Michael Eisenberg (00:01.998)

‍

I am thrilled to be here with my old friend, I think is a fair way to say it, and semi-colleague, because we've been around this tech industry for a long time, Amir Shevat. Amir, how are you?

‍

Amir Shevat (00:13.369)

‍

Thank you for having me. Yeah, I’m excited to be here.

‍

Michael Eisenberg (00:16.344)

‍

So this is gonna be an interesting episode of Invested. Amir and I know each other a long time. We've both been around the tech world for a while, but on very, very different paths in many ways. And so I was trying to remember, Amir, how did we first meet? Do you know?

‍

Amir Shevat (00:32.259)

‍

I think it's when I joined Google and I wanted to, I basically came to Google to try to work with developers, and I met you trying to figure out what's in it for investors in this ecosystem. Me and Eyal Miller, we were trying to map what it's in it for investors, what's in it for startups, what's in it for Israel. And you gave us a lot of insights about how to build a better program for Google for Startups.

‍

Michael Eisenberg (01:00.514)

‍

Oh! I didn’t even remember that. But anyway, take us back to the beginning of your career. Give us your whole background in a nutshell, and then we'll start digging in.

‍

Amir Shevat (01:08.855)

‍

Awesome. So I'm an engineer by trade. So I learned to code when I was a kid. Started with C++, and Logo, and all of that good stuff, and then moved to Java. Started at Microsoft, worked in a couple of startups before that. Then went to Microsoft, worked with developers on SharePoint and .NET. Then joined Google, worked on a startup program for Google, started in Israel–started as what my manager called “A bad idea, but you should start as part of your 20% if you really like startups.” And then turned into a global program for what's called Launchpad around the world, which was pretty interesting. 

‍

Then I moved to the U.S. 12 years ago to run the global program for startups, and then joined another startup myself called Slack. This was two weeks before they actually launched their platform, and grew that platform from 50 developers to 250,000 weekly active developers. Slack then IPO'd and got acquired by Salesforce. 

‍

Then I joined Twitch. Twitch was a YC company that was acquired by AWS by Amazon for a billion dollars. And I was the head of the API for Twitch. So my job was to integrate between Twitch and the game developers.

‍

Then created my own startup, and that was acquired by Jack Dorsey by Twitter. And so Jack brought me to open up the API for Twitter, and then Elon came and closed everything. And that was it.

‍

Michael Eisenberg (02:55.534)

‍

So, okay, this is the perfect introduction to exactly where I wanted to head. So a 20% project at Google to help Google work with developers out of Israel turns into Google Launchpad, which is working with developers all over the world, that turns into a career path in which, in company after company, now we should start with Microsoft before that, from Twitch to Twitter, Slack, et cetera, turns into like being–I don't want to say head of DevRel, developer relations, but in many cases, yes, and certainly building developer communities globally. So when people ask me about Amir Shevat, what I say is “Amir is the king of developers–relations.” You're like the grand pool ball.

‍

Amir Shevat (03:37.311)

‍

Thank you. Yes–not king of developers, maybe some experience in developer relations.

‍

Michael Eisenberg (03:43.682)

‍

All right, so what I'd like to dig into already very quickly is, what was different at the various companies you had, Google, Microsoft, Slack, Twitter, et cetera, in the role you had and in how you interacted with developers, and what you learned about developers, and what you learned about each of those organizations.

‍

Amir Shevat (04:05.881)

‍

So Microsoft was very, very good at working with developers. They understood the monetization that developers are seeking. So for every dollar that we were making on SharePoint, for example, developers working on SharePoint were making $9. So we understood that if you want to be successful, you need to build the right incentives and the right motivations for developers to build on top of your platform.

‍

Michael Eisenberg (04:32.994)

‍

Ben Thompson says that that's the definition of what a platform is, right? If your developers, if people who build on you make more than you do, it means you're a platform. Keep going, sorry.

‍

Amir Shevat (04:41.657)

That was exactly it. So there was very, very strong motivation for other people to promote SharePoint, because SharePoint was their bread and butter. So I think that was a very interesting insight for Microsoft. What Microsoft didn't do well was understand open source at the time. Because I remember I was walking with a Linux shirt one day, and somebody told me that that was a career-limiting shirt. 

‍

And you know what? Today it's not the case. Today everybody's talking open source at Microsoft. But at the time, Microsoft did not figure out how to work with the open source community. It felt competitive with the open source community. And I think that was a mistake at the time. They could have done a much better job at working with developers and understanding developer motivation. 

‍

The last thing I'll say about Microsoft is that they thought that money solves everything. I'll give you an example. We were working on Windows Mobile, and we were paying developers to build apps for Windows Mobile. And that was a mistake, because us paying developers to do something that is not useful for them and not useful for the user wasn't effective, because Windows Mobile was not effective, and developers built it just because we paid them money, not because they could get a lot of users or a lot of monetization capabilities. 

‍

So I think the understanding that you need to build an amazing product first and then build a developer community is something that a lot of people need to understand when they come into this business.

‍

Michael Eisenberg (06:19.638)

‍

And is that because developers are not, pardon the term, coin-operated? It's not what gets them going is the money and they'll build, it's kind of an incentive misalignment? 

‍

Amir Shevat (06:28.739)

‍

Exactly. So I think the key is that you need to have a product that people love. When I worked at Slack, Slack was a product that developers really, really loved. So I could have done a good job. I could have done a less good job, and the outcome would probably be developers loving and using Slack and its APIs. So I can't take a lot of credit for the amazing growth that happened. I probably did an okay job, but the fact that we had a very strong, amazing product at its core really drove developer love. So developers really care about working on something that they appreciate.

‍

Michael Eisenberg (07:06.766)

‍

What makes an amazing product for developers?

‍

Amir Shevat (07:09.943)

I think the first thing is that it needs to be a good product for humans. So developers are a subset of humans. So if the product is cool and fun, like Slack was at the time, then developers feel like they're joining a revolution. The same thing for Android. Android was open source and fun, and Google was a great logo at the time. So developers joined our community because they loved the feeling of being a part of belonging to something that they really cared about. 

‍

Same thing with Twitch. Twitch was super awesome because we run our developer community in TwitchCon, and you can see our favorite Twitch streamers together with working on the APIs with them. So fun and belonging is very important. Then there's like, what they really care about when they do their job, which is amazing documentation, very clear onboarding–so what I call “five minutes to hello world” - I want, from the minute I land on the developer site until I feel successful, I want to be able to be successful within five minutes. Very, very important rule in all my engagement with developers. Get them excited and successful and activated within five minutes, because if not, they'll move to the next application. And you can see that with OpenAI and Claude, and all the rest of the solutions, they do that very well. 

‍

Michael Eisenberg (08:35.374)

‍

I'm pretty struck by what you're saying because at least probably numbers-wise, Microsoft's got the largest developer community on the planet. Yet when I listen to you, I hear it's not because they love Microsoft, but because it's just a big developer ecosystem. But Twitch and Slack, these were beloved products. I'd ask, is that a fair characterization? And given what you said, how is it possible that nobody's undermined Microsoft's developer community?

‍

Amir Shevat (09:01.711)

‍

So I think Microsoft has changed. They had a lot of developer-negative sentiment, and I think they did a good job in changing that. I think Visual Studio is an amazing developer tool that developers really love. I think Azure is now much easier to use. When I worked on Azure, it was like 12% uptime. So developers would install something on Azure, it would not work 80% of the time.

‍

So now, I think they've nailed down the experience, they nailed down the open-sourceness of themselves, they nailed down an ecosystem of game developers which build on top of Windows. So I think there's a lot of renewed love for that platform. And I think you'd see the same thing with Amazon. Amazon started as a book company, like developers couldn't care less, I guess. And then they started releasing things that developers really cared about, which is, I don't want to care about servers. I just want to run things, and I want multiple services that are cloud-based. And the same thing, developers came and used it. But a mistake that I see is saying, hey, I'm going to have a developer community, and my product is not really good. Because it's like, you need to have a good product and then a developer community, not a developer community that creates a good product. That doesn't work that way.

‍

Michael Eisenberg (10:27.406)

Is that predictable or is that like, you know, the famous judge says, you know, “On profanity or pornography, you know it when I see it.” 

‍

Amir Shevat (10:34.593)

‍

I think you know it when you see it. I think you start, so developer community, I call it “herding cats,” because they come, and then you manage their community. It's not the other way around. You can tell them about an amazing thing, but usually, if the product is not good, they just don't care. So I've seen a lot of like, Hacker News launches, and Product Hunt launches that developers come and say, “Hey, we'll check it out.” And then the usage curve crashes. Why? Because the product is not that interesting for developers. So they'll be very open to new things, I think more than usual customers, but they'll really tell you what they think and vote with their usage patterns.

‍

Michael Eisenberg (11:20.216)

‍

So you've been at Microsoft, and Google, and Twitch, and Twitter, and Slack. What's the one DevRel job you wish you had had that you didn't have?

‍

Amir Shevat (11:29.521)

‍

That's a good question. I think Apple would be one of those. 

‍

Michael Eisenberg (11:32.662)

‍

Apple, interesting. 

‍

Amir Shevat (11:58.447)

‍

Because I think it's one of the biggest platforms in the world that drove a lot of monetization. I would try to change Apple's view of developers to be more open. The other one is what's happening now with OpenAI and Claude. If I was 10 years younger, I would just run

one of these platforms, because I think that's the next big amazing thing.

‍

Michael Eisenberg (12:03.512)

‍

Dude, you got another 10 years in you, what are you talking about?

‍

Amir Shevat (12:06.223)

‍

Yeah, but I want to invest in startups, and work with startups–

‍

Michael Eisenberg (12:09.102)

‍

We're coming to that soon.

‍

Amir Shevat 

‍

Yeah I want to invest in startups, and maybe have my own startup. I wouldn't want to do DevRel for another big company.

‍

Michael Eisenberg (12:18.402)

‍

At the five or so companies you worked at, which was the best experience?

‍

Amir Shevat (12:23.183)

‍

The best experience was, I think, two times. One of them is opening Israel Startup Program, because it was so unique and so amazing. And we had so many amazing partners, including you guys, that actually came–you saw the entire community come and celebrate innovation. That was amazing. And then taking it globally.

‍

Michael Eisenberg (12:43.15)

‍

That was at Google.

‍

Amir Shevat (12:49.421)

‍

That was at Google. Yeah, Google Campus. And the second one was definitely Slack. Slack I joined as DevRel number one. We didn't know anything about whether we're going to be successful or not. I took a 50% pay cut, and I joined a small startup. I had to go to San Francisco every day, and it blew up. If I had hair before I started, I wouldn't have hair at the end. It was so amazing to see, to be on that rocketship.

‍

Michael Eisenberg (13:11.694)

‍

And why'd you take the job? Like, it was so early–what'd you see?

‍

Amir Shevat (13:17.527)

‍

I saw a company whose product I really loved, which every person that I talked to had a feeling of excitement about. There's a notion of minimum viable product. I really liked the notion of ‘minimum lovable product.’ Because it generates a feeling that is really important. And I think Slack had that from the early stage, and just joining that was like everybody I worked with was happy, and was motivated, and was working together with zero politics. And I think that was the number one thing that drove me to that.

‍

Michael Eisenberg (13:52.942)

‍

How do you measure a minimum lovable product? How do you know it's gotten there?

‍

Amir Shevat (13:58.127)

‍

I think by retention. Minimum viable product–you look at growth, you don't look at behavior in the product. And I think retention and engagement are the two things that are the indicators for love.

‍

Michael Eisenberg (14:12.886)

‍

Interesting. So I need to ask you the obvious question now, which I'm sure you've been asked a hundred times at this point, which is like, AI is changing developers and changing software development pretty rapidly. The post saying, “Hey, I got Devin, I don't need juniors anymore.” So I'd love to hear two things from you. One is, how is DevRel going to change in the era of AI?

‍

And number two is, start to talk me through the future for developers right now.

‍

Amir Shevat (14:50.105)

‍

So the right answer is, I'm much better at predicting the past than the future. I'm very good at predicting the past. 

‍

Michael Eisenberg (14:57.42)

‍

That's not an option on this podcast.

‍

Amir Shevat (15:18.939)

‍

Okay, awesome. I think engineering will change. I think top engineering, the core, very hard things to build are still gonna be something that human engineers are gonna build. I think that will take another few years, between five and 10, for AI to be able to do hard engineering tasks. Things like, I want to scale to infinite, or I want something that is completely new that hasn't been done before. I think that's the place where engineers are going to shine. The second place where engineers are going to shine is reviewing and managing AI engineers. 

‍

I think we're going to see synthetic humans. Like AI is gonna collaborate with humans in the same way humans collaborate with other humans. And we're gonna have junior engineers that are AI, and senior engineers that are humans. That is for the hard things. For the not-hard things, I think we're gonna move from engineers to product managers, to people who can easily describe software.

‍

And I think we're going to move from coding to specing. I think we're going to write the specs. We're going to write the specifications and move to, more on the authoring stage than actually the implementation stage. We're going to say, “Hey, I want to build a CRM. I want to have this type of login. I want to have this type of user management. I want to have these entities. And AI will implement those specs.”

‍

Michael Eisenberg (16:45.144)

What's the skill set you think you need or need to learn to be that product manager of AI engineers to spec it out?

‍

Amir Shevat (16:53.101)

‍

You need to be very clearly articulated about the process and behavior of software, about user needs, about the way things look and behave, and to be able to iterate your experience about, like, you look at something–what needs to change? What exactly, what exact experiences do you want to create? And I think AI will be very bad at doing that. And I think humans will be very good at doing that.

‍

I think humans are pretty okay in understanding what they need and what other people need. And I think AI has pretty much failed until now in doing that.

‍

Michael Eisenberg (17:36.78)

‍

What should someone study to get to that kind of role where they kind of spec out and manage AI engineers? They have to learn computer science to know how software is built? They need to learn design? Like what do they need to do?

‍

Amir Shevat (17:47.789)

‍

I think design and business and software engineering in the next few years is still gonna be, probably a good idea. But the answer is like, the rapid change makes it so hard for junior people to enter the markets right now. Like I look at a lot of my friends who have kids who are now finishing college and a lot of them are really struggling to find junior roles, because most of the junior roles have been replaced or displaced already. In software for sure. 

‍

Michael Eisenberg (18:19.948)

‍

So the question I keep asking that no one has given me an answer to, but maybe since you are the king of developers you can answer it. How do you get senior engineers if they can't be junior engineers? Like how do you ever get there?

‍

Amir Shevat (18:36.417)

‍

I don't know. 

‍

Michael Eisenberg (18:38.934)

‍

If you don't know then nobody knows!

‍

Amir Shevat (19:06.221)

‍

I can tell you that I have a friend who goes to Miami every couple of years to find COBOL engineers, because a lot of the banks are still using a very, very old–there's no new COBOL engineers, or very few new COBOL engineers. So that guy goes and hires people from like, in their sixties, to go back to banks and solve bugs and problems.

‍

So we as humanity have a problem with knowledge gaps. It's just like, I don't know how to start a fire right now if I didn't have any lighter. I don't know how to create electricity again. I think as humanity, we create these gaps in our abilities. I don't know how to hunt, right? People used to say, “Hey, you have to hunt in order to have a living.” We don't know how to hunt anymore, but we still progress. So I think it’s going to be moving from engineering, from being a senior engineer to being a senior product manager. And I think you can be a senior project manager without being a senior engineer. And I think for the next few years, you still need to have junior engineers. You still need to have senior engineers. So we're going to have this class, but it's going to be much, much smaller. I think the amount of engineers that any company will need will be a lot smaller, which has a lot of consequences on SaaS, by the way, because all the seat-based models are getting broken by that.

‍

Michael Eisenberg (20:07.118)

‍

Yeah, I did a presentation in an annual meeting a few years ago called Peak SaaS. To be fair, I did not call the AI piece of it. I used the other piece of it, which was just a trivial business model. And if MBAs could pull the model forward, there's no alpha in the game anymore. But I think that's another, yet another reason–I think people will just roll their own software, you know, build their own stacks internally. I don't need any of these horizontal SaaS companies and I can kind of have a product manager and some AI engineers and maybe one senior engineer, and pull off 90 plus percent of what paying for Salesforce or somebody like that.

‍

Amir Shevat (20:44.297)

‍

There’s a bet in the Silicon Valley about the single CEO human building the first unicorn. So like, when will we have a single person, a single human managing an army of AIs, creating value at the level of a unicorn? And I think we're about five years from that. Probably five years.

‍

Michael Eisenberg (21:00.566)

‍

And so what's your bet? 

‍

Amir Shevat: 

‍

Five years. 

‍

Michael Eisenberg:

‍

Five years. My guess, by the way, is that in algorithmic trading, you could probably be done faster than that. But I'm not sure that creates a whole lot of value. It may create a whole lot of money, not a lot of value. So I want to go back to what I normally open with right now, because this relates to this developer conversation. What's your core value?

‍

Amir Shevat (21:11.151)

‍

My core happiness, maybe core value ties into that, is to enable other people to be successful. I think that is like a core thing that drove my career, helping other people to be successful.

‍

Michael Eisenberg (21:40.28)

‍

So we're co-invested in a company called daily.dev, which is a developer community where developers can stay up to date. And when you think about developers being replaced by AI, how do you think about the future of daily.dev in this regard?

‍

Amir Shevat (22:02.991)

‍

I think developers as people who create software is still gonna be there. I think it's just gonna be different types of developers. They're gonna be more the types that describe software than the type that code software. So they're gonna be much more effective. So it could be that you're gonna have, daily.dev is gonna have fewer but more senior software engineers. And I think the budget, the overall budget for software creation is just going to grow, because right now a lot of the pains that you have in enterprise, the answer to like, “Hey, we can't do that,” is, shrug, “We don't have resources to do it.” 

‍

And I think that's going to change. It's going to be, “Hey, we need to really do that. Let's make Joe the mini CEO of that application, and we'll do that type of prediction. For example, like look at all our sales, and do a different type of prediction. So I think that the need and usage of software is just going to grow. You're just going to have less people working at the boring parts. like building all the Windows applications, and making sure that all the UI looks okay, that's probably going to be AI. But defining the software itself, I think it's going to be human for a long time.

‍

Michael Eisenberg (23:19.308)

‍

And there will be more of those people in your view? 

‍

Amir Shevat (23:21.155)

‍

Yes, more people who define software.

‍

Michael Eisenberg:

‍

So this is like, this is close to Jevon’s paradox, right? The cost of it's going down, it becomes more ubiquitous and ends up being a lot more of it. The market ends up growing for these things. But so, I guess you're not in the camp that believes that there’s going to be massive layoffs from AI. Although you did mention before that your friends’ kids are not getting jobs.

‍

Amir Shevat (23:46.361)

‍

So I think there is gonna be displacement of jobs. So I think some jobs, like, you know, when I was a kid and I called the U.S., I went through an operator. That was a human that connected the calls. Before that, my parents used an actual operator, right? You called someone and like you say, “Hey, I want to talk to Joe.” And there was a person who connected that operator. That role completely vanished with the automatic call centers. 

‍

But if you look at the studies, they show that like 10 years later, the amount of unemployment, and that was like, three million people worked at that job. That was like a percent of the U.S. population. And the employment 10 years later was not impacted by that. So I think what you will see is new jobs, different jobs. It could be that a person who does, maybe, front end engineering is not going to be a role, but a full stack person that thinks about the entire application flow is definitely going to be more of that.

‍

Michael Eisenberg (24:49.646)

‍

Can you think of an example where, like, so you said your core value, and Nimrod kind of told me to ask you at least a version of this question, not exactly this question–Nimrod Kramer, the CEO of daily.dev, where your value is making other people successful has come into conflict, perhaps with what you do? And so I imagine that there's some sort of AI company that you're invested in or worked with that puts somebody out of a job or some innovation or some other case where that value conflicted.

‍

Amir Shevat (25:20.559)

‍

So it conflicted every Google I.O. Google I.O. is a big event. And every time you release a new set of things, you create an opportunity for 10 startups and kill another 10 startups. So I remember one of my painful experiences as a developer, an Israeli developer coming to me and said, “Hey, you just built a thing that kills my entire startup.” When you build a platform, sometimes you do things that actually–and I see that now a lot with AI, like OpenAI is releasing a feature that 10 startups pitched me that they're going to build that feature. And now it's a commodity built by the incumbent, right? So every time a big platform makes a change, it can potentially kill a lot of innovation, but create room for a lot of innovation. And working with that, like if you move from like the big picture to the small picture, to the human being being impacted by that, then it really hurts. I take it very personally when I have this founder that tells me, “Hey, I need to think about a new thing because my startup has just been demolished by what you've built.” And it hurts. 

‍

Michael Eisenberg (26:31.05)

‍

And how do you deal with it?

‍

Amir Shevat (26:33.763)

‍

How do you deal with it? You go back home. And my wife is amazing. 

‍

Michael Eisenberg (26:39.884)

‍

I've met her, she is, I agree.

‍

Amir Shevat 

‍

And I talk through this. Yeah, she's in my much better half. And you then help another 10 developers be successful. And you realize that like net-net, you could do things that are positive and negative, but if net-net at the end of the day, you help more than you destroy, then you probably do a good job.

‍

Michael Eisenberg (27:01.038)

‍

Before I move on to your seed investing, I do want to ask two questions. You were at Twitch, post-October 7th, there's been a lot of controversy at Twitch.

‍

How would you want to talk about that?

‍

Amir Shevat (27:16.815)

‍

So think Twitch, it's a failure of moderation. I don't think Twitch is intentionally doing this. I think Twitch is just not doing the right job in moderating, and having a core strength of meeting their terms and conditions where they are. I think there's a lot of hate happening on Twitch. And I think that if you read the terms and conditions, you cannot do that. 

‍

So like at the core, they're saying the right things, but they're not doing the right things. So I think very strict moderation in that case would have probably solved that. And they were walking a thin line, because there's two audiences really trying to push them in different directions. And it's hard for a corporate to play the judge–although we're seeing that now happening in the U.S. a lot–but doing proper moderation through your terms and conditions is probably the right move at the time.

‍

Michael Eisenberg (28:18.125)

‍

And so why are they so bad at moderation?

‍

Amir Shevat (28:22.063)

‍

First of all, moderation is a hard, hard problem. A lot of people have different types of understanding of what toxicity means, what free speech means. You see that as a struggle across all social networks. I think Twitch has a unique problem where by traits, gamers are more toxic than other players. The age range is usually 15 to 35 male, which tends to be more aggressive by nature. So I think the problem in Twitch is even more like that. And I think they just, it's a hard problem and they didn't figure it out. I think Bluesky is figuring it out better, because they think about the problem differently. But Twitch just used the same means to deal with a much bigger problem, and that didn't scale.

‍

Michael Eisenberg (29:18.766)

‍

Did they call you for advice? 

‍

Amir Shevat (29:20.985)

‍

Twitch? 

‍

Michael Eisenberg 

‍

Yeah.

‍

Amir Shevat: 

‍

No, they didn't call me for advice. I think Amazon is not very, does not call previous workers. That's not the way they run things. They also think that they might be doing a good job. At the end of the day, a lot of people are judged by KPIs that are not connected to toxicity. They look at engagement, they look at retention, they look at how many people are talking about the platform. And sometimes bad news is a good thing.

‍

Michael Eisenberg (29:52.078)

‍

And what would you have told them to do if they did call you?

‍

Amir Shevat (29:55.847)

‍

To make a stance. To say, “Hey, this is what we do.” It's either, “We're going to allow everything and live with it, or we're going to be very, very strict with our terms and service. And we're a gaming platform and just a gaming platform. Don't get politics in.” And then to stick with what you say. And what they did is, they zigzagged a little bit and that creates that–you lose trust, because when you say one thing and you do another thing, I start to not trust you anymore.

‍

Michael Eisenberg (30:27.15)

‍

It doesn't look like they lost users over it.

‍

Amir Shevat (30:30.295)

‍

I don't think they lost a lot of users for it. 

‍

Michael Eisenberg (30:31.81)

‍

Yeah, so they paid no price.

‍

Amir Shevat (30:34.543)

‍

That is a major problem, because the incentives are tied into retention and engagement, and they probably didn't move the needle on that.

‍

Michael Eisenberg (30:45.208)

‍

Yeah, so speaking of people with other approaches, you got fired by Elon Musk overnight at Twitter. 

‍

Amir Shevat (30:52.654)

‍

Yes.

‍

Michael Eisenberg: 

‍

What happened?

‍

Amir Shevat: 

‍

Elon did not think that the Twitter API was meriting investments. That is the reason. I think he wanted to do cuts. He did cuts across the board. And he didn't think that the API was worthwhile. I tried to talk to his team about the Everything Store, and about the ability of Twitter to be like a platform for developers. But it didn’t resonate, and I think that's okay. That's a decision that you could make. I think it's the wrong decision. I think if you look at extremely successful platforms, you need an open API, and that's why I invested in Bluesky and other companies like that. But if you don't think that API and developers are a part of your story, then that's okay.

‍

Michael Eisenberg (31:50.488)

‍

Judging, I guess, by your social media posts, you're not exactly into Elon. Is that fair?

‍

Amir Shevat (31:55.927)

‍

I think he cares about being out there more than he cares about the merit of things. I think he cuts things without thinking deeply about them. And I think he likes to be in the limelight with a lot of ego. And I don't like any of those traits. I think changes are necessary. I think minimal government is probably better than a very extravagant government. But if you cut things without thinking about it, at least at the government level, people die. And that is, for me, dangerous. putting more thoughts into this would probably be better. So if you own a company like Twitter, you can cut as many people as you like. You can look at it and you say, “Hey, I don't care about API. I don't care about moderation. Take 90% of those people out.” 

‍

That's okay, because you own the company. You don't own the U.S. So like cutting things without thinking deeply about the implications, I think, is a mistake. And I think you need to be more on the caution side. Even if it's a stupid, stupid thing, go through a process of reviewing it, thinking about it, and then cutting it. And what we're seeing now is more of like a show, “Hey, look at all these stupid people, look at all the things that I'm shining,” and I don't like that.

‍

Michael Eisenberg (33:26.062)

‍

It sounds like you were against the Twitter API cut. That seems to have worked, because Twitter's like at all time highs right now, as best as I can tell. And so, at least in my own head, he keeps seeming to win, even if it's an unorthodox style, to say the least.

‍

Amir Shevat (33:42.607)

‍

So the question is, what does it mean to win? So I don't know if you noticed, but there's a lot more antisemitism on Twitter. And when we had the API, we had a lot more studies from Israeli universities about using the API to study antisemitism, to map the people who are driving antisemitism. And antisemitism is just an example. You have other groups that are negative. But a lot of research was done using the API to try to fight these toxicities. So you might win because you're burning something, but I don't know if it's sustainable for the long run.

‍

Michael Eisenberg (34:24.462)

‍

Interesting. You think there's more antisemitism growing on Twitter than on the other social media platforms since October 7th?

‍

Amir Shevat (34:30.635)

‍

No, but I think there's more antisemitism growing on Twitter since Elon joined. That is my experience. I think that it's more unchecked, because there's no public scrutiny. And I think the public scrutiny was done through the API.

‍

Michael Eisenberg (34:36.545)

‍

Interesting. I guess he would argue that it's done now through Community Notes.

‍

Amir Shevat (34:51.407)

‍

I'm an avid member of Community Notes. I think Community Notes is an amazing product, but think how much more powerful it would be if it had an API. If you could do Community Notes automatically, if you could do Community Notes by running AI to see if something is okay or not. There's an area where you can double down on things using developers and APIs, which I think is not being used right now.

‍

Michael Eisenberg (35:20.718)

‍

Interesting. One of your founders, Guy Podjarny, told me to ask you why you went public with your critique of Elon.

‍

Amir Shevat (35:26.573)

‍

Why I went public….I think it was important for me to tell my team that it's okay to go against the bully. It comes back to my core of empowering people to be successful. And when I talked to my team, it felt that they were afraid of Elon. And I think being, regardless of what your opinion is, being afraid of somebody with power is wrong. And I think that if you, as a leader, your job doesn't end when you get fired. I think as a leader, you continue to care about your people after you leave. And I think that was my motivation to show my people that you could do this.

‍

Michael Eisenberg (36:12.814)

‍

So you made a career switch, after being the king of developers for all those years, to become a seed investor. Like, why?

‍

Amir Shevat (36:21.007)

‍

Turning to the dark side, I call it. 

‍

Michael Eisenberg (36:23.308)

‍

Hey, easy buddy.

‍

Amir Shevat 

Why? So a friend of mine called Dror Berman, from Innovation Endeavors told me–

‍

Michael Eisenberg (36:35.95)

‍

Eric Schmidt.

‍

Amir Shevat (36:47.861)

‍

Yeah, Eric Schmidt is one of the LPs. He told me, “Hey, if you want to really deeply understand the other side, try to do it.”

‍

And I started doing angel investments. And I did about 32 angel investments personally. And I found out a lot of things from the other side. I found out that like, founders really find it valuable when you help them at certain stages, very early stages. I can't help a startup at the later stage probably. Like if you're growing from 10 million to a hundred million, and you need someone to re-innovate your sales, I'm not the right person. But if you're thinking about core principles of your product in the early stage, I can probably do a lot of value. And if it's developer tools or developer platform, I can probably double down on that value. And then I said, like, “Why am I doing it with my own money at a small scale? Why not do it on a larger scale?” 

‍

And then I partnered with two amazing friends, and we created Darkmode, which is my fund. And we raised funds, and now we work with developers and startups, and it delights me. I meet nine startups every day. I deep dive into 12 out of those every month, and I invest in one. That's my funnel.

‍

Michael Eisenberg (38:05.966)

‍

Amazing. Of the 32 initial angel investments, how many were zeroes, how many were okay, and how many were winners?

‍

Amir Shevat (38:14.447)

‍

I would say about–and it's like five years, so it's still early–about five to 10% were zeroed. About 20% were growing, are growing what's called markup. I got more money than I invested in. About 20-25%, and the rest are on that trajectory. 

‍

On paper, it looks really good. Can I stop for a beef on our profession for one sec, if I may?

‍

Michael Eisenberg (38:54.158)

‍

Please!

‍

Amir Shevat (39:12.479)

‍

I think when you do product, it's really easy to know if you suck or not. You release something, and you immediately, people tell you, or usage tells you or growth tells you. Reality slaps you in the face really hard. When you're an investor, it really is a slower, much slower learning process. 

‍

So I invested in 32, maybe now 12 are doing really, really well, but they are doing well for their stage, right? And the ones that are doing okay, like if you look at Monday, it's an amazing company, but it struggled for a very long time, and then it went up. And so there's no learning in our profession, or not fast learning.

‍

So it takes a very long while to understand that you're either a very good investor or a very bad investor. And there's so many attributes, there's so much stochasticity, like randomness of the data, that it's really hard to know if you're good at your job or not.

‍

Michael Eisenberg (39:56.8)

‍

You have to fake it till you fake it.

‍

Amir Shevat (39:58.519)

‍

Yeah, exactly. So on paper, it looks like my angel portfolio is phenomenal. But on the actual data, I hope I am a good investor. I hope I'm a better partner to my startups and that I choose good people, mentsches, to work with. That is my thing.

‍

Michael Eisenberg (40:19.374)

‍

Is there any theme in the ones you ultimately decide to invest in? That's like a crazy funnel. 90, you see 180 investments a month, 180 pitches a month, and doing one. So what's the theme?

‍

Amir Shevat (40:34.947)

‍

Yeah, it's more like 90 if you look at weekends, but yeah, the theme is a team that is amazing. Like a team that could really execute, a team that has done this before, or a team that has worked together well. So like, first is the people. I think the people, their ability to hire great people, is very important. Their ability to think about problems differently, their ability to introspect is very important. So that's the first thing. The second thing is TAM, and can I help them? So like, is this a big enough of a domain? 

‍

And do I care deeply about it? And can I help them in that regard? They could be doing an amazing thing in a domain that I don't understand at all. And I would tell them, “Hey, listen, you probably need another investor that knows this thing out.” And the last thing is the technology is like, is there an ability for this technology to have a really strong moat that other people could not easily replicate? That's the three things that I'm looking at.

‍

Michael Eisenberg (41:41.816)

‍

So your friend, investee, co-investor, Micha Breakstone said I should ask you, “If I asked 100 entrepreneurs that you've mentored what they learned from you, what do think the most common answer would be?” 

‍

Amir Shevat (41:55.979)

‍

That's a good question. I tell a lot of people, don't work with jerks. Like I think relationships are important. So I think working with good people leads to good outcomes. So I think they would say that even if I didn't invest, I honestly tried to help them. And I think that returns a lot. And I think you talked about it a lot in like aspects of Judaism–an aspect of tech in general is like, helping other people, and it returns. So a lot of the times I help a founder and they connect me, and I didn't even invest in them, and they connect me to a company that I did invest in. So I'm deeply, I deeply believe in helping others. And I think companies helping each other even without the investor is even better. So, they would probably say that I at least try to help.

‍

Michael Eisenberg (42:53.49)

‍

But that's kind of like, “Hey, I want you to let me into your round, Mr. Hot Entrepreneur, because I'm going to try to help.” You must provide more value than that. So if you had to pinpoint the value you'd provide, what would it be?

‍

Amir Shevat (43:05.711)

‍

The value I provide? It’s that I've been in their shoes, and I've figured out hard things, and I work with them on these hard things before I invest. So I think, every meeting, even if it's a complete waste of time from an investment point of view, I really try to figure out one thing with a founder to try to solve one hard thing. And if you do that consistently, then they pull you into the investments. For example, with Guy, I really worked with Guy Podjarney and Micha on trying to solve hard things for them before I even talked about investment. And usually what happens is that they pull you into the investment, which is pretty amazing.

‍

Michael Eisenberg (44:06.862)

‍

So I want to read something, if I can, from May Piamenta, who is one of your investees. 

‍

Amir Shevat (44:11.871)

‍

Oh wow, now you're gonna embarrass me. My god.

‍

Michael Eisenberg (44:36.962)

‍

When she was on my podcast on the previous episode of Invested, she said the following. It was a huge shout out, Amir. It was actually when I decided I should interview for the podcast. 

‍

“So Amir is one of those people that you just really want to be around, and you know he'll be so unbiased and supportive no matter what. And Amir, throughout the pivot, [her company pivoted], every conversation I had with him, and I was freaking out, and I was like, ‘Amir, how am going to do it?’

‍

And he told me, ‘May, everything will be just fine, and I'm here for you.’ And Amir, you know, he's been leading amazing product teams. He's been around and working in Twitter and Slack and his record is incredible, but he's such an amazing person, and he always has time to help other founders, even if he's an investor or not. So yeah, I'd pick Amir.” 

‍

How come you didn't tell May in the middle of the pivot, “Yeah, listen,” in the middle of the pivot, “Hang it up. You know, it's not working. Let's do something new.” Instead, you told her, “Stick with it.”

‍

Amir Shevat (45:07.001)

‍

So first all, May is an amazing, amazing, amazing founder. She's like, so powerful and so much energy, and so much drive. And your interview with her was phenomenal. I really enjoyed listening to it. So I always like to work with people that I would work for. And she's definitely one of those people. I would work for her in a heartbeat. And that's one another aspect of the people. Would I work for this person is one of the questions that I ask myself.

‍

Michael Eisenberg (45:36.598)

‍

You know, Jamie Dimon, I quoted this in one of my books, the CEO of JP Morgan. I heard him say, he said to me, I was in a meeting with him, he had just gotten back from interviewing a bunch of people in different cities at JP Morgan, and he said the core question he asked himself was, “Would I want my daughter to work for them?” It's a great one.

‍

Amir Shevat (45:52.227)

‍

I love that. I'll phrase it in my words, which is like, if you have imposter syndrome all the time, that's a good signal. Because you work with so many people, and you say like, why are they working with me? Like they're so amazing. I really want to work with them. That's the signal that you're in the right place. Having a continuous imposter syndrome is a very good signal. 

‍

With May, I asked myself, what would I want an investor to actually tell me? Because I've been there, where May was, with a pivot. And an investor telling me to shut down would have been the worst thing that I would want to hear. And I think May has so much potential, and now she's killing it. Now it's amazing. Even if it wasn't amazing, telling her to try and be the person that she could really be, and build the things that she could really build, I would be...sinning all types of sins if I wouldn't tell her that.

‍

Michael Eisenberg (46:54.19)

‍

No, no, but sometimes the right advice is, hey, this dog doesn't hunt. You should shut it down, try something new. You're a great person, try something new. How'd you know that she should stay at it?

‍

Amir Shevat (47:03.001)

‍

Because I think Ron Gura told me that, like, how many runs do you still have in you? He asked me that when I told him that I was struggling. And he said, like,”Do you have another run or two in your soul? Can you do another try or two?” And it really made me think about how many runs do I have with me on this idea. With May, she was struggling, but it was clear she had another 10 runs in her. She could go and go at it. Sometimes you see a founder, and they're struggling, and they don't have any runs in them. They're like, tired. They don't like their founders. They don't like their teams. They're really, it's going to be really hard for them. So you need to see how many runs you still have in the founders that you're working on. And with May, it was like, it looked unlimited. She had so much energy to still solve that problem. And I think that was the right call.

‍

Michael Eisenberg (48:01.646)

‍

By the way, for those who have only tuned into the podcast, the Invested podcast recently, there's an episode with Ron Gura, which I think is probably the third or fourth episode of Invested, that I highly recommend you listen to. He's a special entrepreneur, and a special human being. It's worth going back to listen to that. 

‍

So you invest in Israeli founders and U.S. founders, and I'd love to hear from you what you think are the core differences between American founders and Israeli founders.

‍

Amir Shevat (48:30.745)

‍

I think Israeli founders really excel at technology, and product, and zero to one. American founders in general are very good at product and business and scale. And I think it's becoming blurry. And I think Israeli startups are becoming a lot better at distribution, and scale, and B2C and definitely American startups are being much, much stronger with even core technologies. 

‍

But at the end of the day, it's like, where is your superpower? Superpower for Israeli entrepreneurs is still chutzpah. To think about a problem in a very, very different way that is shocking to everyone else, but actually works. And for American companies, it’s like, how can I take an idea that is okay and really, through go to market and business development in this unfair market?

‍

Michael Eisenberg (49:35.576)

‍

Now you're in Austin right now, right? And you lived in Silicon Valley, right? And you lived in Israel, right?

‍

Amir Shevat (49:37.517)

Yes, I love Austin. Yes. I lived in Israel and I lived in New Zealand as well.

‍

Michael Eisenberg (49:44.738)

‍

And New Zealand. And how would you compare the tech ecosystems between Austin, Silicon Valley, and Israel? Leave New Zealand out.

‍

Amir Shevat (49:52.585)

‍

Yeah, there's not a lot of startups in New Zealand. Silicon Valley is an amazing place for innovation. Everywhere you go, you see people having a startup. There's a lot of availability of money and connection and opportunity. The pay-it-forward is really strong there. And I felt everybody was nice, but...coming into the Silicon Valley to work and to achieve something, and then to rotate. You see a lot of building strong relationships, at least with the people who are coming just for startups is hard because you build relationships and then they move on. They go back somewhere else, or they continue. So I love Silicon Valley, but it felt a little hard when every summer half of the friends left.

‍

I think Israel is an amazing innovation hub. I think Israeli entrepreneurs really, really help each other. They really work hard. I think the problem is that a lot of them think that the next step is leaving to the U.S., and a lot of them do that. And so you see startups succeeding and then moving. And I think that is changing slowly. 

‍

And I think it’s a good change, but you still see a lot of founders moving to the U.S. once they reach some level of success. I think the core thing with Austin is that you come here primarily to live. Like, you come here, you have a house, you're gonna stay here for a very long time. So the community is more based on people who are in your area that could help you, and you see much less churn in here. 

‍

And also Texans are much closer in spirit to Israelis than Californians. Like people say what they mean, they say it straight into your face, and they don't take bullshit. So Texans are much, much closer. Like my neighbors here are much closer to my Israeli neighbors than California.

‍

Michael Eisenberg (52:15.33)

‍

How big is the Israeli community now in Austin?

‍

Amir Shevat (52:18.511)

‍

I would say it's about 5,000 people at least.

‍

Michael Eisenberg (52:24.014)

‍

5,000 people, so like 1,000 families?

‍

Amir Shevat (52:26.659)

‍

At least a thousand families. So one of the things that I really brag about is that I lead the Austin Foodies community. So there's a Facebook group of Israelis mapping good food around Austin. And we have about 1200 people in that group. Only Israelis, only mapping restaurants. 

‍

Michael Eisenberg (52:50.99)

‍

And what are they looking for, hummus?

‍

Amir Shevat (52:56.207)

‍

Good hummus, good barbecue, good everything. Asian food here is amazing. We just had Michelin come through town and give a few restaurants Michelin stars. So everything is becoming more expensive. But we have a great community that is focused not just on tech but also on food, on industry, on a lot of other things which I really like. 

‍

Michael Eisenberg (53:21.56)

‍

Has October 7th, has the war changed anything for you?

‍

Amir Shevat (53:25.099)

‍

It changed everything. It brought in scars from the army. So I had an active army service. I wasn't in Gaza, I was in the West Bank, and a lot of the things that happened there came back. So on a personal level, it was really hard. Seeing my friends at my age go and serve in the army was really hard.

‍

I had a friend who had coffee with his son in Gaza, and just thinking about his wife, and what she thought of like having her son and her husband serve in the army together–that breaks my heart. It also changed my thinking about–I'm leaning towards a liberal, and seeing what the crazy in the liberal side we're talking about really changed my point of view from being a liberal to be a centrist. I think that the center in the U.S. is saying on both sides, both the right and the left, are pretty crazy. So I think it changed my point of view in many, many, many ways.

‍

Michael Eisenberg (54:43.278)

‍

And I guess almost a decade ago I tried to get you to come back to Israel. What are the odds of my getting you to come back to Israel now?

‍

Amir Shevat (54:50.351)

‍

So the word back, I think, is hard. I want to live sometimes in Israel, but I want to live around the world. I want to….my kids, my oldest is now finishing his university in Boulder, in Colorado, and my youngest wants to go to Europe. I think we're going to be in Israel, but I think it's going to be somewhere like living around the world. So living in Israel, and in Texas, and in many other places.

‍

Michael Eisenberg (55:20.782)

‍

Tell me this: I understand that you wrote a fantasy book, is that true?

‍

Amir Shevat (55:25.581)

‍

That is true. So I wrote two books that are technical, with O'Reilly and like very technical, somewhat boring. This is for my soul. I wrote a trilogy about magic coming back into the world. So the theory is that we had magic in the past. And now how do us, as humans, treat the world when there's other alpha species? When you come in, and you have to talk to an elf, and the elf could use magic to get what they want, right? How are we as humans gonna deal with other alpha species?

‍

And I think we're gonna act very badly. And somebody said it's a reflection of us working with AI. And it could be, I didn't think about that at the time, but I think as humanity we'll need to deal with working with other species that are not our own, that could be as intelligent and as powerful as us. And the trilogy is about that.

‍

Michael Eisenberg (56:29.55)

‍

Do you believe in extraterrestrial life?

‍

Amir Shevat (56:32.175)

‍

I think there's an absolute possibility. 100% there's extraterrestrial life. The question is, have we seen them? Have they visited? I'm not sure. But if you look at the amount of stars out there, the chances that there is no extraterrestrial life is almost zero.

‍

Michael Eisenberg (56:49.376)

‍

Almost zero. And so when people reflect on your book they think about extraterrestrial life, AI, and what do you think about?

‍

Amir Shevat (56:58.477)

‍

I actually think about magic. I think about like, how are we going to deal with the fact that you go hunting, and you run into a unicorn? And you can’t shoot a unicorn! If you do shoot a unicorn, they will come after you, and you're not going to win that war. So I think that we need to have some sort of humbleness, which we're lacking as humanity and dealing with powerful things. And I think we're seeing that with humanity right now with like dealing with death, dealing with getting older, there's still things that we don't control. And I think we're not dealing with that with grace. I think we're struggling in that area. And I tried to deal with that in my book.

‍

Michael Eisenberg (57:38.2)

‍

I find it interesting just as a religious person. I think that's what religion deals with in large measure, right? It gives you the humility because there's God and brings you down. And how you deal with death, getting older, and life is very religious in those ways. And you took it to an area of magic.

‍

Amir Shevat (57:56.355)

‍

I think that religion solves that problem really, really well, because you always have someone above you. And I think people with less religious faith or ambiguity there really deal with the struggle of like, what do I do? Assuming that I'm the top honcho, I don't know how to deal with it. And I deal with it really poorly usually.

‍

Michael Eisenberg (58:19.552)

‍

Interesting. Last question, which I like to close with. Who's someone in our space that you would like to say something kind about and you think should be getting more attention?

‍

Amir Shevat (58:54.255)

‍

I think all the entrepreneurs that are struggling and getting all the no's from us... okay.

‍

Michael Eisenberg (59:00.458)

‍

Come on, give me a name. That's like a generic answer. That's like a media training answer.

‍

Amir Shevat (59:06.389)

‍

I've been media-trained.

‍

Michael Eisenberg (59:08.086)

‍

I know, I can tell.

‍

Amir Shevat (59:11.631)

‍

So if I had to give a shoutout–can I have two? Probably May Piamenta is one of my shoutouts. I think she's a freaking amazing entrepreneur that works extremely hard despite all consequences to do something really really good for humanity, and working with her makes you want to be better yourself. So that's one person. 

‍

I think Micha Brickstone is another one. I think you have somebody who's a really successful entrepreneur who is a true mensch, who takes people–like every time you meet Micha, he makes you feel that you are better. And I love people who have zero thoughts about themselves and 100% thoughts about others. And I think that's like, I love his selflessness.

‍

Michael Eisenberg (01:00:10.166)

I'm gonna ask you one more question and then we'll wrap up, which is, what's a strongly held opinion that you've changed in the last year?

‍

Amir Shevat (01:00:24.429)

‍

I think my political views have changed in the last year. I think I became a lot more, I used to be much more radical, and maybe it's age. But I think I've used to be a lot more radical, and I find myself being a lot more centered. Like I think that the fact that us as humanity have two sides and we're always battling each other, it's actually not a good thing. I don't know how to solve that, but I think the edges are controlling the center, and I think that's wrong, and I think that's becoming more and more of a problem. And I want to think more about how to change that. I used to be one of the sides, and I don't want to be a side anymore. I want to be working together with people.

‍

Michael Eisenberg (01:01:10.03)

‍

By the way, that's what my recent book is about, is how the polls have taken over politics and democracy in many places, and economic policy and affirmative action and many other things have caused that, in my opinion. 

‍

But Amir, thank you for taking the time to do Invested with us. And for those of you listening, if you enjoyed what you heard from Amir and the other podcast guests, please rate us five stars on Spotify, Apple Podcasts, wherever you listen to your podcasts. Please subscribe to our channel on YouTube. You'll really enjoy the shorts and the videos there, and Amir's dog in the background. He'll only appear on the video part of the podcast, and not if you listen to us on audio. Amir Shevat, thank you for sharing your wisdom with us. 

‍

Amir Shevat (01:01:51.247)

‍

Thank you for having me. That was a great conversation.

‍