如何在所有人之前发现非凡人才

The most important skill in this industry, or perhaps in the world, is your ability to identify things before the market.

There are so many people in the world who we want to meet, who also want to meet us, and it's just an information problem. What's more impressive to you: someone who went to Stanford, or someone who is followed by Patrick Collison, Elon Musk, and Marc Andreessen?

A lot of people in Silicon Valley have a pretty high bar for who they're willing to publicly endorse. A lot of what we do is finding awesome people and connecting them to our companies, otherwise getting them into the a16z network.

And it allows us to really extend all that network, all that value, all that trust, and really that reputation that transfers and builds over time.

There's this Naval quote I love: “Be so good that people do business through you.” In fact, if I cosign 2 people, they might be able to meet and otherwise build something in a way that I enabled without even having to know about it.

But it's always hard to match the needs and the skills at exactly the right time.

What we've enabled is this. We're super excited to announce Cosign, a product that's around professional reputation for the startup community, and really just trying to be the most comprehensive directory of companies and people.

1. Twitter Versus College Credentials

I want to start with a story, which is my friend David Perell told me. He said, “My Twitter account is more valuable to me than my college education. Even though I paid a quarter mil, whatever it is, my Twitter has given me more value to my career, and also personally.”

Now, we of course just launched our own school, so we're big fans of higher education. But I thought that story was profound because, of course, Twitter's been free for so long, and what he got at was that college is a bundle of education, network, and credential.

And social life.

Yes, of course. Some of that is hard to recreate on Twitter, but exactly.

The education—you learn from having conversations. The network, of course, is that you meet people and discover people. People kind of see it as proof of work: whether they're interesting, whether they're funny, et cetera. If they build something cool, you can discover them.

But I think the credential thing is really interesting, because here's a thought experiment: What's more impressive to you, someone who went to Stanford or someone who is followed by Patrick Collison, Elon Musk, Marc Andreessen, et cetera? I think most people would say the latter, right?

This kind of peer-to-peer credential is, implicitly—although people also do this explicitly—a cosign. It's an endorsement that this person is worth tracking. It's a public signal that this person is worth following. We think there are lots of other ways to express that signal and create this kind of reputational benefit that these people get, which is really the exhaust of it.

This person isn't even trying to give them reputational benefit. They're just trying to read what they write, and the exhaust is this endorsement. If you're followed by these people, now your doors are open. People also want to read what you say. They want to work with you. They want to invest in you. They want to have you be part of their team.

There's a benefit to the receiver, but there's also a benefit to the giver. If you're the first person to discover someone, it's kind of how Silicon Valley operates: you're going to be able to hire them quicker, and you'll be able to fund them quicker. The cool thing about Silicon Valley is you don't know who the next big person is going to be, and sometimes they're 18 or don't come from a name-brand school.

We're all in this, a little bit of a race to discover talent before anybody else and get credit. Zach Frankel's done an amazing job, as so many people have, of building a reputation for discovering people.

The benefit talent gets is that they think, “Oh, I can get kingmade by them.” If Josh Elman or Olivia invest in my company, that's a signal to the market. Angel investing, or investing from a venture firm, is one way to express a cosign or a signal. But a lot of people aren't founders, and a lot of people aren't investors, so we think there are lots of other ways to express this signal.

2. The Three Reputation Signals

I'm really curious: As you've been building Cosign, what are the kinds of endorsements you hope to get from the platform?

We want every profile to have at least 3. One is: Who shaped your career?

Wow.

We think this sort of mentor economy—this is what Silicon Valley runs on—and it also helps both sides. If I was shaped by my mentor, if Josh Elman's my mentor, that means that he is putting his reputation behind me, which is actually true.

Thank you.

Josh saw me as a young person and spent a bunch of time with me, and that's a signal to other people that I'm worth other people spending time with, too. If you respect Josh, you respect me by association. It's also a benefit to you to discover talent early and to be known. People want to work with you if you help them.

The second is: Who would you work with anywhere or in the trenches? It's hard to tell from existing platforms who you were really in the trenches with. Who have you really built bonds with? Who would you go to war with? Who, no matter what you're doing, would you want to work with? That's a much more granular signal than whether I follow them, which could be any number of things.

The third is: Person to watch. Who do you think is going to be really emergent or ascendant? We rank-order these, too.

The platform's just started, so most of it is fair game. In the same way that to build a great investor profile takes many years, but then you've got the credit for doing so, we think you can do that around people, too. You can say, “I think Olivia's going to be one of the greats. I think she's a person to watch.”

It's actually funny: I did this experiment 7 years ago where I emailed a few hundred people—you were definitely one of them—and asked them for their person to watch. I was just checking today. One of them was Russell Kaplan, who's now the president at Cognition, and he was a 22-year-old engineer at Tesla. So there's actually signal in that, too.

Those are some of the ones. We're also excited for people to do it for companies, too.

What are the companies that you think, right at seed, are going to be great? Because, of course, if you're able to, you can express your conviction with your money, but not everyone's an investor. This is almost like fantasy investing or social capital investing.

I love this idea. I feel like I've long felt this, but as I've invested now for almost a decade, I believe it more and more every week, every day, every month. The most valuable data in all of Silicon Valley is who has conviction in whom, and who can give you a real read on a person for so many decisions: Who do you invest in? Who do you hire? What company do you join?

There's such information asymmetry right now. Our job, investing every day, is to ping people, saying, “Hey, do you know anyone who worked with this person?” We do it every day, and we still, I would say, have to work really hard to find that signal.

Absolutely.

How can someone who is maybe evaluating among 5 startup job offers get that level of information? You just can't, especially if you're an entry-level engineer, and yet that could be the difference between joining a complete rocket ship and a company that isn't doing as well.

I know the data has to be handled carefully around who says what about whom, but I think it will be. There's a massive, massive opportunity to add so much value to people's lives and careers.

Talk a bit about this, because you think a lot about talent discovery.

Yes.

Talk about the newsletter you do, because I think it's incredible stuff.

My twin sister Justine also works here at Andreessen on our infrastructure team. We've both been in venture for basically a decade now, and in our very first venture role, we started a newsletter that we ran every Sunday for, I think, 6½ or 7 years.

I was a subscriber.

Thank you so much.

Yeah.

It was basically just aggregating startup job postings, and the thesis was that we both went to Stanford.

We ran a startup incubator at Stanford, and even before we were working at a venture firm, we had no way to differentiate between all these startups and founders—to know who was legitimate, who wasn't, and who was the most exciting. So we were like, “Okay, we have some inside information now. Let's curate a list of the companies that we think are worth joining for young people.”

It really built up a community because I think people appreciated it. You can't get that information outside of Silicon Valley.

Mm-hmm.

We had a network of scouts on campuses who read it, and some of those people are now founders that Andreessen has invested in. They're off running fantastic companies in fantastic places. So that was the less scalable version of this kind of network, and I'm excited for more people to have the ability to put their name behind ideas.

Yeah.

And one thing I've noticed is that as it gets easier and easier to start companies, and as it gets noisier and noisier out there, it's actually getting harder for people to do what you're describing. So this kind of signal is more important than ever. If you think about the average student graduating from Stanford today, they're looking at the startup market, and they have to rely on the signals given by firms who are investing, but even before that, signals given by people like you and hundreds of others we hope to empower here. An example I love to point to on the site is Packy McCormick, who has always had a lot of very good signal around deep tech and vertical integrators in particular. Packy has 37 companies that he's chosen and says, “These are the companies if you want to get into this space.” There are many, many more behind them, myself included. In a way, he's co-signing some of his reputation against that company. He's saying, “I trust this company. I'm putting my money where my mouth is. I'm writing about them, and I think that if you wanted to start your career off there...” So it becomes a sort of living directory of companies and people worth paying attention to.

And free. Crunchbase and PitchBook cost money, et cetera. They're great products. They have to be businesses. But we have the luxury of being able to offer this for the startup community. We want it to be the most valuable place where people can see who invested in what, what Olivia's set of deals is, and reliable company information: who the co-founders are, who the employees are, and what products they shipped.

Yeah.

They're great products. But we have the luxury of being able to just really offer this for the startup community. We want it to be the most valuable place where people can see who invested in what, what Olivia's set of deals is, and what the reliable company information is, like who the co-founders are, who are employees, and what products they shipped. And one thing we've decided is that no product can be all things to all people. There are trade-offs, and we're going to be positive. This is going to be a place where people brag. It's going to be accurate, but, you know, how often do you go to someone's LinkedIn or text someone and say, “Hey, you're connected to someone on LinkedIn. Can you tell me about them?” You're like, “Ah, I don't really know them.”

We're going to have much more granular information to make sure there's real signal there. David, this is something that we experimented with 7 years ago and decided not to pursue because we got busy with other stuff. I'm curious what has always struck you about this idea in terms of the need for it or what we're really trying to do here.

3. Finding Talent Before the Market

One thing I've always believed is that the most important skill in this town, in this industry, or perhaps in the world is your ability to identify things before the market. Those things, if you're an investor, are companies. If you're a hiring manager or a founder, they're people. Sometimes they're products. If you're in charge of procurement for a growing company, maybe you need to identify which product is better before the market.

But there's no way of recording that. There's no hipster leaderboard that says, “This person found this coffee shop first.” Angel investors, of course, want to be non-consensus and right, and they're rewarded for doing so. I can invest in your pre-seed round. If I'm right, then it's lucrative for me.

One of the things that's always drawn me back was the idea that I can invest my personal social capital in you. I am very willing to make an introduction for you. If I really, really believe in someone and no one else has heard of them, I can make an introduction to them—Josh Elman, for example. He will hopefully take that introduction because he sees me putting my reputation behind it. He has no idea who this person is, but he knows that David thinks highly enough of this person to take a chance on them.

To me, the thing that stood out the first time around, and the thing that we're really trying to lean on as we come back, is that when I co-sign someone, I take it seriously. I want that person to be able to take the co-sign I've given them, and they might be able to go straight to Josh and say, “Hey, David co-signed me.” Josh might see that and reply to that cold DM. In fact, that lubricates the wheels of social capital that could empower a lot more people to achieve a lot more.

There's this Naval quote I love from a long time ago: “Be so good that people do business through you.” If I co-sign 2 people, they might be able to meet and otherwise build something, hire each other, or invest in each other in a way that I enabled without even having to know about it, which is very exciting.

4. Turning Interest Into Intros

Yeah. It's really interesting. Related to that, I want to talk about more features of the product. I was talking about the more granular ways of expressing reputation. One of the best events I ever went to was this SV Angel conference in 2018. I think they do it once a year or something like that.

Before I went to the event, they had this list of 150 people that they sent out. Everybody was a baller, and they said, “Swipe who you want to meet—who you want an intro to.” I swiped on almost everybody. I was up and coming, you remember, Josh, and eager to meet great people.

Basically, if they swiped right on me, we would get an automatic intro, and it would be like, “You guys have full context. You should chat.” I ended up meeting with most of those people, and they became really good friends and collaborators. That event, of course, couldn't have allowed me to meet 40 people and have deep conversations with them, but it provided the context for us to become friends and collaborators.

I say that to say that there are so many people in the world who we want to meet who also want to meet us. It's just an information problem. If you can reduce the friction to make that introduction—because right now I have to DM them, but I have to think about something to say. Are they interested? I don't know. This is just passive: Here are people I would meet.

Now what we've enabled is this private signal that's actually a bit more granular. It's not just, “I want to meet them,” though there is that. It's also, “I would fund them if they start a company,” or, “I would hire them if they consider leaving.”

There are a lot of people I would hire, but I can't reach out to all of them. It might be inappropriate, I'm too busy, and I would only want to talk to them if they want to join me or whatever. Having more granularity around professional intent, passively, means there's no risk, because there's some cost in sending a DM. You might feel rejected, or it's time, or whatever it is. If they also express interest, then it's an introduction with high intent that I think can unlock a lot of good stuff.

Yeah. That conference example is so interesting because I feel like the other takeaway for me is that we can all be people matchers now—

Yeah.

—with a product like CoSign. I think about it in all of our jobs: We meet so many people every day. I often meet someone and think, “They are awesome. They have this really specific skill set.” Then I meet someone else 3 months later who's looking for that skill set, and I have to figure out how to remember this person, check if they still want the intro, and make the intro.

So having that store of information somewhere that's more persistent and can work across time is really, really interesting.

I mean, I agree with that completely. So much of what we like to do, both as investors and as people around Silicon Valley, is make these connections, but it's always hard to match the needs and the skills exactly—

the right time, remember that, and be able to codify this. Here are the people I think are truly great designers, or the people I think are truly great PR people who can help you through a crisis, or the people who have really helped me through a career decision that I was making and are the mentors that I go to. Being able to actually endorse that now makes it available so that when you're looking for that—

Yeah.

Being able to endorse that means you don't just have to go, “Josh, who can help me think through this decision?” I have to go through and see who I've recently talked to, who's fresh in my mind that I can think of. But I can really go back and say, “Hey, Josh, I'm looking through those cosigns, and you said that this person really helped you in this one situation. Do you think they'd be open to helping me?”

It allows us to really extend all that network and all that value and all that trust, and yet it's still a way that you're trusting me to try to make that introduction. It's not that any of us want to just say, “Hey, everything's open to everybody.” It's really still part of the trusted community and that reputation that transfers and builds over time.

So building on something Erik said, the private signal is one core piece of it, and I think what we're talking about here is, in a way, the public signal. So we've built the 2 of them. The private signal—I mean, honestly, the product is so fun. You can go rabbit-holing around. It's a bit like Wikipedia, clicking from link to link, to person, to profile, to company, to investor, and back again.

As you go, you're collecting. Maybe it's Pinterest: I'm going to put this person on my personal, private list of people that I want to work with in the future, and I'm going to put this person on my public list of designers that I rate. Every time a founder comes to me and says, “David, I need a new designer. I need a growth guy. I need an engineer,” I can say, “Do I consult my private signals, or do I send them my public list?”

Hopefully everyone has this sort of habit: if I really believe in someone, I'm putting them on my private list. It's utility both ways. It's telling them that I believe in them. It's telling their network that I believe in them. It's also something I can use. I can turn around and give it to a founder or someone I respect, and I can shortcut the work that you're saying you have to do every time someone asks a question.

What's so interesting about that to me as well is that there's often a timing discrepancy that happens in networks. I'm like, “Oh my gosh, I know this person has been trying to get their startup off the ground. This other startup would love to merge with them or welcome them to the team—something like that—if the timing is right.”

But if I go to that founder and say, “Hey, are you thinking about exiting your company?” that's a big risk. It could be offensive or insulting, or maybe the company is working and they want to keep going with it. The idea of this two-sided network, where there's information on both sides and they can match when the time is right, when they indicate it, is really, really, really great.

On both hires, acquisitions, job changes—tons of things.

Totally. Yeah, that's so fascinating. The company-to-company piece, I didn't even think about.

Yeah.

I would buy this company.

Yeah.

Yeah.

Yeah, the—

We'll get the product features shipped tomorrow.

Yeah, exactly.

Yeah.

I think we just extended the roadmap.

Yes.

Yeah, that happens. There are a couple of things I want to add to that. We thought a lot about whether there should be a feed and, if so, what should be on the feed, whether it's free-form, et cetera. Josh helped us really nail down the focus of the feed to be around questions and announcements.

Announcements are so interesting because we think a way to differentiate there is to basically make reputation durable. A lot of reputation is built on social media, but it's ephemeral. When Josh joined the firm, he wanted me to focus on other people, but you're such a good example. A lot of people came out of the woodwork who'd been working with him for a long time but maybe hadn't publicly said anything recently because they didn't need to, and they said, “Wow, I really endorse Josh. You should work with Josh. Josh helped me in my career.”

I said that. Nikita said that. A bunch of other people said that. Then the next day happened, and some people saw that and some people didn't see it. But that type of stuff is so valuable and should exist on someone's profile, like the first time you're looking them up.

For us, any announcement is going to be tied to their durable reputation. When companies fundraise, when people switch jobs, these are opportunities for cosigns, endorsements, et cetera. On the question side, it's really what we were talking about earlier: who's the best comms person for this? Who's the best angel investor in fintech? Who are people I should work with for this?

It's this kind of very specific question about some person, company, or product you're looking to either surface to yourself or learn more about. In that way, it will also build durable reputation, because you get added to a list through that.

No, and I think this idea of bringing those ephemeral moments and conversations that happen on Twitter and other social media—or X and other social media forums—all the time is really powerful. There's so much that happens, and it's wonderful in the moment, and then it gets lost.

I think this idea of building a profile—not a profile that is just what you say about yourself, but one that lets other people be positive and help shine a light on your strengths—I think is really, really powerful. I see a lot of these for you and the great work you're doing here. I see a lot of these for Olivia and David.

We get these comments, and it's just so special, and you feel so lucky to have people you've helped in the past be able to say those things about you now. But then again, it just gets lost.

Mm-hmm.

I think this idea of tracking those cosigns and really making it feel like—

Totally.

—this fundamental network, the web that we are—the web of Silicon Valley, of the tech industry—is really powerful.

I'll give you an example. Do you know the meme where the guy's walking into a party with stick figures, and he's like, “They don't even know that…”? You get a baller cosign or whatever, and you're like, “It's not public.” You know, Marc Andreessen wrote a blog post about me joining the firm.

Yeah.

That's pretty dope. I just want the first thing when you think about me to be Marc Andreessen and joining the firm.

Yeah.

Professionally, it is.

Yeah.

There are all these moments that happen. If we can make them more legible, you can more accurately represent someone.

Yeah.

Honestly, it wasn't until we hooked up the internet firehose to this product that I realized how many of those moments there are. You look at the main feed, and every day there are fundraising announcements and talent transitions.

Yeah.

Thanks to our good friends at X, we're plugged into a lot of those, and we're taking, whenever someone quote-tweets an announcement or replies, a genuinely thoughtful, positive comment. Your point earlier about positivity—those are pulled through, and they're really coming from a person. They're on the profile the person comes from, and they go to a person.

And the reason why it's all positive is because products have to win on some dimension, and this product is going to win on people using it to show off—to show off in a good way. “Hey, this is what I'm proud of. This represents the full manifestation of what I'm capable of on a personal level and on a company level.”

Yeah.

And I really think it's about celebration.

Yes.

So much of what we do—there's so much hard work that goes into making every single product that happens.

Yes.

It's the toil of somebody's love, their idea, the code they wrote, or maybe that they worked on with their AI agent now.

Yeah.

But putting that out in the world is a very vulnerable moment. It's very much, “Do you like what I've tried to make? Do you—”

“Are you going to use it?” I think we need to celebrate that effort and what goes into those moments, and all the hard work that we do, as opposed to—there are plenty of places for critique, and we see all that all the time, too.

Yeah.

But I think that doesn’t need to be recorded for posterity in the same way.

5. Celebrating the Hidden Builders

One thing we were talking about offline that’s interesting is that there are a lot of awards and a lot of shine on founders, justifiably so, obviously. Also, a lot of other people build the companies. A lot more people are critical to those companies, and they’re unknown or not celebrated to the same degree. This is an opportunity to do that as well.

Totally. I think even just positive signal is going to be incredibly informative, because I think a lot of people in Silicon Valley have a pretty high bar for who they’re willing to publicly endorse in any way. When I think about the number of intro requests that come in and how many of those I actually bridge, I think it’s great.

Yeah.

Because if you endorse someone who’s not great, that reflects poorly on you.

Yeah.

Exactly. Yeah.

So there’s a natural reputation-saving mechanism.

Yeah. And it’s a whole thing about whether you share something that you haven’t even played with. Have you even read the article before you even tweet it?

Yeah.

And I think the key is, you’ll get to know pretty quickly who’s just spouting things and celebrating everything that could be celebrated, versus, “You know what? I’ve actually played with this. I’ve spent time with it. I’ve heard the backstory. I’ve thought about the backstory.”

Yeah.

And I just want to double down on one comment you made about this going beyond the founders. I think the founders who take that risk to start a company from nothing—that’s such an incredible thing—but it takes a village to build every one of these companies. It takes so many great designers, engineers, marketers, and product managers, and finance. There are so many parts of every single company that have to get built.

Celebrating that work, when it comes from a specific team or individual, and being able to have that be part of this kind of endorsement and co-signing, I think, is going to be incredibly powerful beyond just the people that we talk about most of the time.

Well, tell a story, or perhaps nudge Erik to tell a story. One of the first times I actually heard of Erik, and then went on to meet Erik, was at the RISE Awards—the 2015, 2016 RISE Awards, thereabouts. So, over a decade ago now. It was around this thesis that founders get a lot of publicity and a lot of credit, but it’s actually the people who are the operators behind the scenes.

Do you want to tell the story? What were the RISE Awards?

Yeah. The RISE Awards were this series of awards meant to be for everybody else—for the engineers, the designers, and the product managers. I believe Josh was a judge. We had thousands of submissions and discovered and selected some great people because of it, and we’re going to bring that back in some capacity.

There’s this phrase: “Fear not the guy who has 1,000 moves, but the guy who does 1 move 1,000 times.” There’s the samurai or whatever. The RISE Awards co-sign is an example; a lot of these ideas have been in the ether for us for a long time.

Yeah.

Yeah.

Yeah, yeah.

And in fact, the thing that stands out about it is that a lot of the people who were RISE Award recipients got to know each other and built lifelong relationships through the retreats they attended, through the groups that you added them to, and things like that. So there’s something around that connection, too.

Yeah.

It’s really interesting, because I feel like we often meet people who are spinning out of great companies to start their own companies. Often, if you get really deep—if you call everyone they’ve ever worked with—it’s like, wow, this person who was, I don’t know, engineer number 4, actually was a huge determinant in whether this company was successful or not.

But to your point earlier about legibility, that’s not something that they can publicize. They’re not going to be the face of every article about the company.

Yeah.

And yet there are so many of these people who I would love to see get to leave that company, if they go start a new company or go to another one, and carry that accomplishment with them.

Yeah.

They can’t really do that right now.

Totally.

Yeah. We’re making the reputation market more efficient—basically, more liquid. For example, there are some people who have an amazing reputation, but there isn’t awareness. Not a lot of people know that they have a great reputation.

It’s more liquid in the sense that, if it were known that they had a great reputation, a lot of people would want to hire them or fund them, and they should get those opportunities as a result. So it’s basically trying to fix that information asymmetry.

I want to return to what you were saying earlier, though, because it’s really interesting. A candidate, or someone looking for a new job, will sit you down and say, “Hey, what company should I join?” There are a lot of people who wish they could do that. But if you’ve got your list of companies they should join, they can kind of do business through you. It’s like, Olivia stamped these companies. They can reach out.

And similarly, for the ones you do meet with, it would be cool if you could just press a button and it would go to all your companies instead of having to—

Mm.

—think through it. Because there are some that you might want to share privately, just to keep to your companies or whatever it is.

But then there are also companies that are like, “Hey, I want a head of comms. Who in my network knows a head of comms? Who should I even talk to?” To the extent that people are co-signing heads of comms or this kind of thing, it can facilitate much more matching.

Yeah, totally. I feel like even for investors, the most up-to-date profile of the companies they care about is maybe if they tag them all in their X bio, and maybe if they put them on LinkedIn. Otherwise, you have to go to the firm’s page and see if they connect them to investments.

If I could do that without ruining the reach and the audience, I would be tweeting about all my companies all day, every day.

Yeah.

Promoting them.

Yeah.

“Go work for them. They just did this amazing thing.” But you can’t do that.

Yeah.

And I love the idea of a place where that kind of thing lives—

Yeah, totally.

—to see at any time.

Yeah. Be shameless.

Yes.

It is interesting. Let’s talk about existing platforms for a second. X is obviously such a great platform, and the exhaust of it is so valuable to so many people. But there are also times when I’m like, “Man, I wish I could just download my follower graph. I wish I could just see that a few thousand people liked this thing about joining my company.”

Why is it so difficult? It’s not built for this use case of discovering people to work for or just getting more signal. Okay, you follow me, but do you want to work for me? Do you want to invest in me? I don’t know. What do you want? What do we want to do here?

And then LinkedIn is such a valuable utility. We’re on it all the time, and yet everyone hates it. Obviously—

Aw.

We love LinkedIn.

We love LinkedIn.

Yeah, yeah. We love LinkedIn.

Yeah.

And Josh used to work there.

Yeah.

LinkedIn is one of the most iconic products of all time. I remember what Reid famously defended when they got acquired, because people were like, “What? Only 26 billion?” It was like, “That’s a great outcome.” It was an iconic product, and yet people have frustrations with it.

Maybe let’s—well, even before the feed, I see people say the feed is cringe or whatever, but even before the feed, I feel like people were like, “I wish there was just more that we could do here.” So why don’t you talk about that?

Yeah, look, I worked at LinkedIn back in 2004 and 2005, when we were just—

Starting to form the network. Reid Hoffman, the founder, used to talk about it very much as having 3 stages of the company. First is growth: we really have to figure out how to make sure we have the largest professional network and give everybody value from just being part of the network, even if they have nothing to do. For a lot of users, signing up, making your profile, and not doing anything else for a while unless something comes to you was a perfectly good usage.

Then we talked about usage. We really had to start building products on top of the network. The real thing was that the network is just the substrate; we need to go build a real job service. I helped launch the first LinkedIn Jobs. We had to go build services: find the best PR professional, find those things. Then we’d build revenue.

So it was kind of the—I called it GRR: growth, usage, revenue. We really had to figure out what those platforms and products were going to be on top of the network.

Now, the challenge was that you built the network, and it was very single-signal: are we connected? In the early days, the beauty of that was the idea that we really know each other. I’ll vouch for Olivia, and if David is asking for an intro to me, and Olivia says, “David’s really good,” will you take this intro? That’s great.

And it was even better if you imagine this whole table. David would ask Olivia for an intro and say, “Hey, can you introduce me to Erik?” Then Olivia would say, “Josh, David’s my really good friend. It looks like you know Erik. Would you be willing to pass him on?” This idea of a trusted graph—I’m so happy to connect with you, and I’ll pass you on to somebody else—was the important condition.

But over time, it started to fade from that strength of connection to, “Yeah, we met once at a party. We high-fived. We had a good email exchange.” All of a sudden, everybody got into collecting. The challenge with collecting that network is that it makes it much harder to figure out where the real signal is.

On the other side, it still is the iconic representation of who each of us is: the professional profile. But when we think about how to use our professional profile, it’s only really good if it adds value.

One of the things I remember when we were first building LinkedIn Jobs was that the goal was to take a current job service where someone just applies and gets a résumé, sees a job listing, and goes, “How do we use the power of your professional network to make it better?” Reid would push us a lot, and he’d say, “What is a social résumé?”

When a job application comes in, it wasn’t just, “Here’s who this person is.” It was, “Here’s who this person is, and here’s who else in your company might also know more about who the person is. Here’s who else might have worked at similar companies they worked at in the past,” or, “Here’s who knows people really well who also worked at those companies.”

So you wouldn’t just get this flat résumé; you’d get a résumé annotated by who else can help you reach that person. That was this big change: making job applications socially enhanced.

And then when you’re looking at jobs, you might just see a flat job and think, “That job looks really interesting at that company.” We said, “No, we’re going to put the hiring manager on the job. Their profile is going to be on the job.” All of a sudden, instead of a faceless job that you’re just applying to and throwing your résumé over the wall, it’s, “Hey, you’re going to get hired by this person. And, by the way, here are the ways that you might have connections to that person, or you might have connections to other people who work at that company.”

So again, it’s about how you use your network to do those things. Those were the beginning epiphanies of how to really rethink something as simple as a job application and job site with this social network.

I think, again, what has been missing, though, is the set of things that happen ephemerally through X, like you were talking about—the set of things that we all have, but we don’t figure out how to strengthen and put into LinkedIn anymore. The true connection, the reason that we were in the trenches together: we worked on this thing; you bailed me out of a crisis.

Yeah.

When we can have that information in the network, it’s going to allow all these use cases to happen so much better. When you get to focus just on making sure the network’s right and getting all that usage, I think Cosine can really open up so many opportunities.

Yeah. It feels like there’s almost a hierarchy of connections, and LinkedIn, for good reason—I feel like we’re all saying it’s an iconic product—I’m an hourly active user.

It’s making a ton of revenue. They’re clearly doing fine.

They’re doing great.

But LinkedIn treats almost every type of connection the same, especially as products like endorsements on LinkedIn haven’t really taken off. For example, 90% of the time when I say to someone, “Hey, I saw you’re connected to this person on LinkedIn. Can you intro me?” I’ll get, “Hmm, I don’t think I know them. Maybe I met them at a party. I’m not sure.”

So there’s the question of whether you actually know the person, whether you worked at the same company, whether you worked at the same company at the same time, whether you were on the same team or adjacent teams that worked together, or whether you managed them. All of those are treated as one connection or no connection now, and I love the idea for Cosine of being able to break that into the granularity that it deserves, because it’s so much more useful.

6. Building The Social CV

Seems like the greatest time in history to be building products that can accumulate context that was otherwise unavailable. You say the social CV—I love that idea. I think, Olivia, in particular, you at the firm have been thinking a lot about talent and the social space, and agents are an interesting way to accumulate a lot of context around a person.

I think a lot of the things that we’re building here are related to that. Could you talk more about that? What does the future hold?

Totally.

A lot of what we do, both on our investing teams and our operating teams, is finding awesome people, investing in them, connecting them to our companies, and otherwise getting them into the a16z network. Right now, a lot of this is scraping these existing profiles and networks, and then doing a lot of manual work to see what the actual strength of the connection is and how we can make this happen.

I think that kind of manual work, which is truly external-signal-based and a binary signal versus a really granular signal, is only going to get harder. Recently, for example, I posted on LinkedIn, “Hey, I’d love to help my portfolio companies hire. If you want to work at a startup, fill out this Google Form.” When it works, it’s almost harder because I get 2,000 results, right?

It’s incredible.

That’s going to happen to every person who posts a job or any kind of opportunity. It’s going to be 2,000 now, and it’s going to be 10,000 a month from now as browser usage gets better and more people use browsers. It’s going to be 100,000.

So how do you, even using Astra at maximum strength, sift and sort? In an age of endless outreach and interest with AI, the scarce thing is the human endorsement, the human conviction, or the human knowledge. That’s where I get especially excited, because I think Cosine is going to be valuable to so many people in the community, and also very valuable to the work that we do and what we try to do.

It’s interesting. We’ve gotten to a place where there are only broadcasting networks to the whole world or group chats with 50 people.

Yeah.

There’s not really this 10,000-person, much more vetted community, and that’s where we are to start and grow slowly. Basically, you put out a post and you’ll just be more confident in the signal that comes in.

Going back to LinkedIn for a second, some things that I wish it had are—remember MySpace Top 8? I want that for business. Who are you in the trenches with? Who do you talk to regularly? Who are your mutuals? How do you know them? How did you meet them? What do they think about you?

There’s more context on the relationship and on people’s reputation. As you said, LinkedIn treats connections all the same, and you don’t really get context on the relationship or on people’s reputation.

LinkedIn is about personal identity. We hope that Cosine is about professional reputation. LinkedIn is about, as they say, people you know. Twitter is about people who are interested in knowing you, and we hope that Cosine delivers both in a more granular way.

At the same time, to give props to LinkedIn, it started 20-plus years ago, and no one has even come close to competing with it in any capacity. I remember Charles Hudson, a great investor, had a blog post about 12 years ago titled, “Where are the LinkedIn competitors? It’s time to disrupt.”

“I don't like the product for XYZ reasons. It doesn't do this.” And we've seen tons of them. We backed them at the firm individually, and people have joined crypto. All sorts of different incentives have emerged to break LinkedIn, and it's just fascinating.

I was joking when I said everyone hates it, but people use it obsessively, and yet they have low NPS on it or something. They know that a better version can exist, and yet it's been impossible to disrupt.

But I think the real thing with LinkedIn is that the NPS is really high for the people who aren't actively in the flow, in the searching, and in the community. LinkedIn has always served 2 audiences. One is—I call them—the people who want to find people, and the other is people who might want to be found at some point in their lives.

Yeah.

When you think about it, all of us would love to be found for the right opportunity if it comes to us. A lot of us might be happy. We used to talk about the mildly happily employed.

There are a lot of people who like what they're doing. They're not in any urgent need to go do something else. But if the right thing popped up, whether it was from someone you know or someone you don't know but it matched all your interests, you might be like, “Hmm, should I actually think about that?” And I might consider it. It still may not be right for you in life at the time.

Yeah.

But for those mildly happily employed people, they aren't constantly checking LinkedIn. If they're an engineer, they're not checking LinkedIn every day for new people or new things. If you're in sales, BD, venture, or investing, you're using it all the time. But that was what was really special.

I think moving tens and tens and tens or hundreds of millions—I think there are over a billion people now—of people who are happily employed and maybe open to the right opportunity, that's the kind of network effect that really sticks.

Yeah.

You think a lot about what a 20-year network effect is. Look at LinkedIn. I say this a lot to new companies: As you're thinking about something new, how do you think about the kind of network effect that could be so powerful for 20 years? Even if some people wish there was a better product, you can't move the masses over.

Yeah. We have looked at and invested in some attempted LinkedIn killers, and I think what they've done a remarkable job with is that the switching costs are just very high, both because they're famously good at bot detection.

That too.

So you can't really scrape it in real time. There are whole businesses that are built off of scraping LinkedIn and selling that data. But it's also because it's such a tight and dense network that, ultimately, if it's something as important as your career, you're going to want to be on the platform where everyone else is.

That being said, I think there's a massive opportunity for a new platform when there is new information or new value to be gleaned or shared, which is very much high-quality endorsements and high-conviction people.

It's generational shifts, and also niche—the hyper-niche that needs its own LinkedIn. For example, GitHub is arguably a better LinkedIn for engineers than LinkedIn itself.

I'd ask: Is there an opportunity to say, “What is the emergent niche here?” Is it the individual who's trying to build their reputation as a talent scout? They're trying to be a founder who wants to break through. Is there a world for a better LinkedIn that represents their taste in people?

What gets me excited about Cosine isn't that it has to even be a better LinkedIn.

Yeah, that's what I said.

It's that it really pulls together all this information that's already latently happening in our industry.

Silicon Valley and the tech industry are so special for this kind of pay-it-forward mentality, for this belief in somebody—the stuff we talked about at the very beginning. It's the belief in a person who may not have had a previously strong reputation but has an insight or an idea, and that alone can gather a bunch of people and a bunch of momentum and change the world.

That's what makes this so special. I think these platforms like LinkedIn are serving so many people now across so many industries and so many domains. I think there's room for something that's really special for the community that we have and that combines the things happening ephemerally with the things happening professionally, with the depth that's going to be great.

If Cosine works great and it still only serves the tech community, that's a huge win.

Oh, 100%.

Other communities and other industries figuring out how to take part in this, too, and bringing some of that Silicon Valley energy to more places, that's even better. But the pay-it-forward mentality and the idea that every great idea can come from somebody who wasn't previously known but truly has a unique insight—that's what I hope this captures, too.

Eric from Benchmark had this comment on the Patrick O'Shaughnessy podcast, where Patrick was asking him, “What do you think about every category?” And he's like, “It's all going to work. Everything is going to be huge—not every company, but just as an index, every category.” I think we've come to a similar conclusion.

Similarly, when people say, “Does the world need another podcast?” There are too many podcasts, and yet another great podcast comes in and it doesn't kill the others. It just all works. Everything that's great works.

Similarly, we just launched a school. Stanford's going to be fine. Harvard's going to be fine, and our thing's going to be great.

Yeah.

Similarly, LinkedIn's going to be great. It's going to keep doing its thing. We're inspired by them in so many ways. We're also inspired by AngelList.

Ten years ago, when they owned the investor graph, it was for the startup community and was really focused around companies, investors, and talent. They made a series of other bets that then made them focus on other parts of that or deprecate the free public social network that used to exist, which we think is an opportunity to bring back.

We're also inspired, in terms of how we're going to do this, by Wikipedia. One, people want Wikipedia profiles; they're aspirational. But two, we're creating profiles for people. That's where we're starting, and we think people are going to be excited about it because we're bragging for them, on their behalf, in an exclusive way.

If they don't want it, we'll take it off. But you're going to see your profile. It's going to be fleshed out with people giving you compliments and also expressing their interest in funding you or working with you—even before you go to the profile.

That's one of the ways in which we're differentiating. Only because of AI can we generate all of this information on people and recreate a professional graph.

But if we didn't go that third-party route, then we would have to get people to sign up. It's a cold-start problem. The place where people go to update their profiles is the most reliable. Often people say, “What's the daily behavior?” Well, we're brute-forcing it by covering every sort of transition. We'll also have reliable profiles with more reputation and benefits.

Totally. I'm curious if you agree, but for consumer products—

Mm.

—the easiest way to acquire users is to tell them that someone else is saying something about them on the product. Especially if it's somebody—it's the Nikita Bier “your friends just answered this question about you” strategy, which worked so well 3 times. It wasn't possible before AI for a network as big as Silicon Valley, and now it is, which is awesome.

Someone called me today the Nikita Bier of founder communities.

It's a real compliment.

That's a true compliment.

Yeah, yeah.

Yeah, so it's worth pointing out: Everything that's being pulled in is coming from public sources. It's what people have posted in a variety of different places. It's being aggregated.

There are a lot of funding announcements; they're the ones that are being announced. So there's a lot of really high-quality information spread across the publicly available internet, and it's uniquely compiled and structured, we think, in one place.

And Erik talks about the AngelList example. Very early in my career, AngelList was probably the first thing. It was like the wedge that I took into Silicon Valley at all. It was the ability to see which companies, which geographies, and which people at those companies.

For a long time, it felt like it could be the LinkedIn for the little niche of people who were ambitious in technology. And then, as Erik said, it moved away from that for a number of reasons. But that's been missing. It's a place that you can go as a founder and say, “Which angel investors do I want to get to know over the next 6 months before my round?” It's a public utility.

It's something that's worth pointing out here. We're putting this out into the world. It's like a thing we'd like to build with people. It's not something that's being presented as a fully finished, perfect product. It's an invitation—an invitation to everybody who wants to come in and edit and submit an update to their own or someone else's profile. It's a real ask: let's build this thing together for everyone.

Yeah.

And submit things to people's profiles, cosign the things you endorse about people. And then, even if somebody has feedback, cosign that feedback so that it gets back to David and the team in order to keep improving things.

Yeah, and just to end on that higher-level note, a lot of the industry, but also other industries, run on cosigns. Paul Graham famously cosigned Sam Altman around 2009, I think. He said the 5 best entrepreneurs were Steve Jobs, 3 other people, and Sam Altman.

At the time, Sam Altman hadn't had a huge exit yet, but that really gave him such a reputational boost. Him running YC, et cetera, is critical to him going on to build one of the most important companies of all time. And there are so many examples of people—Ben Horowitz and Ali from Databricks, for example—who got their cosign from someone older and more senior in the industry and were able to rise up because of it.

Ryan Hoover, who was the founder of Product Hunt, cosigned Erik—

Yes.

He brought Erik on to become one of the leaders in building the community for Product Hunt. It was such a great endorsement of you early in your career and your potential. I mean, look at what you've become since.

Oh, 100%.

No, I'm massively grateful to him. It's so fascinating because one of the things that makes Silicon Valley so special is how much of a meritocracy it is, and how you don't necessarily have to pay your dues for 20 years before you're taken seriously.

One of the most important intellectuals right now is Dwarkesh. He's 25 years old. He's never had a job, as far as I'm aware. He just came to San Francisco, started podcasting, and I don't know if it was Patrick Collison or Marc Andreessen, but a number of people went on his podcast early, cosigned him, and then said, “This is a place where Zuck and others can go on.”

One of the most important founders is Michael Truell, a young guy in his mid to late 20s. One of the most important people, Leopold Aschenbrenner, is, I think, 24 or 25 as well. And so, if you're talented and smart in this industry, someone will discover you, cosign you, and make your career.

Mm.

And so we hope that happens on our platform.

While we're talking about social capital and positive-sum mindsets, I think there's a really important critique of this that I wanted to talk about, where someone might rightfully say, “Why would I share this cosign? This is my alpha. I know that person's good. Why would I tell you they're good?”

If you take a very long-term view of this, you can say, “Sure, if that person's raising a pre-seed round right now and you want to do it, don't cosign them right now. But the person you want to cosign is the one that's raising a pre-seed round 2 years from now, 5 years from now.”

And when they do, they remember you. They say, “Yeah, the reason I'm able to raise this pre-seed round is Olivia cosigned me 2 years ago, and in doing so, she helped me meet my co-founder and she helped me meet the 5 people who were my early believers in my product, and that's why I'm here today. And of course I'm going to call her first for my pre-seed round.”

If you turn it around that way and take a very long-term mindset, this is such a long-term play.

100%.

Right?

And the benefits just accrue to you, in the sense that people remember, by an order of magnitude more, the people who believed in them first.

Mm.

Right?

Yeah.

Brian Chesky, when he goes public with Airbnb, who does he write in the blog post? He writes YC, Paul Graham, his first investor.

Mm.

He's had so much investment all throughout.

Yeah.

So much more money.

Yeah.

So much more time into this company, but you remember the first person. And so, if you are the first person, especially publicly, that's where the value is, because they will remember you for it, and then more people will come to you.

Yeah.

I'm so excited. Let's wrap on that. Thank you so much.

Thank you.

Thank you.