Skip to main content

‘We might be wrong, but we’re not confused’: how Tomer Cohen, chief product officer at LinkedIn, figures out what works best

We dive into managing the relationships between designers, engineers, and PMs.

Share this story

Tomer Cohen, LinkedIn’s Chief Product Officer, smiles at the camera.
Image illustration by William Joel / The Verge

Tomer Cohen is the chief product officer at LinkedIn, and actually, I talked to Tomer twice. Here’s a little secret about Decoder: we do the interviews, and then often, the guest and I just keep chatting for a while. So after my first interview with Tomer, we were hanging out, talking about the perpetual battles between engineers, product managers, and designers. And he said something that completely jumped out at me:

“We might be wrong, but we’re not fucking confused.” 

This isn’t a totally new line — it’s been floating around for a while, you can Google it — but you know I love an f-bomb, and honestly, it’s one of the most simple and clarifying things a manager can say, especially when managing across large teams. So I asked Tomer to come back and really dig in on that idea.

On top of that, we’ve been talking a lot about running social networks lately, and LinkedIn is a fascinating social network because it doesn’t have the same engagement-based success metrics as other social platforms like Twitter and Instagram. Tomer doesn’t care about time spent on LinkedIn; the platform is designed to be successful when people get new jobs. That means his ideas for features and user experiences are just really different.

The following transcript has been lightly edited for clarity.

Tomer Cohen you are the chief product officer at LinkedIn. Welcome to Decoder.

Thank you, Nilay. I look forward to our conversation.

Well, by the end of this everyone will have a new job. That’s the promise of a LinkedIn interaction.

That’s the goal. We have a lot of jobs to fill, so that’s a good outcome for this conversation.

Let’s actually start there. LinkedIn is a social network and it’s thriving. I would say we live in a time of great change for social networks, but LinkedIn is unique. What is the goal of LinkedIn? Do you sign in to it, eventually get a new job, and LinkedIn helped you do that? Is that the purpose of it?

I love that we’re starting there because it’s always great to understand what LinkedIn is for everybody. So first and foremost, LinkedIn exists to create economic opportunity for every member of the global workforce. 

I remember my first real interaction learning about LinkedIn when I came to the Valley in 2008. I went to a lecture about social networks at the Stanford engineering school, and all the buzz was about time spent on the internet and traffic. The founder of LinkedIn, Reid Hoffman, was on stage, and he talked about how LinkedIn exists so people can reach out to professional communities and get their jobs done, whether that’s accelerating their career, partnering, fundraising, or getting a job. That was kind of my first hook in LinkedIn. I was extremely impressed. Later on, Jeff Winter codified it into our vision statement. 

That’s the reason we exist. LinkedIn, at its core, is a community of professionals who come together to share ideas and expertise, but also to get and give help to each other.

Those ideas about time spent and traffic are still the core metrics of every social network and every web platform that I have the pleasure of talking to people about. You have to get people there, you want them to have a good time, you want them to stay longer, you want them to connect and get hired, or maybe you want them to transact — there are ways to transact on LinkedIn. Are those the core metrics for you for LinkedIn? What are the things that you measure that determine success?

“I have no idea what the time spent on LinkedIn is.”

I have no idea what the time spent on LinkedIn is, but I can tell you that we track the opportunities that are created on LinkedIn a lot. For example, right now on LinkedIn, every minute, eight people are getting hired.

How do you track that? So if I get a job on LinkedIn, do I tell you? Does someone else tell you? How does that work?

That’s part of the beauty of the system. We know if you saw a job on LinkedIn and if you applied for that job, and we can actually track if you change your profile later. You can think about it as somewhat of a nice closed ecosystem of seeing job changes. In fact, some of the great economic data that we get is from seeing how the economy is performing for people transitioning in their job, which is a good stat given the macroeconomics right now. 

A year ago the job change rate was 40 percent year-over-year. Now it’s declining year-over-year. We’ve seen massive growth in job change rate, to a declining job change rate, and we see it through the data in LinkedIn. So back to your data, this is just one aspect of where the LinkedIn economic graph comes to life.

You and I are talking on the day that the federal government released the jobs report. The number was higher than most economists and analysts predicted. Did you know it was going to be higher?

We always publish our workforce report in concert and we see a lot of those trends coming to life. For the most part, we actually share data. We share it externally, and we share it with members to show them how it’s performing on LinkedIn. 

Yes, there is actually a pretty strong signal on LinkedIn because we have 875 million professionals. It’s not the world’s 3 billion professionals—the whole workforce—but it’s a pretty good representative sample. We’re also seeing trends in terms of behavior, which I think tends to be some of the most impactful. For example, skills are becoming one of those tremendous behavior changes we’re seeing on the platform as well.

What do you mean by skills as a tremendous behavior change?

There is a tremendous change towards focusing on skills rather than pedigree. 

I think for many years we really believed in the notion that there was a primary focus for every professional, and believed in skill matching between an employer and employees. If I want to hire somebody on your team for a job, I would look first and foremost at their skills. For years, the industry has relied primarily on pedigree and which companies you work for. Now we’re seeing more and more that there is, in my opinion, a tremendous change towards focusing on the skills you have. 

For example, we have tools that we offer to recruiters, which are able to create a pipeline of candidates, to source, and to reach out. We’re seeing close to almost 50 percent of recruiters using skills right now to look for professionals. That’s a big deal. That wasn’t the case before, and it’s increasing a lot.

On the other side of the marketplace, we’re seeing many professionals add more skills into their profile. In fact, this past year there were close to 365 million skills added to people’s profiles, and that’s about a 40 percent year-over-year change. This matters when you really accelerate towards the future. 

My eldest child is my 10-year-old daughter, and I think about the jobs she might have in the future. It’s really hard to tell right now because the job market is changing rapidly. We measured the skill sets for jobs now versus five, six years ago, and they have changed by 25 percent. We think that in the next five to six years it will change by another 50 percent, so the type of skill sets you need for a job are changing rapidly. Being able to focus on the ability to learn and pick up skills is becoming one of the most impactful ways to make sure we have a great match in the workforce.

I think you have a data problem there. I can tell LinkedIn that I have all the skills in the world. “I’m the world’s best carpenter.” A recruiter can search for really good carpenters and now they find me. How do you verify that I’m telling the truth, and that the skills they think they need are the skills they actually need?

That’s such a great question, because when it comes to skills, it is not an easy problem to solve in terms of how you actually certify and verify them. Just a few months ago, we acquired a company called EduBrite and their entire offering is certifications. One of the things we’re investing in the most right now is doing professional certifications on skills you have. Let’s say I started my career as a semiconductor engineer a long time ago, then worked up to embedded systems and think I’m pretty good in C++. Great. Let’s go for a certification process to assess and verify that skill. 

The great thing is that once they verify that skill, we can actually show it to the recruiter. It’s not enough for you to say, “I have this skill.” When a recruiter sees a candidate, they can see what verified skills they have on the profile. It’s a great way to perform a match, because we know being able to say, “I have this skill,” is a starting point. Being able to verify it is where the quality really gets made.

Okay. I have a lot of questions about that, because adding that kind of economics to this entire process probably changes the incentives and changes how people participate. I want to come back to that, because I think that’s really interesting. 

LinkedIn is unique because it is a professional social network. You can layer in things. If you want to tell people you’re good at C++, you can pay some money to take a test to get a check mark that says you’re good at C++. Other social networks cannot do that.

My thesis about social networks is that the product is effectively content moderation. The way that you moderate the content that gets published, the way that you amplify the reach of some content and diminish the reach of other content, the way that you allow some things and not others, and then most importantly, the way that you design the product itself to make it clear what things you do or do not want, these are all the same thing. Is that how you think of LinkedIn? Do you need to make sure that everyone is participating in a professional way? Or is everyone just doing what you told them to do?

Yeah, you’re talking about content in the artifact kind of way, with people exchanging information with each other. If you put it that way, I would agree. I think at the end of the day it’s about matching people. 

When you think about the core of LinkedIn, at the heart of it, you have this extraordinary professional community. Then you have marketplaces built on top of that. It could start at the basic layer. When most people hear content moderation, they think of content exchange or information exchange. For us, it’s about knowledge exchange. Let’s say I have a question about how I set up my payments. “I’m using Shopify, what’s the best way to set up my payment system?” Then somebody responds back. That’s one way of exchange. The other way is if somebody is looking for a job, the content really becomes my profile and my skills. Maybe that’s to your point.

The other aspect is building a business. If I’m a visual effects designer who really wants to build my business and I’m showcasing my abilities and talent on LinkedIn, then that becomes my content. Really the match is done when somebody says, “This is actually a phenomenal skill you have and I would want to hire you for a gig.” That is a real example that happens all the time. If you use content in that manner, that becomes the matching opportunity of those marketplaces. Ultimately the value comes when we see the hire. 

So back to your point, somebody applying for a job is a great starting point for a match. Somebody getting invited to an interview means there’s substance there in terms of that process. Somebody getting hired is also a good example. Somebody becoming an excellent employee is kind of the holy grail. 

That is the funnel. How many folks are coming into that funnel, and ultimately, what is that amazing match in the end?  If it’s in the commerce world, how many are becoming customers? All of those are happening on LinkedIn. Sometimes we can close the loop because we see the full engagement back when somebody’s looking for a job, for example. Sometimes we only see a few steps into the funnel, not the full closed loop.

So the metric that you care about the most seems to be the number of people getting hired and changing their profiles on LinkedIn to say they got a new job.

That’s one of our true north metrics. I’m sure you covered this in previous episodes, but we look at input metrics and output metrics. Basically, what are the input metrics that can ultimately drive value in the ecosystem? Then for our economic opportunity metrics, we look at true north. 

For example, in the hiring marketplace, we look at how many people are being hired in the economy right now. We know in the last year, eight people were hired per minute. We know we influenced a lot more, we are just not able to capture the entire feedback loop, and that’s just one side of the marketplace. If you look at the commerce or the products and services side, we look at how many outcomes. How many buyers are we connecting with brands? If a brand is doing an event, how many people joined that event? How many folks showed interest in that? How many converted to that product? It’s mostly in the B2B advertising context.

We also look at knowledge conversations. This is a really important distinction that we put a lot of emphasis on right now. Ultimately, we believe that the foundation of economic opportunity starts with knowledge exchange. We believe that when you have hundreds of millions of professionals in the ecosystem, they all have a unique craft, a unique expertise in their field. I could be two years into my role, but the niche that I know really well, very few people know in the industry. Being able to exchange that with others becomes extremely powerful as inspiration and for idea exchanges. 

We actually look at how many knowledge exchange conversations are happening in the ecosystem. We know that when they happen, they accelerate the other marketplaces I talked about. For example, we brought in creators to LinkedIn and they talked about their craft, like visual effects, or their ability, like negotiation skills or diversity skills. The biggest feedback we got was that the platform was powerful for them because they were able to get opportunities from it. It wasn’t revenue share. It was stuff like, “Oh, I was hired,” “Three months into the program, I got my gig,” “I got multiple speaking opportunities,” or “I got hired for campaigns.” I think that is the unique part of the platform.

Let’s go from there into what I think of as the Decoder questions. I have a sense of what LinkedIn is, and I have a sense of what you care about for metrics. You’re the chief product officer of LinkedIn, which is a small part of a giant company at Microsoft. Microsoft has a lot of bits and bobs, a lot of pieces of the puzzle. Once you’re on LinkedIn, you can go to GitHub or you can go to Microsoft Excel. You can see a bigger Microsoft strategy to just own work from top to bottom. What do you do as the chief product officer of LinkedIn to align the strategy? What does your day-to-day look like?

Taking a step back, if I were to put it simply, as the chief product officer, I am responsible for everything we build. It starts with setting the product strategy for the company, to then overseeing the teams building it. That includes the product leaders on my team who are leading specific areas, the design team, the business development team, the news and editorial team, and then our content production teams as a whole.

When you specifically think about the role of chief product officer, it’s relatively new in the industry in many ways. I think many companies are shifting to think more around the customer at the center of their offering. From that you go directly to the product side, and then you’re seeing a lot more head of product or chief product officer roles. A lot of the work we’re doing from there starts from how we align to our vision statement as a company, and from there you go to the strategy level and the execution across the teams.

What do you think is the difference between vision and strategy?

Early on, when I joined, Jeff Winter was the CEO. This is something that was an extremely big part of how we operated and thought at LinkedIn. Jeff codified the vision as being the dream. Ultimately, if you’re successful, what change are you making in the world? The mission was more of that tangible aspect that you can almost start measuring on a daily basis. 

Then we go all the way to values, by the way. We call this the “vision to values” process. It’s an incredible exercise that everyone went through in the company. The vision for LinkedIn is to create economic opportunity for every member of the global workforce. Our work will never be done. This would be one of those incredible things, and we can always do a lot more. The mission is to connect the world’s professionals to make them more productive and successful.

That brings me back to our conversation about marketplaces. I can see if I create a great hire or a great outcome for a buyer or seller. I can see if I create a great outcome for a knowledge seeker or a knowledge creator. Then we go from that all the way to the strategy. Obviously the vision and mission are more evergreen. They don’t change. They’re constant. The strategy evolves with time based on the market needs. Then all the way to the values of how we operate. What is the culture of the company itself?

When you think about this strategy changing on a different time scale than the vision and the values… or was it the vision and the mission? We need to make a flowchart. 

I have a flowchart. I can share it with you.

Yeah. You have to share the flowchart with us. We’ll put it in the show notes. 

When you think about the strategy changing more rapidly than the vision and the mission, that to me is the ball game. The strategy of the company is pretty big. Everything should ladder into the strategy, but the market changes, people change, people stop using their desktop computers and start using their mobile phones, then maybe one day we’re all going to be in headsets.

The product has to change really rapidly to get ahead of those big shifts. Do you sit around worrying about the metaverse? Do you think, “Oh man, there was a boom in PC sales last year and more people are working from home, which means more people have the opportunity to have LinkedIn open on their laptops, which they could not do in their cubicles at work.” Is that stuff that comes up to you, or is that happening way below, at a more tactical level?

I sit around getting excited about it. Just a quick story there. So I joined LinkedIn…

Did you join LinkedIn via LinkedIn?

Not really.

Have you ever “dogfooded” LinkedIn? Have you ever answered a cold recruiter pitch on LinkedIn in an InMail?

Of course. All the time.

It’s pretty thirsty, man. The recruiters in the InMail are pretty intense.

When you start leveraging LinkedIn and its full abilities, it’s an incredible superpower skill for everybody.

Have you ever gotten an offer on LinkedIn, then gone to your boss to pretend the offer was more serious than it was?

I have not.

Okay, because that is the true LinkedIn power move, to get the raise just out of the InMails.

That’s not uncommon though. Sometimes those become real. In fact, one of the beautiful things about my job is when you go to meet people and they tell you, “The way I got this gig is through LinkedIn, and it started with a cold email.” 

Three months ago, somebody called me and said, “I can’t believe how I just fundraised. I sent an investor I really respect my deck on LinkedIn and he said, ‘Can we talk?’ I never believed I could do that so effectively on LinkedIn.” They were like, “Oh, mind blown. I can do a lot more on LinkedIn.”

Again, have you just primed the community to open this website or app and be ready for these kinds of interactions? That’s the thing that I keep coming back to. I can cold pitch a deck to anyone at any time on the internet. We can put someone’s deck up on our TikTok channel, but it doesn’t seem like that will convert into a fundraising round. It just doesn’t seem likely.

On LinkedIn though, it does seem likely, and maybe it’s the same people. They’re just pushing a different app icon on their phone, but it’s because you’ve primed them to participate in the marketplace in LinkedIn. That just feels like it has to be a different kind of product.

“You come to LinkedIn to check in, not to check out.”

It is, and it is 100 percent a different type of product. I think if you want to understand LinkedIn, the analogy that I would give is that you come to LinkedIn to check in, not to check out. I think it’s a big difference versus the other social platforms we sometimes get compared to. Ultimately, LinkedIn is a productivity tool. It’s hard to call the other social platforms a productivity tool. That is why it’s really safe to open it up at your workplace. We want you to open it up at your workplace.

I don’t know, man. I feel like if the boss is walking the floor and they see everyone has LinkedIn open, then something is wrong at that workplace.

You need that today as a base to ask questions. Sometimes all you’re doing is seeing what’s happening in the economy, asking questions, or coming back with the recent Twitter news — which is hard to escape.

Let’s go back to your original question on why LinkedIn is unique. The context matters a ton because it brings the right audience in. You wear a hat right now at work — literally, you’re wearing a hat — and I’m assuming it’s your professional hat. When you walk out of work, you’re going to wear a different hat. You might go and use a different tool. That makes a big difference in what type of exchange you expect at that moment. 

That’s why the power of the ecosystem is about who is in it right now and the type of interactions you’re looking for. If I’m going to a soccer match for my kids, I don’t want somebody to go and pitch me why their startup is fantastic. I’m watching my kid’s game right now. But if I’m on LinkedIn, then I’m in that mindset of doing work.

There’s more affordability in my mind to actually accept to bring that in. You asked the question around strategy and evolution, and how technological revolutions play to it. That’s how I joined LinkedIn. In 2011, I had a conversation with Deep Nishar, who was in my job at the time, chief product officer for LinkedIn. We had a chat about LinkedIn and he asked me, “How would you build LinkedIn as a mobile product?” Back then LinkedIn was heavily desktop-oriented. We had a great conversation. Then he said, “Instead of talking about it, how about you come and build it?” That was the way I joined. That was the first conversation going into it. 

Obviously ever since then it has been AI, and it’s going to continue to be AI for a long time. But specifically, when you look at technological revolutions, they inherently change strategies for companies because they inherently change the product you’re working on and the abilities of the product. AI is going to continue to be that forcing function for a long time. In fact, I think it’s just accelerating in power. Ultimately, if you are a tech company, it has to impact the offering, the product, and the way you service your members. That’s why the strategicness of it changes significantly.

I’m going to make a prediction that next year, 15 percent of your podcast in some shape or form will be mentioning generative AI.

Do you think that is because it’s useful and it’s actually changing things? Or do you mean that in the way that people would just say Web3 to me last summer?

No. This is very different from Web3. Web3 I think was trying to find applications. This will have so many applications.

For you, what are the applications?

I think we’re getting started, but think about creativity. I heard from Reid [Hoffman] recently that with AI in a way everybody will have their own co-pilot, their own personal assistant for things. The notion of starting from scratch would no longer exist. You would always have some draft in mind or some knowledge base you can start from. 

I think there’s already a creative base for things, which is pretty unique. It used to be that whole notion of tabula rasa and starting from scratch, but that no longer exists. I think about the applications across healthcare and medicine, across assistant information like decision-making, and obviously across the creative roles and the arts, which we talked about. We’re literally just getting started. You can take me up on that bet we talked about before. I’m pretty positive on it.

You’re a creative person. You had to come up the hard way, and the way a lot of creative people come up is they build copies of things to get good at it. The classic example is art. Students go to art museums and they copy the masters’ paintings. People who design cars grow up copying cars. 

At every stage of being a master of creativity, it is useful to just try to make the thing yourself in the way that your predecessors made it so that you can expand on it. Your example takes that entire piece out of it. Where do you think the expertise comes from in that situation?

You’ll still need to learn how it comes to life, but you’ll have a stronger base to start from. Again, there’s always that implication. That notion that it can be very scary, because it’s going to reform and re-disrupt things, and we don’t know exactly how and how much. There’s the form around, “Wow, the creative base is just going to expand so much and human capacity will be elevated.” I think that is the positive side of it. 

You’re a hundred percent right. I think creativity starts from mimicking and learning from others, then building your edge or your nuance around it. I think that is still going to continue. In fact, it might be elevated, because your base or starting point is going to be so strong.

Is it strong? That’s what I’m getting at. Is it strong? When you see the output of ChatGPT on social media, there is a strong human filter between the actual outputs and what you’re seeing, because people are picking the best, the funniest, or the most interesting output — and that’s cool. Then you actually use it. This is just spitting out a remix of all the stuff that has ever existed before, and the remixes are really cool. 

I saw someone use it to make guitar tablature the other day, and it was awesome. I don’t think the OpenAI people thought that it could make guitar tablature. That is really cool. But at the same time, a person just playing the guitar tablature and a person trying to learn how to play the guitar, they’re different directions in a way. Not being able to evaluate whether or not the output is good and worth building on seems like the challenge, especially for you to build it into a tool like LinkedIn, where people are just going to ask questions and assume that the output is useful or worthwhile.

Yeah. Here’s an example you can think of. Often at LinkedIn, we see scientists who want to share their craft and their research. It’s hard, because how do you bring something that is very complex to the masses? That’s one application of it. You can take something very complex and potentially use it to simplify a concept. 

The flip side is that you can start with a kernel of an idea, then start adding depth to it. It’s that feedback loop between — I guess you can call it bionic — the machine and the human being. Again, it has repercussions that could be very scary, but it has repercussions that could be very exciting at the same time. I think we’re literally just getting started, so we’re going to see a lot over the next one, two, three years that will tell us a lot about the capabilities here.

All right. Here’s the classic Decoder question. We have talked a lot about the kinds of decisions you might make and the ways you might make decisions, but how do you make decisions? What’s your framework to make decisions?

I think you can guess that we start from the vision and mission. We truly do. That’s like the main currency, and I can’t emphasize this enough. Then it goes all the way to our values as a company that kind of ground the day-to-day. 

For example, there is member-first, so we start from members. Early days at LinkedIn, we used to have conversations around, “Hey, can we get this information from members?” It is like, “No, we are first and foremost a member-first company, those are the constituents we serve.” Now that sets the overall framing, and that’s true across LinkedIn, but that framing is kind of wide. When it comes to how we choose what to build, there are three things that matter the most to me, as a product organization. 

First, LinkedIn is extremely interconnected. We talked about multiple marketplaces. At the heart of it, you have a professional community. On one hand we’re trying to connect job seekers and hirers, but we’re also trying to connect brands and sellers. So context really matters, and being a learning organization is a critical part of connecting the dots. You can be extremely powerful at LinkedIn if you’re able to connect the dots between how this amazing economic system works. We have a lot of open forums. We go over performance and we learn from member feedback. I get tons of direct and indirect outreaches from members giving me feedback. Experiments are shared broadly so everybody can learn how experiments are doing. 

Then the second part that matters a lot to me is being very opinionated. As an organization, one construct we have is what I call product jams. When people come in to present a new product or a new strategy, what’s important for me is that they come in really clear on what job is going to be done. What are they serving? What does the audience need? What is the problem statement? 

There are so many times when people think the product statement is clear, but then you realize people are thinking of two completely different problems and that sets them in a whole different direction. Be clear on your principles and what opinions you have there and you can actually have really deep, tremendous decision-making conversations, because there’s so much clarity in the thinking that it leads to, from my perspective, a much better outcome.

Who comes and presents when that happens? Is that a single product manager? Is it a cross-functional team with a product manager [PM], some designers, and some engineers? Is it two people who met in the hallway that had a skunkworks product and they came and showed you? How does that work?

It’s all of those, but it’s really a team of cross-functional leaders: the PM, the designer, the engineering lead, the marketing lead. Depending on the product, there could be the go-to market lead as well. They come and present their local view of what they believe they should be building and how. Then in that session, I have my entire executive team join, and they all bring questions like, “What’s the job seeker’s view? What’s the buyer’s view? What’s the seller’s view? What’s the knowledge seeker’s view?” It’s a great match between the global view and the local view. The more clarity you have in the thinking, the better those conversations become in terms of decision-making. 

We have multiple small venture bet experiments we are thinking about. It could be a very small team of five people, or it could be a team of 500 people. There is a strategy that the leaders of that team are coming and presenting. It really depends on the type of conversation. It could go from 10,000 feet to 100 feet depending on the conversation itself. I would say that when it comes to decision-making.

You run a very professional social network. There’s commercial aspects to it and you have described it as a collection of marketplaces. I don’t think people think of consumer social networks as a collection of marketplaces, even if they have some marketplace functionality. If you walked up to a regular Instagram user and said “Instagram is a collection of marketplaces,” I don’t think they would have any idea what you were talking about. Whereas I think if you asked a LinkedIn user, “Is this a bunch of marketplaces?” They might just say yes. They might perceive the whole product that way.

I think it’s going to be really hard for Twitter to add a bunch of those paid products. Maybe they’ll be successful, maybe they won’t, but I think it’ll be challenging. There’s a flip side though. LinkedIn has tried to add a bunch of consumer features to the more-professional social network.

This is a show about decisions. You launched Stories and then you took it away. There was a bit of a Clubhouse competitor, a live audio space program, but it seems to have diminished. Talk me through the decisions here, including, “We’re going to do Stories like everybody else is doing and then we’re going to shut it down because it’s not working.” And then let’s talk broadly about why those consumer-style features may or may not work.

Let’s take a step back. With consumer products, ultimately you are trying to support the member in whatever they’re trying to do. Earlier I mentioned the “job to be done.” If I’m trying to express myself, what’s the best way to express myself? Stories is actually a great example. For us, there were two reasons why Stories was a really interesting experiment — and I classify it as an experiment because that’s exactly what it was. One aspect was, it’s a better way to tell a professional journey. I can have the multiple taps. I can have that imagery and rich media coming through in a visceral way, and I can tell my story better. If I’m trying to communicate, “I have advice on how to fundraise,” I can do that in a story-like way.

Stories didn’t work on LinkedIn because its users are not looking for ephemerality.

The other hypothesis behind Stories was that we used to hear a lot from members that one of the concerns around potentially sharing on LinkedIn was that a post persists, it’s always on their profiles, and Stories had that ephemerality notion that could potentially solve that. What we learned was exactly the opposite. Things remaining on your profile is a feature, a benefit of LinkedIn. What you share on LinkedIn, you see as part of your professional identity. People actually share and engage on LinkedIn. They were not looking for ephemerality. In fact, they were looking to feature it on their profile, not to make it disappear. 

I think that was a phenomenal underscore for us on the role of content in your profile. When I share on LinkedIn, it’s not about virals and likes and comments coming through. It’s really about showcasing to everybody who’s looking at potentially hiring me or selling to me or partnering with me. What do I stand for? What’s my thought leadership? What’s the stuff I care about?

This is all personal marketing in a way, right? What you’re describing is a large ad for me and the services I can provide for you. That is what my LinkedIn profile is and what my LinkedIn interactions are communicating.

It’s your professional identity. If done authentically, it’s who you are.

The reason I frame it that starkly is that almost every other social network is getting away from that, right? Facebook, once upon a time, was about a profile and you would curate an image of yourself in your profile. Instagram famously was the most curated identity platform of all time, where people would create these artificial versions of themselves being extraordinarily beautiful, whether or not that had anything to do with reality. I may or may not have been guilty of this. Who knows?

LinkedIn is still that thing. You are creating a professional identity for yourself that you are marketing and you don’t even want it to be ephemeral. You want it to last because it stacks up over time and accumulates value. Everyone else is down to make a piece of content and hopefully it will go viral in the slot machine. Why do you think that has broken in such different directions?

Honestly, I just think it showcases how LinkedIn is unique and different in its professional context.

That’s very good. LinkedIn is special, but why do you think it’s so different?

Yeah because when I interact with you, being able to understand who I am... Think about any interactions. Our members are looking to fill a need. I’m looking for my next interviewee for my podcast. Who’s somebody interesting to look at? Okay, look at the LinkedIn profile. I can see what they’re sharing and talking about. That could be interesting material for me for my show. I’m looking for a great person to partner with me on my startup, be my engineering co-partner, be my product co-partner. LinkedIn is a phenomenal way to understand who this person is, their professional identity, and what they say.

This is another Twitter question, I apologize, but it’s right there. It’s the Death Star hanging over this conversation. A lot of people say to me, “I need Twitter for my job.” Reporters especially. “This is where I find my sources. This is where I market my material. This is where the other reporters are. This is where the editors are.” Authors, too. There are publishing houses in this country that will give you a book deal if you have enough Twitter followers, basically without knowing what your book is going to be, because they know they have a marketing channel. There are lots of comedians. There are just tons of people who are on Twitter and the reason they’re on Twitter is like, “I need it for my job,” which is a really weird way to think about Twitter, but it’s how they think about it.

Isn’t that LinkedIn? Shouldn’t you be going and trying to get all those people to come use LinkedIn, especially in this moment when Twitter seems like a disaster?

I think it depends on what they’re trying to do. We actually learn a lot from journalists that they’re doing a lot of their job on LinkedIn, trying to find leads to stories. If I’m looking for a cybersecurity story…

In this moment, you’ve got a huge audience of people on Twitter that maybe for the first time ever is at scale reconsidering its relationship to this platform. I can’t think of another time this has happened for a social network like this. Usually, the network effect just takes you from Friendster to MySpace to Facebook. It’s usually not, “Something bad happened and we’re all going to quit.” Twitter’s in the middle of that moment where everyone’s reconsidering the relationship to the platform, and a lot of them are like, “I need this for work. I’m panicked about what I’m going to do in my career if I don’t have this tool.” Are you doing outbound marketing to say, “Hey, come work here. Actually, we are better for this because we are work”?

We’re not doing any outbound marketing. I think in many ways it depends on how they see the value they’re getting from the platform. For many of them, they’re already on LinkedIn. Many of them are actually getting their job done on LinkedIn. That’s the stuff we hear. So for example, you mentioned audio events briefly. Two days ago, TechCrunch ran an interactive audio event on LinkedIn. It was about Twitter. It was their entire editorial team having conversations, taking questions from the audience regarding their thoughts about the business model, what should be done, and what should be the interaction model going forward.

LinkedIn is highly professional, so it was a professional conversation about Twitter. So I think what’s naturally happening is people are realizing you can actually get a lot of what you’re trying to do on LinkedIn already, and that’s coming through for the system as well.

Here’s another example. One of the very successful products we have right now is newsletters. People are building new newsletters. It’s a great way for individuals to cultivate intimate relationships with the audiences. So we have 150 million newsletter subscriptions running right now on the platform, growing extremely quickly. Forbes launched a newsletter-plus-group combo right now, so they run the newsletter and it gives them direct access to their subscribers. And then there’s also a group for discussion about the material shown in the newsletter. It was done naturally. It was done before the Elon news. It will be done many times after. And I think generally, publishers, audiences, whether it’s professional identity as a person or as entities, are always looking for a better way to get their stuff done. LinkedIn tends to be a phenomenal way for them to do so.

The best product teams are usually composed of a product manager, some engineering people, and some designers, and ideally they’re all aligned. You said something to me that I am just never going to forget, which is, “You can disagree, just don’t be fucking confused,” which is incredible. We should make T-shirts of that.

ChatGPT is like that. These tools are the place where a product team like that is going to disagree. We can even put that into practice here. You have product teams and you have to manage them. They often disagree. No one knows how to use these tools, no one knows what the right outcomes of these tools are, no one knows if they’re moral and ethical in some cases, and no one knows if they’re useful. They’re cool, but we haven’t put them into practice, so there is no set of best practices or industry secrets to learn from. We just have to invent new patterns. How do you manage a product team through that?

“We might be wrong, but we’re not confused.”

Yeah, the exact quote is, “We might be wrong, but we’re not confused.”

Oh yeah, okay.

It’s a very powerful way to lead a large organization in general, but there is also some clarity of thought and clarity of execution for everybody.

Do you want to say that again with the F-bomb? Because the F-bomb was very powerful to me.

Well, we can play it through. It’ll be more natural as we go into the podcast.


It might come up a couple of times. Last time, you called it tension between designers, product managers, and engineers. I think what you find is that they’re solving different things, and that’s the problem. They’re not saying it, but they are solving different things. 

This is where the confusion starts. That’s why I say I would rather be wrong than confused. To come with clarity of thought lends itself into clarity of execution. I love that you and your podcast talk a lot about decision-making and frameworks. Those are super useful, but there is very little discussion about actual clarity of thought in those decisions and carrying it through with focused execution. 

Whether you’re leading an organization of thousands of people or just leading yourself, I believe that when you are confused, it becomes impossible for you to rally people, especially when your target is really challenging and you lack that confidence. It’s about striving for clarity and conviction, while at the same time trying to remain humble and self-reflective to make sure you can still go through the process.

Start from what we talked about before, with designers and engineers. I think you’ll find that if you take each into a room and ask, “What are you solving for with this proposal?” They will tell you different things. There’s no point discussing if they’re both coming in with different problems. Maybe you should discuss the problem first. You can start a conversation by saying, “Hey, what is the problem we’re solving for?” Don’t give me a headline. Give me a nuanced, articulated definition of a problem. In fact, get uncomfortable with the problem. Give me the trade-offs of the problem. 

You’ll see people saying, “Oh whoa, that’s not what I’m solving for, but are you solving for that?” This is a guesstimate based on experience, but the engineer might say, “Hey, I want to make sure we can leverage the platform capabilities we built. We worked for a year, we built a platform, and now you’re building something very different. Why the fuck did we build that?” They’re right. The designer would say, “Wait, we’re not innovating. There are so many innovative patterns and motion graphics today, we can start playing into this. This looks like it was built two decades ago.” They’re also right, but they are solving for different things.

Going back to, “Hey, let’s just align with the problem,” that would probably be the most important thing you can do as an organization. This is true for everything. I don’t want to talk politics, but we can talk politics really quickly — you’ll see they’re solving for different things. I think that’s like number one. Once you have that and there is alignment on the problem, it’s actually shocking how easy the solution can be, because sometimes we just spend that time talking about the problem a lot. 

When we come into the solution, I ask my team to write down their principles. When we do a product jam and bring in a new product strategy or design, I ask them to lay out their principles. Good principles have teeth. You’re supposed to see the trade-off from a good principle. A bad principle is like, “We should have simplicity in the product.” That’s amazing, that’s beautiful.

For example, we should add some friction to our signup process so we can make sure we don’t allow bad actors in at the expense of signup growth. We can debate that for hours, because you want to build the most frictionless experience possible. But sometimes with frictionless experience, you’re also introducing an entry point for bad actors. You don’t want to have that, especially in a product like LinkedIn. We want to make sure the trust is very high. Then you go into, “Okay, is there a way to build a frictionless experience?” You can spend time debating that for a long time. There’s potentially something innovative that you could do there, but the constraint is very clear. 

A bad principle is like, “We should optimize for the members having a great experience and for the customers on LinkedIn.” That’s great, but you didn’t tell me much. I don’t feel uncomfortable. I think it’s the moment you start feeling like, “Wow, if I decide on this axiom, this type of principle, then I’m letting go of something else.” I think great principles have trade-offs embedded into them. 

Now, sometimes when I talk about how I might be wrong or confused, it can come across somewhat as, “I know I’m right.” It’s not about knowing I’m right, it’s about being humble and letting go of being right. It’s about not caring whether you come across as smart or not, but caring about if the company is successful. If you go back to those examples of team dynamics and tensions, there’s a lot about the interests for my organization, for my function, for what I need, and it’s less about the company. 

People get attached more to company goals, clarity, and focus, and they get detached from being right and wrong. You’ll get phenomenal conversations. Now, this is just the opinion part. We haven’t gotten to execution. Many companies tend to think that if they have clarity on the decision, the execution will just come for free. That’s not correct. There’s so much confusion in execution, and we see this quite a lot.

This actually relates to something I’ve been wanting to ask you about. When you think about product organizations, most people think about what you have described at LinkedIn. A product manager [PM], some designers, and some engineers. The PMs in most situations are powerful. That culture of product managers being very powerful is pervasive throughout the tech industry. We had Tony Fadell on the podcast, and he will go on and on about how having a great PM is basically the secret to all success in life. If you go to Google, PM culture there is like its own thing. It’s its own universe. 

There are companies that run in other ways — I’m thinking of Apple — where design is much more prominent and they have product marketing managers who connect the designers and engineers to the customers. Maybe these are all just different names, but the companies and their products are expressed very differently. There’s an inherent trade-off to that: how are we going to structure the jobs? How are we going to give them titles? Who’s going to have the ownership of the decisions? How do you think about all of that?

It’s a great call. With product management and different organizations, it really depends on the company you join. If you look at product management as a skill, some designers do it amazingly well, and some engineers have phenomenal design skills. Some of the best people I work with can play all roles. They can be very technical, they can be focused on the customer and the business, they can focus really well on the craft of the design, and it depends from company to company. 

When I started as an engineer, we didn’t have any product managers. As an engineer, I was the product manager. We used to pride ourselves on building great technology, but we didn’t actually understand who it was for. So we had the best technology, but not really a solution for it. It was kind of odd. 

To your question about the product management side of the world and how it comes to life, for me, it’s kind of a good description of roles and responsibilities. What about execution? I run an organization that is composed of product managers, designers, business development, an editorial team that is responsible for our content and plays a big role in decision-making, and our content production team, so it’s a very diverse team of product builders. We have a lot of clarity on the roles and responsibilities, and it could be different for different projects. 

We have a process we call RAPID [recommend, agree, perform, input, decide]. For every project there is a decision maker and there is a recommender who tends to be the subject matter expert. The decision-maker could be somebody who has multiple objectives they’re trying to prioritize in running the organization. There is a performing team, inputting team, and approving team. You can think of legal in some cases as an approving team. We want to make sure we stay away from doing anything illegal and that we do everything with high trust. We make sure that whenever we start a process or a project, there is clarity on roles and responsibilities. 

“Only one person can be the decider.”

Only one person can be the decider. That allows for velocity of decision-making. For example, design gets what we call “the complete D,” which is the decision on how the product is going to look. Engineers have the complete D about how we build it, and product managers have that for what we are building and the features. The roadmap is a product management decision. Now, they all brainstorm with each other. We hardly have any conflict within core teams of PM, design, and engineers. I can’t think of the last time it happened. Usually it’s a bit across different product teams that have competing objectives based on resources. That’s usually where the tension comes from.

You know that I’m really interested in structures. You’re saying you have these core teams and you oversee it all, so it all ladders up to you and maybe you just break all the ties. I see it at other companies. They have a design department and an engineering department, and then they make cross-functional teams and say, “Make this number go up.” Then there might be some argument within the larger organization that prevents the team from making that number go up. Do you see this? 

I can just pick any large company, like Microsoft. This is a thing that happens at any company that hits any scale, so I don’t mean to pick on any one company. I hear about it all the time as I talk to people like you. These are the actual problems you solve in your day. How are you structured? Are you structured to have an engineering department that lends engineers to cross-functional teams? Do you have PMs that run teams of engineers and designers?

At the company level, we roll up functionally to our CEO, Ryan Roslansky. So I run the product team, my counterpart runs the engineering team, we also have the legal team, the marketing team, and the sales team. In my organization there are the product managers, the designers, business development, content production, editorial, and so on. We then run product areas. We talked about how we have a team dedicated to the talent marketplace, connecting job seekers with recruiters. We have a team dedicated to our products and services marketplaces, connecting marketers and sellers to buyers. We have a big team dedicated to knowledge exchange between knowledge seekers and knowledge creators. 

Those run separately, and they have every function represented. So, the job seeker team has a PM, a designer, an engineer, and a marketer. It’s rare — I can’t think of the last time it happened — that within a team there is a need for an escalation, because they usually start from a very clear understanding of roles and responsibilities and there is a roadmap. It is not rare that there is escalation across teams. For example, the trust team and the growth team talked about our signup flow and how much friction there was. This is where we actually encourage escalations. 

“Are we disagreeing or misunderstanding each other?”

There’s another thing that I use often that I learned from a good friend. “Are we disagreeing or misunderstanding each other?” You will find that in large organizations, teams debate stuff forever and usually it starts with, “But wait, I don’t understand you.” What they mean to say is, “I disagree with you.” There’s a big difference between misunderstanding and disagreeing. I would tell my team that if you disagree, escalate and move it up. If you misunderstand each other, spend the night explaining to each other until you’re both clear on what everybody is trying to accomplish. If there’s a disagreement, it’s not worth the time to just try to convince each other to death, just move it up.

In a large organization, when that type of conflict is healthy, it is very efficient. When it is unhealthy it gets very toxic very quickly. How do you manage that kind of conflict in an organization as large as yours?

To your point, I think it is culture. It starts from the leadership and your role model at the organization. 

For us, we have what we call a clean escalation. You escalate with everybody else on the thread. You don’t escalate separately, where everybody escalates to their boss and has their own information that is not shared. Everything is out in the open, because honestly it’s all about LinkedIn for us. It’s about delivering its incredible vision and purpose. If you let go of your ego, it’s less, “Am I right or wrong about this?” and more about LinkedIn becoming successful. 

It’s very easy to do clean escalations. It’s very easy to say, “Okay, all I want to do is continue on my job.” Most of the time all people want is to just have progress. The last thing you want is to just stay put. That is the most dysfunctional organization possible. You’re not doing anything when you’re debating things forever and trying to be right or wrong, versus focusing on the company and the purpose. 

All the way from Jeff, our former CEO, to Ryan, our current CEO, it’s about role modeling that all we care about is fulfilling the vision and purpose of the company. Whenever things linger for no reason or somebody sends me an escalation in an improper way, I will add everybody to it and ask for a clear understanding of what we are trying to solve for and the principle thinking behind it. It has to be role modeled. This is one where if you don’t show it to your team, they will not do it.

You mentioned designers having a decision about how things looked, product managers having the decision over what features would be built, and engineers having the decision about how it’s built. If I had to sum up The Verge in a sentence, I would say it’s a publication that has persisted for 10 years because we are very interested in buttons, who puts them there, and why we push them. It is the simplest thesis statement of our entire publication that the presence of every button that you encounter on every product is actually a long and complicated story, from the genesis of the button all the way to how some person is going to push that button at some rate.

I often find that at big tech companies, the politics of, “we should put this button here,” and, “here’s what it looks like,” are not two different decisions. It’s often the same decision. How big the button should be and what order the buttons should be in are not design decisions, they are product decisions. The product managers and the designers often fight quite a bit over what buttons go where and in what order, or how big they should be and what they should look like. How do you make the determination between the two?

That’s an easy one because you can just experiment. Honestly, it’s not worth a discussion on if the button should be blue or yellow.

But that pisses people off. That example carried all the way is Google testing 41 shades of blue and the design culture of Google just leaving.

So here’s a good example. Now if somebody says, “Hey, I want to use multiple types of fonts all over the site. I think if I can get somebody to squint their eyes all the time, it will lead to more engagement.” This is one where you say, “I don’t think your design judgment is very good, even if you can show me some movement in numbers.” That’s not a good design if you think about the aesthetics of how the product looks and how simple it is. Sometimes it’s 100 percent a belief system. Data cannot tell you everything. Data should give an answer based on your hypothesis, but it should not tell you data without a hypothesis. It should start from something.

There is a notion of the simplicity of the experience and how you build it, and sometimes those translate over time, so you want to build something in. “I think the button should be here or there,” doesn’t make any sense unless the design itself doesn’t make sense when you zoom out. Like, “Hey, I think we should have ‘sign up now’ instead of ‘join now.’ Test it.” That is not worth the debate. This is a good example of debating the sand, not the big rocks. 

I think great debates come from asking, “What type of experiences are we building?” LinkedIn serves so many audiences — job seekers, marketers, we try to serve every functional hat on LinkedIn, and that could lend themselves to a very complex product. 

Imagine everybody wants to have a tab. “I want a tab for job seekers, I want a tab for buyers, I want a tab for recruiters.” That’s a really complicated experience to build. The flip side of it is, “How do I build something very simple that caters to every audience?” That’s a great thing. With no specific audience having its own tab and experience, that’s where you start to push the design into true innovation. You say, “Okay, to really perform on that, we have to have incredible relevance.” If I want to make sure nobody needs their own hub and their own home, and I want to build it for a really simple experience through search, feed, and notifications, then we better understand the intent really well when somebody comes to the site. 

Then when Nilay comes in I can serve him a great experience. I don’t have to say, “Oh, Nilay is looking for a job, so he has to go to the job seeking tab.” “Oh, Nilay is now recruiting somebody for his team, so he has to go to the recruiting tab.” “Oh, Nilay is trying to market something,” et cetera. That’s where you start building what I would say is like an octopus, very much a complex experience. 

Building something simple is very hard. That’s where you start pushing the elements. I think those are the discussions where healthy confrontation and good tension are actually amazing. You can start pushing the experience tremendously with those.

A lot of that is healthy confrontation and good tension. You started off by saying good principles have the trade-offs built inherently into them. I want to bring that all the way back around to where we started, which is that in AI — particularly in generative AI — there are no best practices for this stuff yet like there are with almost everything else. 

“Should the buttons be bigger or smaller? We should test it. We shouldn’t argue about it.” There is now a universe of product design’s best practices that might at least tell you where to begin or where the pitfalls are. There’s an institutional base of knowledge that good PMs can just access. There’s also an institutional base of knowledge for good designers and good engineers. There’s none of that with generative AI. How do you avoid those pitfalls? How do you engage and test that stuff? That is a blank slate for groups of people who have to make pretty intense decisions that could have gigantic positive returns for your business and also gigantic negative consequences.

I think that is the biggest question right now for every technologist who has had a chance to go deep on this technology. What are the best practices? I think we had years to develop best practices with technologies that we understand really well that we can go deep on. When it comes to the capabilities of generative AI, I think we’re going to start seeing some good but potentially scary disruption in the market. Building that playbook is going to be one of the most important things we can do.

I think we’re getting to a place where technology is leapfrogging society’s ability to comprehend it, and we’re starting to put more muscle and more intelligence around building guardrails for what is okay and what is not okay. I’m a big fan of the work happening out of Microsoft with responsible AI. There are so many elements to it that are really key, from fairness to equity to how transparent it is. Sometimes those are constraining AI, but constraining it in a way that helps society move forward with it. A lot of that is needed from many companies. We haven’t talked about the roles of governments or third-party institutions in this, but I think that’s going to be the biggest conversation coming upon us in the next few years.

We can talk about governance and regulators. I think they’re pretty far behind the curve in general.

Especially in the US.

Yeah, that feels like a safe statement. But a big company like Microsoft is not. It’s a big company that has a responsible AI initiative. It is a leader in many of these services, for many of these kinds of bleeding-edge technologies. You run a pretty sizable team that is part of LinkedIn, which is part of Microsoft. Is there a place at LinkedIn where the internal Microsoft AI regulatory body shows up and says, “I need you to stop helping people write their own resumes on LinkedIn using our generative AI product until we get it right”?

I would say that it is aligned. We are a proud part of Microsoft, and they are leading the industry in this. They care tremendously, and there is ethics built into every element of how responsible AI is built. We learn from them in many ways. 

At the same time, LinkedIn is unique. User-generated content happens a lot more frequently on LinkedIn than on the other parts of Microsoft tools. There are parts we see that are really unique to how we build it very correctly. Sometimes on purpose we will constrain it, because it’s valuable for our vision. And it’s helpful for us to go deeper into what type of elements we’re talking about. 

“Not everything is great on LinkedIn. It’s not conducive for every type of conversation, but we want to make sure it has the most productive conversations in the workplace.”

For example, when it comes to content. By definition, we say on LinkedIn that we expect professional content. Not everything is great on LinkedIn. It’s not conducive for every type of conversation, but we want to make sure it has the most productive conversations in the workplace. People talk about work, how they work, and their craft. We can go deep on a vertical like workplace conversations, so that’s something we can contribute back to the brother Microsoft elements. 

We are so aligned on this. There’s a massive understanding of the importance of it, because of Microsoft’s diverse portfolio and the different aspects of their product — from search, to social professional networks like LinkedIn, to gaming, to productivity tools like Office. I can go on and on. Every business brings their best intellect and principles to the table to showcase how it’s unique for them. I think we’re pretty unique within the Microsoft landscape.

I want to do a two-minute, “Would you allow this AI feature?” lightning round to end this conversation.

This is a new one. You just thought about it?

I just thought about it. It’s just in the context of this. It’s user-generated content.

I might steal it. We’ll see.

No, I want royalties for all of these, but some of them are going to be crazy as I think of them. Generate a profile picture of me where I’m wearing a suit.

Your face?

I’ll take a selfie. Make a profile picture where I’m wearing a suit, which is the most LinkedIn thing I can think of.

For us, the principle is about proper reflection of who you are. I think that I would have to look at the picture of you in a suit to make sure it’s a good reflection of who you are. As long as somebody can recognize you easily. For us, the authenticity of who you are is really important. If it’s in line with that…

So that’s a yes?

That could pass.

With some guardrails, that’s a yes.

That sounds like a yes.

Write me a resume where it says, “I know how to operate a forklift.”

“Write me a resume…” I think right now you can write yourself a resume and bring it to LinkedIn.

I’m saying I want to type into a box on LinkedIn, the same way I would in ChatGPT, and say, “Write a resume that says, ‘I know how to operate a forklift.’” By the way, just to be very clear, I do not know how to operate a forklift. I think it would be cool to try, but I do not at this time know how to operate a forklift.

You and I have similar childhood dreams, so I think it comes to life really well. It’s important that it starts from your profile and is authentic to who you are. To emphasize something for both the product builders and the audience, it starts at, “What do you exist for?” For us, as long as what you’re presenting is authentic. “Hey, here’s my work. I worked for this forklifting company or this inventory company. This is my work, but help me craft it better.”

The line for you here is “don’t lie,” because right now I could lie.

Don’t lie.

You don’t want to help people lie.

If it’s about potentially better crafting or better articulating the work you’ve done, that’s a nice way to elevate and make it clearer for others about what type of work you actually do. Most of our members and most of our engagement is international, but the vast majority of the profiles are in English. Somebody comes and says, “Hey, I don’t have great English, but this is my work,” and they write it in Chinese. “Can you make sure it’s written well in English?” I think that is a great example of generative AI.

I picked “operate a forklift” because I feel like the truth comes out in that scenario very quickly. You get hired to run a forklift and you show up to your first day of work, it’s going to become very obvious that you lied. What about writing me a resume that says, “I have great interpersonal skills”? Write me a resume with some examples of my great interpersonal skills, which is obviously impossible to measure, but that will take months, if not years, for the truth to reveal itself.

“People’s ability to write something intelligent is going to be somewhat commoditized, because you can easily copy-paste from a ChatGPT bot.”

Here’s a nice generative AI prediction for you, which I think will happen. People’s ability to write something intelligent is going to be somewhat commoditized, because you can easily copy-paste from a ChatGPT bot or something like that. Authentically showing it through a video that is something we do today.

Oh, interesting.

When you want to showcase your soft skills, you can make a video. Let’s say you’re trying to get a sales rep job. How would you sell this product? You do a one-minute video of how you would sell. There’s a lot that comes across in video, that even if you’re running off a script, you can see live. That could work. For that, there’s nothing very unique about it, because people can have somebody else write their soft skills, but video would be a good element of showcasing it in a more authentic way.

Last one. I think this is one that you, Google, and basically every platform will be confronted with first. Write me a 2,000-word blog post about content marketing that I can just spam to LinkedIn all day and night.

It’s a great example. What’s going to happen really depends on your identity. Let’s say somebody is a scientist in the field of genome research. If you can show credibility to what you just shared through the work you’ve done and your experience, if you can add content to it and somehow attach it to the things you know about and work on, I think that’s actually really interesting.  There’s a community there to potentially challenge you and contribute back. 

I think this is where people talking about their craft would be amazing. Maybe they come up with some idea and they want to add more depth to it. Maybe they have a lot of potentially complex notions and ideas and they want to simplify them. This is where the combination of identity, what you’ve done, who you worked with, and what you have to say, your knowledge, will be a really important combination. Just sharing something and coming across as smart will not be enough. Showcasing it depends on your craft and the stuff you’ve worked on. I think your reputation will start to be more and more important in this world where you can generate things pretty easily like you’ve never been able to do before.

That sounds like a yes on the content marketing blog post. I cannot wait for this feature to roll out. Tomer, this was great. Thank you so much for talking to me. I love chatting with you about this stuff.

Thank you, Nilay.

Decoder with Nilay Patel /

A podcast from The Verge about big ideas and other problems.