(0:00) Jason tees up today’s segments!
(1:21) Breaking down the Reddit situation: Why 2000’s-era web companies relied on third-party apps built on their API 
(3:10) Origin and current state of the Reddit revolt, CEO Steve Huffman’s AMA and comments
(10:27) Jason’s hot takes on how to solve the third-party app API pricing issue, Reddit’s IPO price, and more
(17:00) LinkedIn Jobs - Post your first job for free at https://linkedin.com/twist
(18:28) Brad Gerstner joins live from Angel Summit! Importance on non-consensus conviction, first-party research, how Brad first invested in Facebook
(29:00) Vanta - Get $1000 off your SOC 2 at https://vanta.com/twist
(30:06) Brad breaks down the greatness in Zuckerberg and Meta’s reaction to his “Time to Get Fit” letter
(34:08) Current state of the markets
(38:05) iConnections - Get 20% off iConnections Miami 2024 event at http://iconnections.io/twist
(39:48) Predicting what the AI innovation cycle will look like when compared to past cycles: pc, internet, mobile, cloud; why now is the most important time for AI investing

Subscribe to This Week in Startups on Apple: https://rb.gy/v19fcp

FOLLOW Brad: https://twitter.com/altcap
FOLLOW Jason: https://linktr.ee/calacanis

Thanks to our partners:
(17:00) LinkedIn Jobs - Post your first job for free at https://linkedin.com/twist
(29:00) Vanta - Get $1000 off your SOC 2 at https://vanta.com/twist
(38:05) iConnections - Get 20% off iConnections Miami 2024 event at http://iconnections.io/twist

Listen here:
Apple: https://podcasts.apple.com/us/podcast/this-week-in-startups-audio/id315114957
Spotify: https://open.spotify.com/show/6ULQ0ewYf5zmsDgBchlkr9
Overcast: https://overcast.fm/itunes315114957/this-week-in-startups-audio

More from us:
Twitter: https://twitter.com/twistartups
Instagram: https://www.instagram.com/twistartups
Official site: https://thisweekinstartups.com

Subscribe to our YouTube to watch all full episodes:
https://www.youtube.com/channel/UCkkhmBWfS7pILYIk0izkc3A?sub_confirmation=1

Subscribe to TWiST Clips for all the best moments:
https://www.youtube.com/channel/UCS7tJlcUA6PzVHEMo-X7ddg?sub_confirmation=1

#startups #entrepreneurship #investing #angelinvesting #tech #news #business

First up on the show before my spectacular fireside chat with the fifth bestie Brad Gerstner.

I want to talk about this Reddit situation.

As you may have heard, Reddit, the most famous and successful forum or community site ever created in the history of the internet, let that sink in is currently dealing with a strike a strike by the members and the community managers of their own website.

This Week in Startups is brought to you by LinkedIn jobs.

A business is only as strong as its people.

Post your first job for free at LinkedIn.com slash twist Vanta compliance and security deal breaker for startups to win new business.

Vanta makes it easy for companies to get a SOC2 report fast.

Twist listeners can get $1, 000 off for a limited time at vanta.com slash twist.

And iConnections is a platform to connect and meet with elite capital allocators through their online platform and bespoke events.

The first 25 VC funds to sign up for iConnections Miami 2024 event in January of next year will receive a 20% discount.

Head to iConnections.io slash twist to sign up today.

Well, the Reddit moderators or for short mods are on strike because third party developers are revolting after Reddit said they were going to charge for its API.

Now, for background, what this means is there are apps in the App Store when you type in Reddit, and you'll see there's a half dozen apps, maybe a dozen.

In the early days of the internet, there was a culture of API's where people could build around other services.

So if you had access to Google's API, you could use their search results or Yahoo had 1.

And it really started at Yahoo, they had something called search monkey, and they inspired people to build apps and use their data.

And so to Twitter, Twitter said, Hey, here's the API have added.

Well, they were constrained in the 2004 to 2010 era, when a lot of these sites were launched, there wasn't a lot of money available to startups.

And those startups had very few developers and developers were a precious resource.

So they said, Hey, if you want to build something like surmise was a search engine for Twitter, If you want to build tweet deck, a desktop client for professional Twitter users, the developers would make it and the understanding was, hey, you're giving us free developer resources, you own your code, you're using our data, but you're making our service generally more valuable.

And then once the services became big, like Twitter, Reddit, etc, they decided we're going to deprecate these services because they became cash rich, and they wanted to control the experience with users.

And doing this means that there's going to be some hard feelings and the feelings have gotten people a little sensitive.

So as of Monday evening, 8000 subreddits, in other words, topics have gone dark.

This means either they went private, or they stopped allowing new posts.

Well, the way Reddit works is you can create your own subreddit.

Now there is this illusion that you own that subreddit.

We've had people on this podcast who created Wall Street bets and other things who have felt bad about having their subreddits essentially taken away from them, which if you're building your house on somebody else's land, at some point, they can take the land back.

And that's essentially what's happening here.

The people who are the mods are saying you can't post and you can't even see the content here.

Now they could, it's gonna be very hard for Reddit with this many subreddits going dark for them to fire everybody, it'd be essentially like turning off their entire community.

Remember, back in April, we covered that the New York Times did a profile of Reddit CEO Steve Huffman, really brilliant entrepreneur, who co founded Reddit with Alexis Ohanian.

The title of the article was Reddit wants to get paid for helping to teach big AI systems.

And basically, the article is about how Reddit was going to start charging companies that use its API to train their large language models.

And we've been talking about LLM had nauseam on this program.

You know, in order to guess that third or fourth word in a sentence, or try to essentially come up with good ideas, you have to have a model, what's 1 of the best models in the world of human content creation?

Those serve a certain niche core is 1, Twitter is 1, Reddit's 1.

And then of course, you have things like open source projects, all of those data sets are what modern LLM are built on top of but you need permission.

And people didn't get permission from Reddit.

So Reddit said, Hey, this is a way for us to make money.

And it's quite interesting, because for my whole career as an angel investor, and a venture capitalist now with a fund for the past, I think 12 years, I've been doing this, people have always told me, Hey, we're people, somebody is gonna want to buy our data.

But actually, there is a market for data, we think it's possible that these language models will have enough free data that they can use that they won't have to pay Reddit.

But if it was a reasonable fee, why wouldn't you to make your large language model that much smarter For somebody like Microsoft, open AI, closed AI, or Google, there's no price that Reddit can charge, that's going to be too expensive, or that Twitter can charge.

But if you made a free version of a mobile app, which these third party developers have done, well, you probably are not making any money or a small amount of money.

So Reddit started in 2005, and didn't launch its own mobile app until 2016.

Remember, I said before, these startups were typically constrained in terms of their revenue.

So they went 11 years before launching their own mobile app.

And Reddit users have relied on third party apps, alien blue was 1 Apollo, those apps would not be possible without the API and Reddit, of course, got more users to use the service on their mobile phones, thanks to those folks.

But the API, while it's great for people to tinker with the platform and to create these things, when you start looking at how successful these apps have become, well, Reddit wants to charge something around $12, 500 for every 50 million requests.

But at that usage 0.1 Reddit app, Apollo says they would need to pay 1.7 million a month.

So now, Apollo, and a bunch of other popular third party apps, they announced they would be shutting down June 30.

Last Friday, CEO Steve Huffman did an AMA and AMA and ask me anything on the API issue.

And he made 14 replies in the thread, generally, really poorly received about the API pricing.

The folks on Reddit, I mean, they're largely from anonymous accounts.

And the currency of Reddit is being dank, a word for being dark slash funny slash irreverent, let's say.

And so Steve's Reddit username has been spaz, SPEZ for since the beginning.

How do you address the concerns of users who feel that Reddit has become increasingly profit driven and less focused on community engagement?

Alright, so he responds quite reasonably will continue to be profit driven until profits arrive.

Reddit is trying to be profit driven, especially because they have an IPO coming up.

Unlike some of the 3 p apps, third party apps, we are not profitable.

So in another exchange, username Sri you can said the announcement of the API changes felt very abrupt to your New York Times initially just this was in response to the rise of chat GPT and other large language models LLMs.

Were these API changes or in the pipeline prior to chat GPT?

Or is this really a knee jerk response to cut off get a cut of LM training data to which Huffman replied, yes, and no.

2 things happened, happened at the same time the LLM explosion put all Reddit data use at the forefront and our continuing efforts to reining costs to make read it self sustaining put a spotlight on the 10s of millions of dollars it costs us annually to support the 3 p apps third party apps.

He's being completely honest with the community.

And when you look at these third party apps, well, they are siphoning off the top users.

Well, the Reddit app is not as good as these apps.

These apps have been in at it longer, these apps only have to do 1 thing, Reddit has to do everything.

So they're always going to be slightly better than Reddit's app, or chances are, if you have this, you're going to have 2 or 3 apps that are better than Reddit's.

What does that mean the top users then are going to go to these apps.

So for Reddit, which wants to make money from advertising, and perhaps some of these apps like the Twitter third party apps, which have been deprecated as well, in a very similar fashion, those apps were taking out ads on in Twitter's case.

So not only were they removing ads, they were taking the top users away from the platform who advertisers covet the most.

So a lot of this has to do with the top users.

And, you know, having insights into them.

Remember, these third party apps are getting all the data of all the usage.

As I said, 8000 subreddits with over 28, 000 moderators have gone dark, some subreddits we shutting down indefinitely, whatever that means.

I think that means until they run out of fish sticks and hot pockets, but you know, they could be back sooner.

3 reasons why this is super interesting to me, you know, a lot of Reddit values created by these unpaid moderators, and they are doing subreddits, really, because they're passionate, and they found friends.

So it's affiliation, in some cases, it's recognition, you know, being on the leaderboards, some of them could be on the edges, you know, drunk on their power, but they don't get to participate in the profits.

I think the Reddit members are doing this because they know they can.

And it's interesting because AI is forcing their hands.

Now, there are a couple of hot takes I have.

And here's a nice middle ground that I'm sure Steve has thought of, and maybe he'll deploy is I would offer to either buy the third party apps and hire those folks who become developers and then maintain them.

So you have maybe different apps that, you know, have different focuses and different themes.

So 1 could be for power users who are using the site all day, 1 could be for mods, etc.

But that could be a path forward where they buy these folks out, give them some equity, turn them into, you know, advocates and promoters of Reddit as opposed to detractors.

So if you think about these folks, they're the ultimate promoters of Reddit, who have now been turned into detractors, and you want to flip that.

So how do you flip it, Either you buy them and give them equity in the company and say carry on or how about saying this to the different Apollos of the world in these apps, we'll split subscription with you revenue with you 5050.

So you can be an official third party app, we want 50% of your revenue, you know, we get control of your subscription data, whatever, we have co access to it, we get data about the users, and then we'll give you 30% of the revenue that people see.

And we'll know that because we'll know the number of views you get in your app, everybody wins revenue take, and you make it a 5 year deal or something.

So now, these folks are going to make more money.

In other words, you let them eat, you let them get their beaks wet.

And then if you think about the top mods, at Reddit, I had offered back in the day, when I was at AOL, we had a clone to dig and Reddit were inspired by FARC, and which is really the and slash dot, which are really origins of these news plus community sites, where these community news sites, And we just hired, I think 20 people from Reddit and 20 people from dig and we just offered them $1, 000 a month.

So I went to AOL and I said, Listen, for a half million dollars, I can get you 40 of these people who are already doing it to program the homepage and program some topics and it worked.

So I think you could offer the top mods a job in mod support.

And this maybe they're doing this already at Reddit, but the ability of Hey, if you get your subreddit to a certain level of critical mass, hey, maybe we'll hire you for 5000 a month, or maybe we'll split ad revenue with you.

This could change the entire nature of Reddit, you have to be careful, because again, a lot of people who are doing this are not doing it for the money.

And then if they do it for the money, they're going to compare it to their day job.

So if you're a developer making a quarter million dollars, you know, at Uber, you might not want to make you know, an extra $50, 000 a year from your subreddit in advertising.

But if you're a journalist, and your subreddit makes $50, 000 a year, you might feel great.

And you might quit your journalism job to try to double that 50, 000 get to 100, 000.

So I would think of a sub stack, Patreon like model here where you could enable mods to make some money, ie I subscribe to this specific subreddit, I give $1 I give $10 a month, whatever, And that gets split 5050 7030.

And that would be very appealing, I think, to the folks on Wall Street.

I know if I was going to buy the stock right now as an ad based system with a little bit of a possible data business.

But if it had a revenue share like YouTube, that would be groundbreaking.

And that would be what I would pursue if I was Steve.

Interesting take from our guy already Buko.

You know, already Buko from the Sopranos went on to a great career in venture capital doing Buko capital.

He says Reddit is walking a tightrope now need to choke out third party apps to serve ads on core app have now pissed off all their free labor mods who are now revolting.

Ben Thompson put a while the adventure that is the end of free money continues.

And that's really referencing Zerp 0 interest rate policy.

If Reddit is a worth $10 billion or $20 billion, at some point, the free money invested into companies like Reddit, WordPress, whoever, they're going to need to basically demonstrate revenue and go public.

And the Zerp part of this, where there's free money everywhere is truly over.

You see that with Airbnb, and Uber, most of all, those businesses were in dogged competition with infinitely funded competitors that were taking all the margin out of these businesses, same thing with DoorDash.

And luckily, that Zerp phenomenon is now over.

And we're actually getting the nature of these businesses and the nature of Airbnb is it's printing money.

The nature of Uber is it just tipped over into free cash flow and is going to, you know, start printing billions of year billions a year in free cash flow.

I believe that and I'm still holding a large percentage of my shares from the seed investment.

Reddit is expecting a $15 billion valuation on their IPO.

The last known revenue number was 400 million in 2021.

But 2022 was a horrible year for advertising.

I wouldn't be surprised if they didn't grow much, maybe they had a flat year 500 million 600 million.

But I don't think they're tripling revenue or doubling revenue, it might be like 30%.

And that 15 billion number based on the 2021 10 times 400 million is 4 billion 20 times 30 times 40 times is 16 billion.

Probably not, especially if it's growing less than 30% 40% year over year in which we don't have that number.

So maybe it's going to be worth 20 times revenue, 15 times revenue, we're having a little bit of a tech rally, but I think they need something exciting and fun.

I really hope they consider my proposal of sharing revenue with the mods, sharing revenue with the app developers.

All right, let's get to this awesome fireside chat with my bestie Brad Gerstner from Altimeter.

We talked about macro macro macro trends.

And it's an awesome fireside chat that happened at our angel summit, the angel summit is 100 capital allocators, VCs, LPS, angel investors, fund to fund sovereign wealth funds, endowments, etc.

Coming together for 3 days in Napa, we have it every June.

And hey, Steve, when you're out of your quiet period, or whenever the time is right, please do come back on this weekend startups.

Okay, here's Brad, creating a job post and finding qualified candidates can be time consuming.

Like just the amount of time it takes to find somebody unless of course you use LinkedIn jobs, because LinkedIn is closing in on 1 billion users, as I predicted, well over 900 million members now.

And think about how many insanely qualified people are out there right now looking for work, just go post an open role right now on LinkedIn, and you're going to be 100% certain that you have access to the most qualified candidates available in the world.

Cal, that's right, go to linkedin.com slash twist to post your first job for free.

You got nothing to lose everything to gain.

And 1 of the things you're going to gain is that purple ring.

You ever see the purple ring around people's profile, that means that you're hiring.

And when you see that everybody in your network is going to know you're hiring and you get some friends of friends, you know, those are always the best candidates anyway.

1 of your friends says, Hey, I know a great salesperson, they were laid off, or they were part of a riff, no fault of their own.

And maybe you land that person who is a game changer, a bar raiser for your startup.

That's exactly what I did for launch for inside.

And I found so many amazing people on LinkedIn jobs, it's a faster way to find better candidates.

I'm doing some programming right now in your brain.

I say LinkedIn, you say better candidates, faster on LinkedIn jobs, better candidates faster, LinkedIn jobs will help you find qualified candidates that you want to talk to faster post your job for free linkedin.com slash twist as linkedin.com slash twist to post your first job for free.

Next up is Brad Gerstner from altimeter capital Brad, you were in the room for Mars presentation.

I'm curious when you hear her framework for picking hits at the early stage.

Juxtapose that with where you spend a lot of your time, public markets, and frameworks for making decisions about picking future hits there and working with those founders and supporting those founders in the context of you made a big bet on Nvidia recently before the big run up that seems like a pretty good example.

Snowflake seems like a pretty good example from your portfolio from the private side.

And then of course, maybe supporting founders sometimes with tough love in the example of Facebook.

But let's start with just what you thought of her framework.

Like I think, you know, Mara's presentation, You know, I'm sitting over there taking pictures.

I'm like, I was live blogging, Mario, your blog or your presentation to the besties.

And she had that slide and it was Mike Moritz and Bill Gurley.

And of course, they love that they're having a field day with that.

And Gurley's like he's because he didn't put his hand on his chin.

If you notice Moritz and Gurley both had their hand on their chin, Rodin style.

you know, so it's like not going to happen, but

I've been leaning into, you know, the Steve Jobs 5 fingers pose.

but that was a that was a great presentation.

And picking is picking whether you're in the public markets or you're investing in Snowflake when it was pre-revenue.

And 1 of the things she talked about, you know, Clive talked about is that luck befalls the prepared mind.

And what links those 3 people together, what links the besties together, is you have people who spend 20 hours a day thinking and preparing the mind and debating and messaging 1 another with analysis, with rebuttal, sometimes quite sporty, And we're deeply passionate about how technology is moving humanity forward.

And unless you like authentically feel that, unless you get up every day studying that, trying to learn with a curious and a fresh mind, I think then it's very difficult to compete.

Because the byproduct of doing that is that occasionally, you have an observation about the world that is orthogonal to the observation that other people have.

And yes, there are patterns that get developed along the way.

But everybody, when I went to Harvard Business School, they taught about the efficient market hypothesis, that everything is known or quickly will be known, and so things are priced efficiently.

NVIDIA was priced at $125 a share in January of this year.

Of this year, the consensus sales forecast was that data centers would be down 6%.

This is all of the world's best sell-side analysts.

Goldman Sachs, Morgan Stanley, et cetera, thought that data center growth in the year of AI would be negative for NVIDIA.

And so, I think people take comfort in crowds.

They build models that all look the same with linear deceleration.

And, you know, like, and they don't want to be too far out of the crowd because it's an uncomfortable place to be.

But when I have entrepreneur after entrepreneur walking into my office saying, the single greatest bottleneck that we have in the world of AI today is our access to H100s.

It was not hard for us to get our head around that.

And by the way, Gavin Baker is going to be talking later today.

He's forgotten more about silicon and about chips than I'll ever know.

But sometimes that's a burden, not an advantage.

And I think 1 of the advantages for us is, I try to take a very fresh look at these super cycles, just listen to what I'm being told, look at it relative to what the consensus thinking is.

And, you know, Snowflake, when it was pre-revenue, the consensus thinking was that cloud was too costly was not as performant it was less secure that was 2013 that was the state of the world right but if you went and you talk to every entrepreneur did they like having to set up a data center?

So, and you looked at the trends that were happening, it was pretty clear that those lines were going to cross.

And if they did cross, the biggest winners in this thing called cloud were going to be databases that were re architected cloud native.

And so we found the founding team out of Oracle, you know, who had been working on this for 2 and a half years with some great early investors that were in it, and we took a chance.

I think another incredible example is when we watched revenue and the headwinds that the advertising market was facing, Google, Facebook, Snap, and we were looking at how fat these companies had become and just how much money they were spending.

In the Dow market, we started having this conversation at poker and just in group chat about well, if the earnings are going to go down, and they're going to get compressed, well, if you stop spending money, or you cut costs, well, then the earnings can go up.

And there was this incredible resistance that nobody would ever do that.

It was just nobody had the sheer will of force to say, you know what, maybe we don't need as many people, maybe we don't need as many departments, maybe we don't need to give the free lunch or whatever it is.

And you and I were talking about Facebook.

And I was like, God, it just seems I heard your whole sort of rationale of how they could turn things around by just changing maybe how they thought about spending.

And that Sunday night, we were having the conversation.

I saw you say, Look, they're cutting 10, 000 people tomorrow.

And I said, Whoa, if Zuckerberg doesn't like to lose, he's only 1 in his whole life.

And that 1 turned out to be except exceptionally correct.

Again, a non consensus thought Silicon Valley could do more with less when we just lived in a culture of just keep spending.

So maybe you could talk about that call that you made, which is really great call, obviously, it's up, I think, 2 or 3 x since then.

But more important than that call the sort of operating philosophy and how that's impacted these companies.

And maybe we'll use as a way to segue into some slides after I answer it briefly.

Listen, we've been talking about this for a year, right?

Over poker on the all in pod, you know, etc.

See all This Week in Startups transcripts on Youtube

Reddit communities revolt + Brad Gerstner on state of private and public markets in 2023 | E1761