#120 Data Trends & Predictions for 2023

Hi, I'm Maggie, VP of Curriculum at datacamp, and I know firsthand how overwhelming it can be to keep up with the rapid changes in data and AI. But staying ahead doesn't have to be hard. That's why our curriculum team has curated the best learning content to help you and your team stay ahead, covering everything from data and AI literacy for busy executives to cutting edge generative AI for developers, and everything in between. Whether you're upskilling yourself or your team, we make learning practical, engaging and flexible. It just works. Power your data and AI transformation journey. On datacamp.com. You're listening to Data Framed, a podcast by Data Camp. In this show, you'll hear all the latest trends and insights in data science. Whether you're just getting started in your data career or you're a data leader looking to scale data driven decisions in your organization. Join us for in depth discussion discussions with data and analytics leaders at the forefront of the data revolution. Let's dive right in.

Welcome to Data Framed. This is Ritchie. I hope you're all feeling suitably rested after the end of year holidays and are looking forward to 2023. It's shaping up to be an exciting year for data science and for datacamp. So today I'm joined by two of the founders of datacamp to make some predictions about what's going to happen. Joe Cornelisson is datacamp's CEO and Martin Thurwissen is datacamp's coo. They've spent the last decade worrying about how the data industry is evolving, and so they're in a prime position to identify some trends. We're going to be talking about data skills and jobs and tools, along with some sneak previews of what's been worked on at datacamp. There's a lot to cover, so let's dive right in. Hi there Joe and Martin. Welcome to Data Framed. Thanks for joining us today. So, just to begin with, do you want to tell us a little bit about what you do at datacamp? Joe, what does being a CEO involve?

Lots of things actually, and it's one of the reasons I love this job. It's been changing every year. We started the company nine years ago. Right now my main tasks are focused on setting the vision, hiring the right team, and then making sure there is enough money in the bank to actually execute on that vision.

And Martin, what does being Chief Operating Officer involve?

So also a lot of things actually, but I'm actually mainly active on everything that's doing our commercial side. Lots of interactions with our B2B customers, those who use Datacam to get digital transformation going, mainly in the data field. So doing lots of work with those and working also with our media team of which this podcast is a part of. So in a way this is an operation for me as well.

Always happy to have your support on the media team.

Thank you.

So with that out of the way, let's get into some predictions because our core audience at DataCamp is data scientists. I'd like to start by just chatting a bit about the data scientist role, whether you have any predictions for how this role is going to evolve in 2023.

So I think my mental model for the data space in general is somewhat similar to the software engineering space, but then it kind of delayed by 10, 20 years. So I think what you'll see in 2023 is further specialization. Whereas five, 10 years ago there was kind of one role, the data scientist that became really popular, made it as a secondary role, the data analyst. I think what we've seen in the last couple of years, and I would expect that to continue its further specialization as the space matured. So think about new roles like the analytics engineer, that's kind of sitting between the analysts and the data engineer on the engineering department, that's now becoming a role in and of itself. So for those who are unfamiliar with the concept of an analytics engineer, their role is essentially to create clean, easy to use data sets for then analysts or other people in the organization to use and kind of self serve ideally. And that's a role that I think is fairly new and didn't really exist until more recently it became a specialization. But it's just one example of a general trend of specialized roles within the data space. Another example would be an MLOps engineer, somewhat similar to how in software engineering you have traditional software engineers, full stack software engineers and you have DevOps engineers. I think in the machine learning space there's a similar trend where you have the machine learning teams themselves, but they need kind of an infrastructure and operational team and that's kind of where the MLOps engineering role comes in. So those are two examples. But the bigger trend is really specialization.

So I think a lot of data scientists are going to be happy about that because the space has just become so huge that learning absolutely everything is impossible at this point. So having some sort of sense of specialization and just having a more niche field seems pretty helpful for a lot of people in this area of employment.

Another role that I wanted to add, and it's maybe out of the data field, but actually, actually what I tend to see is that they're more and more within L and D departments. They're getting dedicated L and D data people. So people that are responsible solely for getting the digital transformation, the data transformation, going through educational aspects. When we started Data Camp nine years ago or even five or three years ago, that role didn't really exist. And I hardly encountered it when talking to some of our B2B customers. And now you see that more and more companies are incorporating in IT and dedicating someone like just to that. So not a typical data role, but definitely something that they can describe as a data role.

Absolutely. And talking about digital transformation and data transformation is a recurring theme on the Dataframe podcast, so I think that is an increasingly important role as well. On a related note, we talked a lot about data skills just then and how the roles are changing. Do you see any changes in who needs data skills in 2023?

Yeah, so maybe a little bit of context. Our vision has always been that data skills are not just for the data scientists, for the kind of nerds, the PhDs in physics. Data skills are essential skills for everyone in an organization. And kind of the level or the types of data skills people need are just different, depending on their role. I think it's fairly obvious if you're a data scientist, there's things like RPython, there's statistical models, there's machine learning, but if you're on a marketing team, there's a whole set of skills around a B testing and things like that are very useful to know as well. Or if you are on a finance team, there's a similar set of skills that are very useful. And so what we've done at datacamp is we've developed a data literacy curriculum that's suitable for the masses and for a large group in every organization, because there is concepts like data security, data governance, where the more people in an organization are aware of them, the better kind of data transformation will be executed upon. So it's not just one set of people who can benefit from data skills. I think it's a very broad trend, a little bit similar to how when the Internet came about, everyone had to kind of upskill and gain a new set of skills.

So this goes a little bit back to what you were saying before about specialization and different people are going to need different sets of data skills.

Absolutely. Yeah.

Nice. Martin, was there anything else you wanted to say on the subject of broad data skills?

No.

I think data literacy will go mainstream. I see that more and more, and I think that trend Will only, like, continue 2023 and much beyond.

Nice. So one thing I've noticed personally is that younger employees, they tend to be very open to the idea of using data. It's often been pushed in school or university, but it tends to be the older employees who, you know, they've been doing job for a while, they've got a bit of intuition and they tend to want to use gut feeling more than using data. So I would never suggest this of YouTube, but it's quite often that the senior executives that tend to want to just stick to gut feeling, they're a little bit more skeptical about the use of data. So have you seen any progress in terms of getting senior management to adopt the use of data?

Totally. And I think we may have a little bit of a biased sample here because the people that contact datacamp, it's usually when they're already kind of at a certain point of data maturity in their thinking and in their execution potentially. But there is a general trend where a lot of executives at organizations now are very well aware that it's crucial to invest in data science, in AI, but also as a starting point to invest in data skills and shifting their culture to be more data driven. So I do think that that is a general trend in the market and I don't think that's going to change anytime soon. If anything, it feels like it's only accelerating.

So just related to that idea you talked about investment in data, so do you see any changes in how these data investments being made?

That's a great question, because I think if you look at the economy right now, things are cooling down. Things are cooling down quite fast. That being said, I don't think that's going to shift necessarily, the fact that there are more investment in data and AI. I think what it will do is it will shift how people invest in data and AI. So to be specific, whereas I think a year ago a lot of the investments were probably around how do we gain more insight, how do we deliver better customer experiences? That will remain important, but I think the investments definitely will shift towards how do we ensure we can understand the ROI of different areas in the business and how do we ensure we use data and AI to actually make the business more cost effective? So the general trends will continue, but how people invest will be different, I think, in 2023 compared to the previous decade, where we've seen generally much better macroeconomic circumstances.

Yeah, cost saving does seem to be on everyone's mind at the moment, for sure. So I'd just like to move a little bit outside talking about data at work. Because you mentioned before, both of you about how data skills are becoming very widespread for everyone really. So can we just talk a little bit about maybe data outside work and how do you see data sort of further permeating our society?

I think a good example in this area is like all the stuff that we do with classrooms and universities and the programs that we have there, like where we give university professors like six month access to data camp for free to use it for their students. Like things that we see is that more and more schools and universities as well as like they're nonprofits now that are focused on this as well, start to do programs for data skills around data skills. Like I wouldn't say like there's this major shift going on where like there's a very structured approach behind it, but you see there's these grassroots projects and they're coming up more and more where universities professors take initiative to rethink about like, okay, data is this really important thing. It's not going to go away. It doesn't matter if you're studying marketing, finance, HR or like data science itself. We need to do something with it. We need to develop the skills for it. And we see like lots of pilot projects going on. We see in our data that they're taking lots of programming courses, like in areas where you wouldn't even expect it. Like I gave the marketing example, they're upskilling themselves in Python. There's this shift going on there and I think the innovation is happening at the base. It's not being pushed down. That's great to see.

That's really cool that people are actually kind of wanting to learn data skills in these. A lot of like kind of a small scale grassroots projects as well as sort of more sort of high level strategic big projects.

I wanted to add to that. I think there's definitely a positive shift like Martin said. I do think it's important to realize the traditional educational system tends to shift slower than what Martin was describing. What we've seen compared to five, 10 years ago when we started a company is that there's much more focus in a lot of university programs on data skills. And so that's really encouraging to see and we've been very supportive of that. Where I think there's still a huge shift that is starting to happen, but is definitely not there yet is at the high school level. There's definitely still a big focus in a lot of high schools of things like calculus. And I think in Today's world, it would be amazing if we can shift that to data skills, or at least in addition to the skills that are currently taught. I think it would be very valuable for people to gain a basic level of data literacy while they're in high school. And that is something we would be excited to support even more in the future. But we've seen mostly traction at the university level and much less at the high school level right now.

That is pretty interesting because you don't necessarily need to be 18 years old in order to start getting some data skills. Certainly a lot of the stuff around descriptive statistics, data visualization, this is something that almost any moderately smart teenager can understand.

Absolutely. And the market needs these refills. Right. There's still a supply demand imbalance. There are more jobs for data professionals than there are people who are qualified for those jobs. So the market really needs those people. And the earlier we can upskill folks, the better. Even at a country level, it can become a competitive advantage. There are some countries that are definitely ahead of the curve in integrating data skills in their curriculum, both at the university level and the high school level. And then there are some countries that are definitely behind in investing in data skills of their younger population. Yeah.

Let's hope that 2023 brings some changes to the educational curriculum for school kids. All right, we've talked a lot about data skills, so let's change it up a bit and talk about tools. So it kind of baffles me that even now, Excel is the world's most popular data analysis tool. So there are hundreds of different tools around for analysts, for data scientists, and other data professionals. Can you maybe talk us a bit about how the tooling landscape is changing?

Maybe a little bit of a disappointment to you, Richie, but, like, spreadsheets are still very popular. I don't think that will change in the next couple of years, but I think it's definitely true that market share is moving away from that. And I think we see two types of directions. On the one hand, there's the low code, no code movement. If you think about platforms like Power Bi, like Tableau, so the business intelligent platforms, like, we see that more and more that they're taking more and more share. And more and more companies find it important that their employees know how to work with these tools. And they're doing quite a lot of investments in making sure that people, like, know how to use these tools. So it's definitely a change, like, compared to a couple of years ago. Like, there's really, like, a transformation going on and the amount of people that get access to it, the type of data they get access to it, all thanks to these low code and no code platforms. And they can do a lot more in those than they can do in Excel and a lot easier. So I think that's one movement away from spreadsheets. I think the second thing is around notebooks. Notebooks are already pretty popular with data scientists, data analysts. I think a really good example there is Jupyter notebook. The way that they're used today is still very much within that space of data scientists and data analysts now. So they go in there, grade some stuff, they do their analysis. Now I think we're moving towards a situation where also this notebook is getting more and more accessible by the non data scientist and the non data analyst and the non machine learning engineer in the company. I think it's going to democratize the workflow that you see today. Like for example, like we're building tool called Workspace and what it does is like it does not only allow like the data scientists analysts to like very quickly do their own analysis, it allows it to share very easily, it allows for people to comment on it, to make their own edits and that like lowers that barrier to go in and start like doing your own stuff using a notebook. Maybe just to summarize, like, I think that what we'll see is that next year notebooks are going to go more and more away from just a tool for data analysts and scientists to like, a tool that's going to be used more and more by data practitioners and we're going to get more and more data practitioners in the world given like how important everyone starts to see it.

I want to second that last point. I think it's kind of crazy. If you look at the design space, you have tools like Figma that really leveled the playing field and basically brought product managers, engineers, everyone kind of in a central place to collaborate around design tools like Notion, who've done that for general documents. But for data professionals there are no standards tools yet that really have spread wildly that kind of do this. And I think it could be super, super impactful and it's almost certainly going to happen in the next few years. For some reason the data space has just been behind other areas here.

Yeah, it does seem like there are many tools, but there's no one sort of a standard tool.

Yeah, and that's an excellent point, Richie. That's one of the things that makes collaboration so hard in the data space. It's like there's a Huge fragmentation of tools. And depending on the data literacy level of the individual, they tend to use different tools. So people on the BI team, they use certain set of tools. People on the data science team use a certain set of tools. People on machine learning team use a different set of tools. And that fragmentation makes collaboration really, really tricky. And I think there's going to be a trend towards kind of standardizing on easier to use tools that really put collaboration at the center rather than specific use cases.

That's interesting. So maybe we can talk about this sort of market consolidation a little bit. I really like Martin's point about how there are sort of two different ways this is going with the business intelligence tools and also the notebooks. Maybe we'll take business intelligence first. So, of course, with datacamp, we've been focusing on Power BI and Tableau in the BI space just because they're the market leaders. Joe, do you want to talk a bit about how you see this changing?

I'm not an expert in the design space, first of all, but I like it as a comparison where Adobe has this whole suite of kind of creative tools for designers and other creative professionals, and you can think of that similar in the data space, you have Power BI Tableau. Those tools are not going to go away. If anything, I would expect them to gain strength, especially Power bi. We've seen an incredible surge in demand for training on Power bi. It speaks to the fact that once Microsoft starts pushing something, they have an incredible power to set a new standard. All of that being said, I do think in parallel, there's going to be another shift towards collaboration tools that enable not just the analyst and the BI team, but enable a broader set of people. At least that's our expectation that both things will be true. So on the one hand, you're going to see continued dominance of some of the BI players and they'll continue to grow. On the other hand, there's going to be a shift towards kind of cloud native tools that focus on collaboration, that enable a broader group of people.

Interesting. Martin, is there anything you want to add to that?

Nothing in particular, but I think what's worth pointing out is that the BI tools also become more and more powerful, like their ability to connect with data that's stored in the clouds, their ability to make reports available or the dashboards available to the entire organization, like not even on just desktop, but on your mobile phone, so that people are maybe in the field working like salespeople can check them. So it goes way beyond, like what you think of, like if you think of an Excel, you think of a spreadsheet and that's it. And you open it up on your laptop and you need to understand a bit like the opportunities that BI gives you, like, but they're building on top of it and working on it, like, goes really far. So I'm really bullish on what's happening there.

Yeah, I definitely agree with what both of you are saying that, like, well, okay, there are tools for analysis already, but the tricky part is the collaboration between team members and sharing things with people who are maybe less technically literate. So that's a sort of a big thing for business intelligence tools, is they get more powerful. They, you know, make it easier to collaborate and share and things like that. How about the case with Jupyter Notebooks? So, of course we've got datacamp Workspace, which is our own hosted notebook, but there are a few other players as well. How do you see the notebook space evolving?

Yeah, so maybe to give a little bit of context, because this is one of the most exciting changes in the space, in my opinion. So Datacam had a unique view on the space in the sense that our students kind of come to us and tell us, like, hey, this is where we want more content. And one of the areas was just local setup. A lot of our learners continue to struggle to kind of download all the packages, set up the jupyter notebook environment. And it's a real kind of unnecessary friction to kind of transition from learning to doing data science. And there's several companies now investing in cloud native notebooks on top of the jupyter ecosystem. Typically, some of them are completely rewriting it. Some of them are really building on top of Jupyter Notebook, and I think that's going to be extremely powerful because it takes away the friction for people to get started with notebooks. It enables easier collaboration. And I'm very excited about kind of all the innovation that we're seeing in that space and different people taking different angles. Some of the tools like Hexer, seem to be really designed for the absolute expert data scientists, and then there's other tools that are kind of in the middle. And then our approach is really to focus on making it really easy to get started, really intuitive to get started and appeal to as broad of an audience as possible. But all of the innovation is, I think, really going to change the way data professionals work in their day to day life. A little bit similar to this has already happened to the software engineering space. If you look at how a lot of engineers do Work today, it's very different from 10 years ago and their whole workflow kind of shifted from local environments to a lot of it happening in the cloud. So I think with notebooks, that's the big shift we're starting to see. A lot of individuals are kind of exploring to shift. A lot of organizations are still kind of in the old world here, so that's going to be exciting to see that shift happen.

Absolutely. And I do think the software development space is a really good sort of, like, leading indicator of what's going to happen in Data and having all these tools move to the cloud. One thing you mentioned, though, was saying that a lot of, like, new users, they really struggle with getting all the software installed and whatever. I have a confession to make because I've been working in data science for nearly 20 years now, I still struggle with Python package management. It's horrible. So it's not just the new users that sort of need this.

It's one of the hardest problems, I think, in the data space. Package management.

Absolutely. It's just like getting the right version of Pandas going. Cool. Do you want to talk a little bit about what's happening with Workspace over the next year?

Yeah. Just to give you a sense of where we are kind of in our journey of developing Workspaces, a cloud native collaborative notebook. I think the product today is already really good in educational context. So we see a lot of learners use it to kind of do projects to build their portfolio. The portfolio they then show up on their profile and they use in their job search. And I think the next year is really about like, perfecting that use case first, first of all, but then also enabling the team's use case, where right now I think we have one team actively using Workspace, which is Data GIMP itself. Our data team is very actively using Workspaces, similar to how the design team is using Figma. So they're pulling in other people in the organization and we're kind of building that use case and perfecting that use case and that we want to do a big marketing push to shift a lot of our teams users to not just use datacamp for learning, but also start using it to get work done. And that's a really exciting kind of transition we're going through as a company.

Absolutely. And certainly there are very different needs from, okay, I'm just going to write some analyses in a notebook myself to I'm a company, I've got to manage like thousands or tens of thousands of different notebooks and have them all findable by the right people and have everyone have the right permissions and know what's what.

Exactly. So there's a whole set of features that you'll need as a team that are less relevant for an individual or for a professor who's using it in a classroom setting. Because that's something that's increasingly starting to happen already with the current version of the product.

All right, moving away from tools, one of the sort of biggest stories in data in 2022 was in generative AI. So we're talking tools like ChatGPT, we've got stable diffusion, we've got Dall E for generating text and for generating images. So how do you see these tools having an impact in 2023?

One really interesting one is like, because you talk about GPT3, like one interesting descendant of it is like the OpenAI COD and in essence like what it does is like it auto generates code and the claim is like, thanks to that, like as a developer you can focus more on the high level problem and then focus less on like the syntax and maybe your Python package management hopefully soon. But they're actually like already like real world, like use cases of that. Like for example, like GitHub Copilot uses it and they released some interesting stats where they said that the software developers that they saw that use GitHub copilot, like they saw like a 40% increase in productivity. And that's an important number. It's an incredible number because it shows how much productivity gains there can be made. And it's also like counterintuitive because I think we always thought, like, it's going to be the technical roles, the creative roles that are going to be least impacted by AI and it's showing that it might even be the opposite. If you think like this OpenAI Codex, like makes software engineers 40% more productive and maybe over time, like even fully replace them. Like, don't think that will happen. But like to a large degree, if you think about Dall E and all the creative people and like being able to auto generate your brand logo or the design of your marketing email, I think all these, like, what's coming at us now, like, is going to extrapolate in 2023 and like it's going to be super interesting because I think that the GitHub copilot is just like scratching the surface on like what is possible. And it's interesting to see like how it's going to increase productivity, how it's going to lead to wider adoption, because all of a Sudden you don't need to know any nitty gritty detail about the syntax so it becomes more accessible. That's one of those things that I'm really excited about.

Yeah, Certainly just is having an impact both on the creative side and on the sort of more technical side, especially once you have the code writing capabilities. Joe, is there anything you want to add to that?

Yeah, I just want to second, I think it's one of the most exciting innovations in this area in probably the last 10, 20 years. I think this is really a game changer. These large language models like GPT3. Rumors are GPT4 is going to be even better, significantly better. And so I think this is going to be a massive, massive shift where we're only at just the beginning. I think some practical advice for people maybe is this is going to make it even more important to understand conceptually what's happening with certain statistical models, machine learning models, like Martin said that the syntax part is still going to be important, but it's going to be a little bit less important in the future. And I'm also excited how this is going to enable us to build better educational interfaces, quite frankly. Yeah. I think what datacamp has done is we've innovated on kind of the didactical approach by creating this kind of automated feedback system. As people progress through courses, these models that are being developed now are going to enable us to get much closer to what our initial vision was, which is we want to automate the instructor, we want every student to ultimately have their personal tutor. And I think we're now kind of seeing the technology innovation that will be required that will enable that use case where eventually every student will have their personal AI tutor. And I think we don't even know how much impact that will ultimately have on society because very well established, like private tutoring is probably the best way to, and the most effective and the fastest way to learn anything. But a lot of really great private tutoring has historically only been available to a very, very small group of people in the world. And this is going to open up like the best type of education ultimately to the entire world. So I think it's incredibly exciting and it's really going to change every single sector eventually.

Absolutely, yeah. We're just sort of scratching the surface of impact at the moment. Actually, you mentioned AI tutors. There's a really great sci fi book called the Diamond Age by Neal Stephenson. It's about a girl with an AI tutor. So definitely worth a read. Total side note, but one thing you Both mentioned was the idea that learning concepts is important. And I do agree with this. So even if you've got an AI that can write your scikit Learn, Fit and Transform method code for you, it's not going to tell you what the model means or how to interpret it. So that sort of idea of understanding what your model is doing does seem incredibly important. Moving on, I'd like to talk a little bit about the jobs market. And so at the moment in many countries around the world, the worries about inflation, worries about possibility of recession. So is data science still an attractive field to get into?

So we're obviously not completely unbiased here, but I think if you look objectively, the growth expected by the US Bureau of Labor Statistics for data roles is 36%. That's way above kind of almost anything else. If you look at Glassdoor like top 50 jobs in America, it's currently number three. So I think that data scientists and data roles in general are still one of the most kind of sought after or one of the most lucrative and popular jobs to go after. That being said, we will see in general because of the economy cooling down, the interest rates going up, I do think we will see a kind of slowdown in terms of salary increases. We've gone through a decade of an incredible job market and a fast seller increases in many roles. So that's probably going to slow down a little bit. But I think within the potential jobs that people can choose from, data roles are definitely still going to be all the way up there in terms of future potential value for both the individual as well as for the organizations that end up hiring these folks. In the short run it's maybe worth saying like some sectors are going to be more resilient. Datacamp has now a recruiting product where we help organizations hire the right data talent and we are seeing some, still some real strong traction in industries like healthcare, for example.

Yeah, I think there's also another angle of like data science and the data scientist job, but there's also the angle of data science as a must have skill to get maybe a job in marketing or in finance or in healthcare. And I think if the economy will cool down and, and it will be more competitive to get certain roles or jobs, I actually think that having that data knowledge is going to be a differentiating skill and that more and more organizations will ask for it. If you look at the other investments that they made and we talked about bi and we talked about some of the other technologies and the notebooks and so on, having it as A distinctive skill is going to give you an edge. At least that's my belief, given everything else that's going on.

Yeah, that does sound very plausible. I mean, certainly at datacamp, you mentioned the example of marketing, and some of our people in the marketing team are very creative focused, some of them are more analytically focused, and there is that kind of divide. And so I think having those data skills to go, okay, this works, is a good competitive advantage there. Are there any other things you think people can do in order to help them get a job in a tough climate?

Absolutely. I think this is going to be at least the next one to two years. It's going to be a job market where it's crucial to stand out. So I totally want to second what Martin just said around. Hey, even for non data roles, if you can show you actually have incredible data skills, that's a huge plus. If you're applying for data roles, I think it will be really important to stand out, for example, by having a portfolio. Just like software engineers often have a GitHub profile where you can kind of see the projects they've worked on, maybe they've contributed to some open source. I think for data professionals, it's incredibly important to have a portfolio as well, to stand out. You can, for example, use datagame Workspace and then add it to your profile. We're seeing an increasing number of students actually take advantage of that. There's certifications, so certifications from specific tools like Power Bi or others, and then data. Can we launch our own certification program as well where we're certifying whether people are job ready for the job of a data analyst or a data scientist? And all these things really help you stand out because it is definitely a crowded job market now where in the short run, especially for the more junior roles, we've gone from companies really struggling to hire for those roles to companies getting hundreds of applications for open roles. And so the challenge now has shifted from kind of finding the right talent at the junior level to filtering through resumes. And everything someone can do to really stand out is really essential, I think in the next few years.

Absolutely. I do like the idea of just creating a portfolio because if you can't show off your actual work, then it's very much a gamble on whether or not you should be hired, because an interview can only tell you so much. How important do you think credentials are in data science?

It's a tricky question. I think ultimately the market has shifted to really valuing skills. And so I think the most important Thing is having credible and fast ways to demonstrate that you're skilled. Credentials are one instrument to do that. You can get credentials from datacamp, Microsoft, other tech companies. There's definitely value in that. I think there's still value in kind of credentials from traditional educational institutions as well as boot camps. The challenge there is that, especially in the US and in the uk, that remains really expensive. So a Data science master's or bachelor will easily set you back tens of thousands or even hundreds of thousands of. And even if you say, hey, I'm just going to go through a boot camp, that will easily cost you 5 to 20, $25,000 as well. There's some companies that have gotten creative with the way they charge that money, but at the end of the day, it's five to $25,000, which is a lot for a lot of people. A lot of money for most people, frankly. So I think there's value in that. I think it's just a question of what's the ROI and for how many people are those things available.

Yeah, certainly you get a very long Data Camp subscription for tens of thousands of dollars. That's a good lifetime subscription there, I think. All right, so just going back to the idea of recession and potential economic problems, we talked a bit about what it means for people who are trying to get a job. How does it change things for hiring managers?

Yeah, I think for hiring managers, it might be easier and might also not be easier. So on the one hand, if you open a few, a data science role and you have zero applications, like, that's not great. And then you need to go out and hunt there. And I think there's been a little bit of situation in the past couple of years now. At the same time, like, okay, if you get like 100, 200, 300 resumes in, it's also not that great because then you need to sift through them and figure out, like, okay, who actually has demonstrated skills here versus who doesn't. And so I always think, like, okay, the ideal world is like, you get one applicant in, it's the perfect candidate, and you hire the person. That's actually a little bit what we're trying to do with datacamps recruit product. Like, for us, it's not about giving a hiring manager hundreds and hundreds of resumes in their mailbox. It's about like, okay, tell us what you need. And then we're going to try to figure out, like, okay, what are the two, three, four people that meet your needs? And we're going to try to connect you. I think that's like, we're moving more and more to a world where that's going to be valued. Not like finding the, the situation where you only have zero applications, but like the situation of like, okay, how do I find a tree for white people?

So it's really a question of search and filtering out all the candidates who aren't appropriate just to get the right person from a hiring manager's point of.

View and knowing the skills of these candidates and knowing what they have demonstrated. And that's where I think, for example, our certification is really cool because we know what these people have demonstrated and so we can link that to what they're searching for.

It's about a trend towards, I guess, increased transparency both on the candidate side and on the hiring side.

Yeah. So instead of like just searching, like give me everyone from Harvard University, like you can also search on like, hey, give me somebody who is in the top 5% of Python skills here. You can search on demonstrable skills.

Okay, so are there any other trends you're seeing or do you have any further predictions for 2023?

It's partly something I'm very curious about. In the context of these large language models like GPT3 stable diffusion, we don't have clarity yet. And where there's definitely going to be a catch up is the legal framework around some of these things where like the power of these models is ultimately resting on input and a lot of that input might be copyrighted. And I think this is so new that we just don't have a legal framework. There's all kinds of lawsuits that are starting to happen. I think it's really interesting how that kind of shakes out. And I think that could either be an accelerant or something that slows down that process. So the prediction I guess is like, there's going to be more talk about that in 2023. I don't know where we'll land on it though.

Yeah, certainly interesting times from a sort of legal and ethical point of view with all this generative AI and new sort of applications of data. All right, thanks, Martin. Do you have a final prediction for 2023?

Yeah, I think one thing that I think will become more and more important is making sure that your data is correct. We got a lot more people now introduced to the power of data analytics, data science. A lot more people get access to it, like across all the functions and departments, which is great. But we all know on this conversation how hard it is to make sure that what is presented is actually correct and the underlying data is correct so that the conclusions are right. And I think there's still quite a lot of work to be done on the data side to make sure that these data observability platforms like Monte Carlo XL Data Data Vault, like, they solve that kind of problem and that's fixed. I think there's going to be more and more attention to that.

Absolutely, yeah. Trust in data is such a hard thing. I do like a quote from Dennis, our financial data scientist, who sort of checks that all of our data gap numbers are correct. And he talks about doing paranoid data science where you have to thoroughly check absolutely everything just to make sure things are real. All right, so thank you both for joining me. I hope you've enjoyed the experience.

Yeah, thank you, Richie. This was really exciting to do.

Thanks, Richie.

You've been listening to Data Framed, a podcast by Data Camp. Keep connected with us by subscribing to the show in your favorite podcast player. Please give us a rating, leave a comment, comments and share episodes you love that helps us keep delivering insights into all things data. Thanks for listening. Until next time.