per month with a very, very human workflow from the research to the outlining and what we call a content brief to the draft, editing and publishing. Now AI can do a lot of those steps. They can prepare the brief. They can write the first draft, which then can be edited by a human. And what that has allowed us to do is really transform the service into something that we can now deliver instead of 4, 6 or 8, we can now deliver 15, 20 or 30 pieces of content.

So we've practically 5x the amount of the volume of content, or you can also think about it as we can move 5x faster with the same resources and with the same cost basis. And we can pass that value along to our clients and actually make that service even more profitable. And I think that a lot of people think not in the lens of this can help us to move faster, but rather that at the current speed, that this is what's going to replace what I'm doing today, and that can lead to fear.

And so I think as an employer and hopefully people that are listening that are on marketing teams, most likely the leaders of these teams are thinking, how can we use these tools with our current people to just accelerate what we're doing now, move faster and create more value. And if we do that, we're going to need even more people to do that because we've created, you've created value. And when we are creating value, we're going to be growing and we're going to be growing the teams.

It's just that the people are going to be doing the work that they're doing is going to change.

And it should actually become more interesting too.

(20:45) Shay:  Absolutely. And, and I think that, you know, one of the things that we could look at is look at industries that embraced automation early on aviation industry, for example, people might or might not know that. But today, a modern dreamliner takes off, flies across the country, lands in a different airport, taxes on the runway without human intervention, right?

That allows a very small team of maybe two pilots and an engineer to fly passenger airplane with 300 or 400 people on it, right? So the efficiency is exactly that. So I think if we look at industries where automation is almost taken for granted, we're starting to get a sense of what you could do.

So what's the parallel dreamliner example in digital marketing, for example, can you take a small dedicated team and just do a lot, fly your customers to new heights. I think would be the metaphor I'm aiming for. That doesn't necessarily alleviate the fear of an individual being used to do a very specific job.

Right? If you're a photo retoucher that has been using Photoshop and made a decent living for the last 10 years doing that for agencies, and our machine can do that not only more accurately, but a hundred times faster. Then yes, the fear for your specific job is understood. And I think that there's enough research right now.

And again, I mentioned Carl Frey earlier. One of the research papers he put out with another researcher named Osborne. So Osborne and Frey, you could look it up. Susceptibility to automation. You can look at different professions, you can understand, which professions are going to get automated if given occupation or given professions and high susceptibility to automation, people within that profession should be looking to make career transitions and hence, by the way, the name of the company that we started retrain AI, you know, we didn't call it skill fit AI.

We didn't call it AI match or anything like that. The core idea is that retraining is probably a new normal. People entering the workforce today are gonna get to retrain many times over in their career, whether it's within the same company or whether they would make stops between different jobs.

But we named the company Retrain because we think that is the new modus operandi of the markets. There's gonna be new technologies coming in, and people are going to have to be retrained. And it goes back to your earlier question about kind of what are those missing skills. Many of those skills are actually not about the use of tools.

They're not about knowledge. They're about soft skills. Like teamwork and collaboration and creativity and how to bring those forward within any given profession regardless of whether you're now working in Photoshop and retouching focus and tomorrow you're doing something else.

 (23:19) Midtro

(24:02) Paris:  Let's talk about Retrain a little more because this is a really interesting platform. And I noticed that you started in April, 2020 and the generative AI revolution started in like end of 2022. But when you started the company, were you looking forward already to AI and what was coming or have you now more repositioned around what's happening in AI now?

(24:27) Shay:  So I joke with my fantastic partners, Avi and Isabelle, two kind of industry veterans, we often joke that ChatGPT was the marketing campaign we could never afford, because all of a sudden people understand the power of AI, they understand what large language models, LLMs, are, they understand how these technologies could be incorporated into existing workflows, and they couldn't do that before ChatGPT kind of came to the market.

So in that sense, we didn't anticipate it, but we're happy to say that it's coming. AI has been around for a long time, you know, various versions of AI have been developed for the last, maybe 20 years with a big, big jump in AI that moved from kind of rule based systems into neural networks, maybe the last decade or so real, real big jump in the last six or seven years.

So in that sense, when we started the company, we named it Retrain.ai, we already knew it's an AI company. It's going to use natural language processing it's going to use modern AI techniques in order to better understand and facilitate these transitions for retraining, reskilling, upskilling inside the enterprise.

We knew that when we started the company. What was kind of surprising and exciting for us is to see the rate at which these technologies are kind of capturing the public imagination. And again, ChatGPT was a watershed moment in the sense that, it's the fastest technology ever to reach 100 million consumers.

So it's kind of wildfire and people now understand. Oh, wow. I understand what AI is now. What people less understand is that something like a ChatGPT is like a Swiss army knife. It's really good to do mundane things all across the market, but you're not going to use your Swiss army knife to shave and you're not going to use your Swiss army knife to cut your salad at home, right? You would use a dedicated proper chef knife at home and you'd use a razor to shave. You and I don't really shave. So maybe you don't know that. But,

(26:18) Paris:  I still do some trimming though.

(26:19) Shay: Exactly. And you're not going to do that with a Swiss army knife, are you?

(26:22) Paris: No. Maybe around the campfire.

(26:25) Shay:  ChatGPT is kind of the same. It's a very powerful technology. It's really good for the baseline. You want to use dedicated tools in each domain. And one of the things that we see happening across the market is that in every vertical markets, there are going to be the parallel ChatGPTs.

That are going to be developed. It's true for oil and gas. It's true for marketing automation. It's true for e-commerce. It's true for healthcare. It's true for HR. They're going to be dedicated tools that use the same foundational technologies as ChatGPT, which are large language models and knowledge graphs and neural networks and many of the kind of core natural language processing techniques.

And image processing techniques and computer vision that are being used by a ChatGPT. But instead of a chat interface that has kind of a general body of knowledge, those are going to have dedicated bodies of knowledge per vertical. They're going to have dedicated integrations per vertical with the existing stack.

And most importantly, they're going to have enterprise consumability. They're going to have the security models, the privacy models, the auditabilities that are necessary to run. Particularly for the domain we operate in in HR, it's also a highly regulated industry because we're dealing with privacy information and individual information of individuals and you don't want to just send that type of information to general kind of GPT.

So when we started the company, we said, okay, we see a lot of these technologies. Could we bring it together to build the world's most robust HR copilot, if you will? And what Retrain is building over the last three years is a talent intelligent platform intended to get incorporated into the HR tech stack.

We don't want to replace the applicant tracking system. We don't want to replace the human resource information system. We don't want to replace the human capital management system. We don't want to replace the digital classroom or the learning management system, all of those systems are great, but they're great at what they were designed for, which is keeping records, running a digital classroom or keeping an employee record database or taking track of a long applicant pool.

Right? What these systems have not been designed for, and therefore they're not very good at or they don't do it at all. They're not learning systems. They're not decision support systems. They don't learn from examples. They don't make recommendations. They don't deduce. They just keep records. And the opportunity, now that we've been exposed to the power of AI, is to develop a set of technologies that use AI in order to help the humans do what humans do best.

Just like the tip I gave you about digital marketing, the same is true for HR. Within HR, think about every large organization. There's a big HR department and that HR department is in charge of a set of core processes from hire to retire. Thinking about which people the organization needs to hire, where to find those people, how to hire those people, how to conduct interviews, how to generate an offer, how to make sure that the company benefits are in line with the level of seniority, how to create an effective onboarding plan, how to create a continuous learning plan, how to create career passing, benching, retirement, exiting.

Every step of the way could benefit from those same techniques of amassing large amounts of information, learning from past patterns, being able to make recommendations, learning from mistakes, etc, etc, etc. This is what AI is very good at. And in HR, there are multiple opportunities across that continuum to let machines do what machines do best, which is to gather large amounts of information to take stock of all these past examples and to make recommendations.

So this is what we're building. We're building an HR copilot. It's going to make those systems of records smarter, more intelligent. And we look at intelligence and the two meanings of the word intelligence as in intelligence gathering, being able to just look at all these vast amounts of data that today are in separate systems, and look at them in a common language of skills, and intelligence in the word, in the meaning of it's intelligent, making intelligent decisions, so that for Paris or Shay, we can make career recommendations, or for Paris or Shay as candidates for a new job, we can make a career transition recommendations or help employers decide whether they want to hire you or hire me or hire somebody else and whether they even want to hire because they're hiring for particular skills.

 (30:41) Paris: That's fascinating. So you're kind of building a layer that's going to sit on top of the classic HR tech stack,your applicant tracking system, your LMS, your HRIS, and it's going to ingest all of the data from those systems and then be able to learn on that data. And really create a ChatGPT like experience for HR employees where they can say, give me some recommendations in our candidate pool for who might be a good fit for this role, or help me to redesign our benefits scheme or how can we build a better career path for someone who's in this role?

Am I on the right track? Is that how people are going to interface with the Retrain?

(31:27) Shay: Yeah, absolutely. And you know, we can't share any visuals today, but of course it's very visual, but I think there are two faces to that same technology. One is the employee or candidate view and one is the HR view. So the experience for it, candidate for a job, it's very simple.

You could start with some QR code. You apply for a job. The system allows you to basically generate your skills passport or skills profile by either answering very simple questionnaire or just uploading and scanning your LinkedIn profile or your CV and then seeing whether the machine actually got it right and you have a chance to correct it and and help teach the machine a little bit about you.

And then it automatically matches you with jobs. If you're a candidate or matches you with career opportunities from marketplace, if you're an existing employees, that's the employee experience, a very employee centric control of your skills, destiny, control of your skills passport so that you can help expose the best skills that you have so that you can find opportunities, whether you're entering the job force or entering a company or making career transitions and enhancements within existing organization.

The flip view of that is an inaggregate basis. So for a few, an HR manager and you need to hire a thousand engineers, how close or far away are you to those? Are you willing to drop certain skills? Because you understand you're going to have a hard time in a world where we said there are two open jobs for every candidate.

Are you willing to drop some of those skilling requirements because you're going to do the skilling internally, for example. So that allows you to see dashboards and benchmark where do you stand against the industry in terms of developing those skills. What could you learn from your competitors that those questions you're not asking because there are skills you're not hiring for, and maybe you can learn something from the analysis of the market.

So at Retrain we've analyzed tremendous amounts of data. coming in from private and public sources, including all the available data through government websites and publicly available data on sites like LinkedIn and then companies that give us data so that we can help them in a give to get model. The upshot of that is that we can, on one hand, create a very candidate and employee centric experience, but on the other hand, give dramatic insights to the HR professionals, primarily in three areas.

The first of which is planning. Through a skills architecture model that allows an organization to ask the basic question, which is, which skills should I be hiring for? Are those skills grouped into tasks and roles that can create a strategic process plan? That's area №1. Area №2 is in talent acquisition.

Now that I understood which skills I'm hiring for, let me actually hire, create requisitions, connect with my ATS, create those job openings, but do the entire process based on skills, whether it's based on title or educational background. And №3 is that now that I've gotten these people into the organization, but I'm still missing some skills, how do I make sure that my strategic upskilling requirements as an organization level are matched with those individuals that actually want to learn and upskill and re-skill inside my organization? So that's about benching, career passing, creating individualized training pathways, etc. So those three areas, three different models in our software, and we sell those.

Is there individually, is there a package to organizations that are interested in undergoing that skill-based organization of churn?

(34:40) Paris: So let me make sure that I understood this. Your Retrain does have a employee-facing interface and it has an employer-facing for the HR teams?

(34:50) Shay: Correct.

(34:51) Paris:  But is your core business model to, is the customer, the HR? The HR user, is that the paying customer? Do employees pay for this?

 (35:00) Shay: The employees pay nothing, candidates pay nothing. Two types of licenses, license or technology. Is there a government organization in every country, there are organizations that are Department of Labor or vocational training centers or people from kind of a government or NGO sectors that are in charge of developing fully-functioning, productive economies. So for them, it's about the national level, creating training programs, being able to make sure that young people can enter the workforce and understand where vocational opportunities lie. And if not, what training opportunities lie. We did a big project with the government of Israel, for example, called Skilayel, where we've helped more than 200, 000 people create a profile of themselves, understand skills gap and get directed to one of 5, 000 training programs, many sponsored by the government so they can enter the workforce higher on the skill that, for example, so hundreds of thousands of people are really using those systems. They are shifting their skill link to the language of skills.

They understand what are the skills in demand. And we've connected that with some government data about demand, open jobs and salaries within those markets so they can take informed decisions about skilling, as an example. So that's one type of customer, kind of a government, national level assistance, but our most prevalent customer are enterprises.

These are organizations that understand that in a world where there is war on talent, the methods of yesteryear are no longer relevant. They cannot get enough talent. Remember what I said earlier, we started a company because we understand there's a new normal. There's not, the problem is not that there are not enough jobs.

The problem is that there are not enough people with the right skills to fill those jobs. And every CEO and every CHRO and any global 2000 companies understand this, there are areas of their business where they cannot find enough talent. We hear that time and time and time again. Many of those areas are specifically related to what we also discussed, which is AI automation.

Their tech departments are starving. People don't know this, but at a big bank like a Goldman Sachs, there are more programmers than bankers, right? All of these kind of traditional companies are becoming technological retailers that are moving to e-commerce. Healthcare companies that are moving to digital health, all of those companies need those 21st centuries.

They're having a very hard time finding them. So they start doing it and they understand that they need to change their ways and they hear about the idea of the skill-based organization and their first reaction is, let's use the tools we've always used, which are sticky notes on the fridge in the cafeteria.

Well, that doesn't scale. So they move to Excel. And they sometimes hire a BCG or a McKinsey or somebody like that and they do a 2-year project to do the planning and say, let us bring this organization to its 21st century. And let's generate this big Excel that's going to show us what we need to be hiring for.

And soon enough, just like the sticky notes in the fridge in the cafeteria, the Excel breaks. Does it break at 500 lines? Does it break at 1000 lines? Does it break at 2000 lines? Does it really matter? Most of these projects fail because by the time you're done filling a thousand lines in Excel about what you want to be hiring for, it's already too late.

You've got to start that work all over again. And many of those organizations understand that they need to automate the process of understanding automation. Then they come to us, they come to one of our competitors, understanding, oh, we're in an organization, a skill-based organizational journey.

And we need a skilling platform to allow us to better plan, to better hire, to better retain using the language of skills.

 (38:32) Paris: You mentioned the role of government and the imperative of governments to take some degree of responsibility to retrain and upskill society for these new skills that are going to be needed in the future of AI. I also believe that education institutions like universities also have to some degree they have a ethical imperative to do the same thing. And they also have a huge opportunity to do that and one thing that i've always been frustrated with over the last I don't know 15 years that we've had this agency is that the educational institutions never did prepare, really prepare their marketing students for real digital marketing skills.

They gave them a lot of theory, but not a lot of practical know-how and how to work within platforms like Google ads and Meta and to do things like SEO and keyword research. These are a lot of the fundamental skills of digital marketing.

And I think that higher education dropped the ball and is still today dropping the ball in most places. Are they going to continue to drop the ball on AI skills or?

 (39:47) Shay: So first of all, I share your frustration. And I think that what you're feeling in digital marketing, the same is true in programming. You know, a programmer out of a good school after a four year college degree in programming doesn't even know how to start. You know, they maybe understand computation and complexity, but they don't know how to build a website.

And the same is true for lawyers that, you know, finishing law school have to spend now six months just passing the bar exam because they can't lawyer at all, etc, etc. So I think that's true. That in itself doesn't say that higher education is failing, but I think that the gap between the real world application, or if we want to discuss this in terms of skills gap, the gap between the skills that are necessary to participate in the economy and what people learn in higher education is widening instead of shortening. And in that sense, I think that higher education is due a dramatic shift. There are a lot of voices within the industry and within higher education that understand this. It's not that you and I have a secret. They understand it. There's been a lot of talk about micro credentialing and about small degrees and micro degrees and digital certificates and a whole slew of other things, solutions.
I don't think anybody found the pixie dust. Yes, there was tremendous amount of excitement about digital platforms like Udemy and Udacity and Coursera, and some of them have done a better job than others. Udacity, for example, has started something very promising was just sold to Accenture, actually, a few weeks ago, and Accenture was going to turn that into its AI training platform. But they bought it in some sort of fire sale. As a business, it didn't really work. It didn't prove that it could offer an alternative to higher education that was actually business viable. Others like Udemy and Coursera actually had a better chance at doing this while working with enterprises to be able to do this.

So I think we're going to see a lot of pressure, both internal and external on these institutions. And I think that sadly for some of them, not all of them are going to survive because one of the things that digital platforms allows you to do is concentration at the top and kind of a winner takes all mentalities and kind of economics.

If you can take your digital marketing Master's degree at Harvard, why would you want to go to your regional college? Well, maybe cost is one factor, but if government is maybe going to sponsor that, then what's going to happen to all those middle of the road colleges, middle of the road schools that are maybe not so famous for quality and historically have been locked in with their audience because of regionality or geography.

What's going to happen to them? I think that's a very big question open right now in higher ed.

(42:19) Paris:  Yeah, so that's a great answer. And I agree with you that there's going to be some degree of reckoning within higher education, and I think it's only going to accelerate. These conversations have been happening for many, many years. And now that AI is on the scene, the pressure for higher education is that, as you said, that the gap is getting wider, not more narrow. I think that there's going to be a shakeout and the low end of the schools are probably, I mean, it doesn't make sense to pay those types of tuitions anymore.

(42:52) Shay:  Definitely, definitely not. If you take my earlier assumption as true or somewhat true, which is that people are gonna have to retrain so many times since their career. I've seen some statistics that on average people entering the workforce right now are gonna have to make between 10 and 15 career transitions.

And there's an old joke that no Jewish grandmother ever wanted her son or daughter to be an Android developers. They all wanted them to be doctors, doctors and lawyers. And, Android developers and soon enough prompt engineering are probably better paying jobs.

 (43:27) Paris: Yeah. That's always a fun exercise to try to explain what you do to your grandmother, whether it's digital marketing or in AI. This has been fantastic, Shay and I think we could go on a lot longer, but probably we need to wrap this up and maybe have you back on again.

 (43:46) Shay: Absolutely. Well, it would, this is a fantastic conversation and I'd be very happy to come back.

(43:51) Paris: Yeah. As we wrap up, is there anything that I didn't ask you that you wish I would have asked? Or is there anything else that you feel would benefit our audience around this topic?

(44:01) Shay:  I think we covered a lot of ground here. So, thank you for all those great questions. And if people want to learn more, we're always happy to talk directly so people can look us up at Retrain.ai. and my email is shay.david@retrain.ai. Always happy to learn from the audience and take questions and explore these themes.

(44:23) Paris:  Great. I have just realized that Shay I've been mispronouncing your name the whole time. So my apologies.

(44:28) Shay: No, no, you know, Shay is a Hebrew word. It means gift in Hebrew. But my Hebrew name is Shay David and my American high school has been going by Shay David. So high school was interesting.

(44:40) Paris: We should have clarified that and my apologies, but Shay it's been a real, it's been a real pleasure. It's been enlightening. I think you, you have a tremendous vision for the future of both of AI and how organizations need to reskill for this future, and it's fascinating to learn from you. So thanks. Thanks for coming on the show.

(45:00) Shay: Thank you. Fantastic, Paris, and we'll talk soon.

(45:03) Paris: All right. 

(45:05) Outro