Keith Rabois: OpenAI is winning the AI battle, Opendoor needs a new CEO and AI-native transformation

to this and debate Google more on this show, but we have Keith Ra Boy joining the stream. Welcome to the show, Keith. Good to hear from you. How you doing? Great. Pleasure to be back with you. Fantastic. Uh, before we go into all all the news that's shaking up the timeline. Do you have a take on on Google right now?

We were diving into the fact that semi analysis seems quite bearish on Google and is is c is is saying that uh uh openai will basically steamroll them once they get into agentic commerce and and checking people out properly monetizing the user base and all the economic activity that they're already driving. Yeah.

And then Ben Thompson's kind of saying like, hey, it's this college campus. They have a bunch of great AI researchers. Like something will come out of this and there and it's it's just a lindy company. It's been around for a long time. It's not going anywhere.

Well, I think Chat GBD is the fastest growing consumer app of all time and as long as that continues, Google's significantly threatened. I think normal people are substituting uh what used to be searches and queries into prompts and that really does threaten Google.

Even if they have top tier AI talent, they have yet to productize it in a way that undermines the trend towards TBT. If Google were to fuse together all the information they have about you, your Gmail, your YouTube, etc. , and make everything personalized out of the box, that'd be interesting.

We'll see if they can ship that product, um, internally, uh, substantively, and whether it resonates with users. But they're losing, you know, the AI battle to Open AI, period. It also threatens the revenue.

I don't think that performance-based advertising which is the mainstream you know driver of Google success will be the way that consumers uh expect to be monetized in the future.

I think uh all the AI applications are showing that consumers are willing to pay for value and so I think direct value direct capture uh from consumers subscriptions etc might be better um and that would you know undermine the entire advertising business model that really has propelled Google into the stratosphere was there's product innovation initially engineering innovation data innovation but ultimately is advertising innovation and they had a more efficient advertising platform and that may not be the future of the next 20 years.

Yeah. Yeah. In some ways, Google's Google's margin and how much revenue they generate for me as somebody who throughout the year is going to make a lot of searches that that result in a lot of purchase activity.

And if I can substitute that by just paying 200 bucks a month and eliminating a lot of that, you know, it it Chache BT can can become a probably a trillion dollar company by just eating into eating into that and being willing to give up that that incremental revenue that they might get from advertising.

I just had this experience. I bought a Nintendo Switch 2. I went to chat GPT. I fired off agent mode. I had it uh go around check every website for what's in stock. It decided that Target was in stock. I clicked on that. Had to go into the Target app and set up an account and it was like a huge hassle.

Could have just done it for me and taken a cut, but Google wasn't involved. So, it's kind of a crazy time. Yeah. I mean, you can also do some rough math. Uh you can calculate Google's revenue per user per month or per year. Yeah. And then what consumers are willing to pay through chatbt and similar products. Totally.

I think exceeds that. Yep. No, it makes sense. Anyway, uh enough on Google. give us the high level on what's going on with Open Door. Well, I'm very excited. A bunch of retail investors led by Eric Jackson have really take a spotlight and focus it on the potential of the company.

I think for a lot of reasons the company for the last three years has not really uh capitalized on its disruptive elements, its innovation and people have sort of forgotten about it.

um in the public markets attention is really important and now there's a a very strong shining spotlight on the potential of the company which I think should be like Carvana.

Carvana is a $40 billion company with in fact more competition than Openoro has but Open has not been able to frame the narrative about the innovation as well as Carvata and I think that's going to have to change.

I think substantly the company needs better leadership both on the storytelling side and on the innovation side. I like tweeted out, you know, really with about 10 minutes of thought ways to fix Open Door this morning. Yeah.

And there's specific uh specity behind every one of those elements that I could dive in deeply, but the company's not executing on any of those things. If anything, it's taking a step backwards and partnering with these legacy real estate brokers, which makes no sense whatsoever.

How much of the like selloff in the stock or just kind of like you know the like getting into the doldrums is and the trouble that open door has run into has been just the interest rate dynamic in the post 2022 2023 era where we've been in this high high interest rate environment that that abstractly has a big impact on home buying.

But how direct is that on the open door story? It's moderately direct. So in a hot market, maybe 5 to six million homes transact a year. And Open Door gets paid every time it transacts. So in a in a very high interest rate environment, that number went down to 4 million. So 4 million verse six. Like that's that's right.

It's not like people stop buying and selling. It doesn't go to zero and it doesn't go to infinity. So 4 to 6 million. The problem for Open Door financially is the cost structure of the company at a GNA level, not marginal level. GNA level is just way too expensive to make profits at $4 million.

And this was a known problem. The node post who I work with warned the current management team, myself included, um, in 2015, that this was going to happen. Real estate has roller coaster rides to it. And you need your fixed cost base to be low enough that you don't have to transact to offset your fixed cost, your burn.

Unfortunately, the base was too high. And when the Fed raised interest rates like five times sequentially really fast in kind of an unprecedented way very few people transacted. So there wasn't enough profits to offset the fixed cost of running the business. It was a mistake. It really did cause significant strain.

Arguably you know the entire management team made that mistake didn't listen to the node. Whenever that's three years ago the company's made no progress on cutting the GNA cost over the last three years. The good news is it's actually really easy to cut the GNA cost these days.

AI allows you to substitute for most of what the people that work at Open Door do.

So you should be able to bring the very the fixed cost of running the business down to here and then just decide on every home purchase are we going to make money or lose money and just purchase homes in any interest rate environment where you will make money.

What are the what are the key activities within the business that you think are are most ripe for replacing with agentic workflows? Honestly, I think real world stuff like there are companies that are actually doing inspections with AI.

The idea that you need this like we have labor going to inspect you know repairs etc etc that that stuff can be done videos and AI much better than uh traditional people and there's companies that are specializing in this. So I I just think that you don't need all these people and it's going to continue.

acceleration AI is more and more obvious and more and more stark every day. So can you cut it to 100 people?

Probably yes actually maybe less and then you have a then you all have variable costs and the only thing you need to do is model the value of a home correctly which is complicated but open door has been excellent at it. One thing that people miss is with a exception of one quarter in the history of the company.

The company has priced homes correctly and purchased successfully and profitably, but the marginal cost uh the fixed cost sorry has not been able to be offset in when no one's transacting. The company has significant market share in many markets.

So the company will just mint money as long as it's gets its cost under control. Now it needs to be a massive company which is the potential. Like if you think about it, let's take a top down perspective. Yuri Mililder actually made this point to me seven years ago.

The largest real estate platform in the West is worth about $18 billion. That's insane. This is the largest asset class. Period. The idea that the most innovative company in the entire Western universe would be worth $18 billion in residential real estate makes no sense whatsoever.

So somebody's going to build a 50, 100, $150 billion market cap company. And this management team doesn't think that way. They think about, you know, moving one basis point here and one basis point there versus innovating to build a hundred billion dollar company.

There is no reason this shouldn't be a 50 to$100 billion company. You just need the right leadership and we're going to fix that. We have a couple questions from the chat. Are you interested in going back on the board of Open? Are you interested in stepping in the CEO seat, getting a new role?

like what what do you see your involvement going forward? How do you think about that? The co the company needs a new CEO. If I can be involved in identifying, assessing and or closing the proper candidate, I'll be happy to be involved. Um I I do not plan to be an executive. I have a very busy full-time job.

Um and I this company needs a full-time dedicated CEO who's intense, who's creative, who's innovative. Um, so if I can help that person, y if I can encourage that person, that would be wonderful. Regardless of the structure, maybe they're listening right now. Uh, DM Keith if you're the one for the job.

If you're a great if you if you're anything like Will Gabbrook at Stripe or Greg Bman, head over, please call. Come on over. Uh, I want to talk about uh transformation and actually implementing agentic workflows, implementing AI and then cutting cost. Is there a sort of uh like capex type cost?

I mean, I imagine that there's some folks on the private equity side that are buying up small businesses and then hope to use AI to cut costs, improve margins. Uh, and maybe they're doing that internally and defaying those costs of knowing which tools to pick uh internally across a portfolio.

Then there's McKenzie and the big uh the big consulting firms that are going to Fortune 500s and saying, "You're going to pay us a ton for a slide deck, but you know, in theory, we're going to get you set up with something that will save you money. " But for a company like Open Door, um, what does it look like?

Because it feels like it's probably too complex to just say, "Okay, yeah, we're going to like sign up for chat GPT and that solves our problem. " There's probably going to be some implementation of these AI tools. Uh, and that might have cost in the short term, even if it drives longer terms better earnings outlook.

How do you think about that? Well, I think that's why we need a CEO who's AI native, AI uh, insightful. I think that business transformation requires real skill.

I think you are right though that the hottest area of venture capital is venture capital is chasing after old school businesses and trying to turbocharge businesses uh with AI whether they're turbocharging on the top line which to me is more interesting or improving IBIDA with AI substitution of cost.

That's that's pretty cool. But you know we funded a few companies. There's at least two VC competitors of ours that I know of that have dedicated funds that do nothing else except roll up traditional businesses and try to turbocharge them with AI. So I think you know this is very common at the large company level.

It is happening apparently. Um actually apparently Amazon has had a lot of success with this. They don't talk about it and speak about it but apparently it's true.

Um, and then I think you're going to see more private equity firms insist upon their portfolio, whether they're large market cap portfolio companies or small, um, apply AI in a thoughtful creative way.

Do you think do you think I mean it feels like private companies have like an extreme structural advantage in terms of doing like true AI driven transformation because open AI uh sorry not not open AI open door you know every uh over the last week every 10 posts on X has been some you know different retail investor having strong opinions about who the management should be and and what the board's doing and all this stuff.

Um, yeah. Do do you think that um do you think that if if the right person were to come to the table that that like a that that Open Door would do better if it was a private company for the next like few years? Yeah, I don't think so. I think you can transform yourself in the public domain as well.

I think you could take advantage of those suggestions. I think first of all, let's just take a step back. I think retail investors having a point of view and being excited about an opportunity is a great thing.

I think the whole point of markets is to allocate capital that's why we have markets right that's why we have public markets it's an allocation function and consumers voting with their feet especially for consumer brands saying I want more of this I want less of that is actually a proper capital allocation like if the company did this I would spend more money with them that should encourage capital allocation the this is not some people have like this negative perception of retail investors I think it's actually better when retail investors say I'm going to vote with my feet, I'm going to vote with my dollars.

If product X or Y or brand does it does X, Y or Z or brand represents Z, I'm going to spend money with them. That is a reason to allocate more capital to that company. It's fundamentally sound.

How do you think about the the storytelling around AI as a silver bullet versus a core competency that will be a compounding advantage? I'm thinking of the Amazon example you gave. I completely agree that Amazon's been a beneficiary of AI all over the place, but I don't think it's happened in a single quarter.

I think they ramped to a million robots across all of their different facilities. They bought KA a decade ago. They were using they were using AI recommendation systems to tell you, hey, you're buying a computer. Do you want to monitor? That's AI. But, you know, just a couple decades ago.

And so, it feels like the right person for the job. They might be able to come in, rip off a band-aid, get things right sized, do the hard work, but then it's really like you can't take your foot off the gas. I think you need to do both.

I think there's bottomup transformation, which is blocking and tackling, persistency, consistency. It's like going to berries. You have to go every day like every day for like for like a decade. And you know, you get sometimes sometimes sometimes three, four, five times that. Yeah.

And then there's I think there's top down. I think leadership involves sometimes just putting a stake in the ground and saying thou shalt not like we are just not going to do this anymore. We are absolutely out of that business. And that's why you need a founderdriven CEO.

Truthfully, transformations with emerging technology require the moral authority of a founder just saying absolutely no. I know this has worked in the past. I know this has worked in other companies, but we're just not doing that anymore because the world's going this way and we want to be ahead of the world.

How do you think Open Door's relationship with with traditional real estate agents should evolve? Well, I I think it should be we should be innovating so that there's no comparison.

Like the value proposition Open Door provides a consumer, whether a buyer or a seller, should just be so much better that no one wants a real estate agent. It's not a bad thing.

Like real estate agents used to do X and Y and there's a bundle of services they provide, but what Openor provides is this and it's just like a no-brainer. And if it's not a no-brainer, the company is not innovating and is not creating enough value period.

Uh how do you think about other levers like zooming out to the macro that could just increase the velocity of uh of real estate transactions or just make homes more affordable in America?

I don't know how high this up is how high this is on the current administration's agenda, but it feels like something that people have been clamoring for for years on both sides. Are there obvious wins that you're optimistic about in the next couple of years to just improve the quality of housing in America broadly?

Well, affordability is a top tier issue certainly for my conservative friends. Yeah, we need to make housing more affordable for more people as fast as possible. There are some things that have short timelines and some things that take longer. So, building is great. We need more. We need more supply. Supply works.

Supply has worked in local environments. You can prove it in in in city action and citywide. The cost will come down if you build. But you can't build a house, at least right now, without more robots, more automation. You can't build one overnight.

So supply does take time, but we need to start working on supply and getting rid of all the blockers and excuses for lack of supply, particularly in California. Secondly, we do need interest rates to come down. U we do need a new chairman of the Federal Reserve.

The best thing ever, you know, for open door would be replacing Carrie as CEO and replacing Jerome Powell as Federal Reserve chair. Fortunately, I think both are going to happen. I hope both happen in September, maybe before. Okay. Can interest rates ever be too low? Probably.

You know, like interest rates, it's basically related to a time value of money. Yeah. And so if interest rates are too low, people's willingness to part with money to get paid back in the future gets reduced. And that that is investment. That's that's really what the definition of investment.

So they could be too low, but they're they're definitely too high right now. I guess I've just been thinking like there are there are a lot of green lights, green flags in the market. New companies are going out. The stock market's at all-time highs. Everything feels really strong in the economy.

Even the CPI is coming back flat and GDP is printing. Everything seems pretty good and it feels like except the fast casual restaurant chain. Yeah. Yeah.

They're having trouble except the slot market is but in general things to be seem to be going very well in the American economy and when there's a risk of okay, we could be overheating.

I feel very reassured by having let's lower interest rates as an ace up our sleeve in case we get over our skis and there is a market correction.

the Fed does have some has some tools in the like our friend our friend Joe's point is like the in his view the the argument to lower rates is like the data that's coming out of the labor markets but again that's also being debated and rehashed and and uh I don't think anyone really well let me let me take a step back though please I think the foundation that growth equals inflation is just wrong so from 1950 to 2010 Each decade we average 3.

6 years with over 4% growth without inflation. It's only the modern world post 2010 with qualitative easing that people equate growth with inflation.

The good news about AI and productivity gains is it's very easy to see how you can grow fast three, four, 5% consistently without sparking inflation because all of the growth is not propelled by labor cost increases which is what causes inflation.

So I think the modern world of the next 30 years if it's managed correctly, if the leadership in the political sphere is dialed in should allow consistent growth which will eliminate the debt and make not a non serious problem like 3% 3% 3% plus without inflation.

And we need to get people who are sort of educated a century ago out of this mindset that every time you see growth, you need to put on the brakes. That's just not that. And that's why the Federal Reserve keeps making that mistake. And so we've got to fix that.

But part of it is AIdriven and technologydriven innovation will allow for great growth, consistent growth without inflation. Is that your current outlook? Um, not necessarily a fast takeoff and we're growing at 10% GDP a year. Uh, Satin Nadella says, "Call me when we're growing at 10% a year.

" Uh, but but but a materially improved economic condition for the United States on a on a long-term basis. Yeah. Scott Bess likes to talk about 33 and you want the 3% consistently. You could beat three and I think we will beat three and I think he wants to beat three in the next couple years.

So I subscribe to his perspective on the world. I think he's right. But it's technology that's the magic wand that allows consistent growth without actual inflation. And that's what we need. That is how that is raising taxes is a disastrous policy. It's not going to fix any problems.

Consistent growth without inflation will. And we need a Federal Reserve chair who understands that. We have a Treasury Secretary, fortunately, who really does understand this. Yeah. Yeah. And we and we've seen that with those charts of the various goods and services inflation over the past decade.

Education and healthcare goes through the roof where everything that's on the technology adoption curve like TVs and dishwashers that all has gone down in price. And so the more goods and services that you can put on the deflationary curve, the more growth you get and the and the better health of the American consumer.

Good stuff. Jordy, anything else? I think that's it. Thank you for jumping on on short notice. Thanks for taking the time, Keith. This is always great. come back on whenever you have uh more thoughts on uh on if you if you can think for 10 minutes and get a post up, you can jump on the show, please.

We'd love to have you. Great pleasure to be with you. I'll be back. Cheers. We'll talk to you soon. Bye. Take care. See you. Uh let's go to the medicine list. We updated the Met list, our ranking of the uh top 128 now AI researchers. Uh it's