Grammarly is now Superhuman! CEO Shishir Mehrotra explains the rebrand | E2201

Today's show: *Why rebrand an app with 40 million daily active users? On TWiST, Grammarly CEO Shishir Mehrotra swings by to explain why he's renaming his company after the mail app he just purchased, Superhuman. Plus he's giving us a first look at the startup's new AI assistant, Superhuman Go, digging into their new agent-centered strategy, looking ahead at the potential for AI job displacement, and more. Jason and Alex also talk about our new nuclear deal with Westinghouse and Japan, OpenAI's conversion and what it means for Microsoft, 1x's tele-operated Neo-bot for chores, and much more! Timestamps: (00:07:56) Jason is headed to Tokyo, and NOT just for skiing this time! (10:00) Uber AI Solutions**:** Your trusted partner to get AI to work in the real world. Book a demo with them TODAY at Uber.com/twist (00:12:21) Untangling this US government-Japan-Westinghouse deal for nuclear power… and why nuclear power is SO important (20:00)DevStats translates complex engineering metrics into a shared language everyone at your company can understand. Get 20% off by going to https://www. DevStats.com/twist (00:20:04) Why Jason says sabbaticals and 4 day work weeks are for established brands, not startups (30:00) Perspective AI: Real insights, straight from your customers, and your first two months are on us. Just go to getperspective.ai/twist. (00:33:01) Will income inequality become a concern, even for the wealthy? Jason says YES. (00:37:29) Shishir Mehrotra from Gram— NO WAIT, Superhuman! joins us to talk name changes (00:38:16) Using Grammarly as a model for the new AI assistant, Superhuman Go. (01:08:39) Nvidia shot past $5 TRILLION in value… How is no one freaked out about this?! (01:14:52) OpenAI executed their conversion… but Jason says "public benefit corporation" status is pretty meaningless (01:25:02) Checking out the Neo robot from 1x… he wants to help you with chores and he's for sale! (01:32:09) Uber's speeding up its self-driving plans Subscribe to the TWiST500 newsletter: https://ticker.thisweekinstartups.com Check out the TWIST500: https://www.twist500.com Subscribe to This Week in Startups on Apple: https://rb.gy/v19fcp Follow Lon: X: https://x.com/lons Follow Alex: X: https://x.com/alex LinkedIn: ⁠https://www.linkedin.com/in/alexwilhelm Follow Jason: X: https://twitter.com/Jason LinkedIn: https://www.linkedin.com/in/jasoncalacanis Thank you to our partners: Perspective AI - Real insights, straight from your customers, and your first two months are on us. Just go to getperspective.ai/twist. DevStats translates complex engineering metrics into a shared language everyone at your company can understand. Get 20% off by going to https://www. DevStats.com/twist Uber AI Solutions: Your trusted partner to get AI to work in the real world. Book a demo with them TODAY at Uber.com/twist Great TWIST interviews: Will Guidara, Eoghan McCabe, Steve Huffman, Brian Chesky, Bob Moesta, Aaron Levie, Sophia Amoruso, Reid Hoffman, Frank Slootman, Billy McFarland Check out Jason's suite of newsletters: https://substack.com/@calacanis Follow TWiST: Twitter: https://twitter.com/TWiStartups YouTube: https://www.youtube.com/thisweekin Instagram: https://www.instagram.com/thisweekinstartups TikTok: https://www.tiktok.com/@thisweekinstartups Substack: https://twistartups.substack.com Subscribe to the Founder University Podcast: https://www.youtube.com/@founderuniversity1916

Speaker names marked ? are inferred from speaking time and the show’s known hosts — they are not voice-verified.

• 0:00Speaker 00

  I want to bring up something else, Jason, that's a little bit off from where we just were on the energy side of things, and instead talk about humans. There's been a theme on the show for ever since I got here that companies are either not hiring or reducing their team sizes. And so when we think about taking care of employees, we don't talk about attrition that much. It's just not the thing we're kind of tuning for, I think, as an industry. But our friends over at 37Signals, DHH, and Jason Fried, are talking today quite a lot about their use of sabbaticals. This is a six-week break for workers in their third year at the company as part of their overall kind of like kind setup for their staff, including four-day work weeks from May through September. And this struck me as just out of the blue because we're talking about laying off 14,000

• 0:44Speaker 00

  people at Amazon, not trying to keep people. So, one, are you in favor of sabbaticals of this type? And two, are they just drinking different Kool-Aid than everyone else?

• 0:54Jason Calacanis?

  So, they are a highly profitable software company that has existed for a long time with a loyal base of users, and it is pretty reliable.

• 1:07Jason Calacanis?

  That those revenue streams, people who use Basecamp, they just love it. They're never going to stop using it. Okay, period, full stop. So, when you start a company and you don't have a bunch of investors, which they don't, I think they bought out Bezos, who was their only investor back in the day, they can build the world they want to live in. Now, you remember, they had a whole DEI woke thing happen at their company where people wanted to bring all their opinions to work and debate it on Slack, just like at Coinbase, and we had this sort of pendulum swing the other way. Hey, if you're at Coinbase, you're here to do one thing, you know, provide crypto solutions to the world. If you're at 37Signals, they took a similar approach.

• 1:49Jason Calacanis?

  Hey, if you want to be here, you're here to make great productivity software, full stop, and that's it. And then here's an incredible pay package. I think both companies just said, hey, if you want to opt out, we get it. So during that culture war, which we forget about, but I guess that was 2017 to, you know, COVID, basically, they went through that. Then whoever's left, what's happening today is companies are getting more efficient. If the company gets more efficient and you own a business... You really appreciate stability. You really appreciate the fact that Alex shows up three days a week and has a suit and tie on, or I didn't tell you about the ties. He has a fresh shirt on. Don't make me buy ties. You never know.

• 2:34Jason Calacanis?

  So you start to appreciate that, right? And then when you lose a couple of video editors, you know, because they want to go make the next great movie or they want to be a screenplay writer. You're like, oh, my God, I own this business and I'm going through this again. I got to hire that person. Oh, you know, start the process again. It becomes a distraction from growing your business. So these kind of offerings seem crazy if you were running a large company, and they would be. They would create entitlement, they could create abuse on the margins. However, with a small company, you can make a lot of decisions like this that are quixotic. But you can be 20% inefficient over the summer or for four months.

• 3:17Jason Calacanis?

  You can be, if six weeks is like giving, over three years, is like giving people an extra two weeks of vacation, essentially. If you've built a company that's resilient and you have plenty of people and you're overstaffed by 20%, it doesn't matter if people go on sabbatical for six weeks. So you can build companies like that. Now, if you tried this strategy and you were up against OpenAI and XAI, you would fail miserably. So you do need to understand the game on the field. If you had venture capitalists and you tried this and the company wasn't growing, they'd be like, hey, maybe people should come in on Fridays since we're not growing. But we will see in the United States, I predict, by the end of our lifetime, the emergent of, the emergence of a four-day work week. I think, you know, a four-day, you know,

• 4:02Jason Calacanis?

  not this like five, six hours that people are working and going and getting matches and yoga, whatever, like, you know, the full-on 10-hour day, four 10-hour days might actually, for some people's rhythm, be better than, you know, five, eight-hour days or five of these like six-hour days where you do a life in the thing. So I think it's worth noting here because you have to know the game on the field and the context. The context is highly profitable company, extremely stable revenue, no investors, two owners. They can do what they want and they can pursue whatever they want. I remember when I worked in New York, I would go to Chanel and we managed their IT systems, like the luxury brand, and they all were allowed to leave work at 12 o'clock or 12:30 on Fridays during the summer.

• 4:47Speaker 00

  Summer Fridays.

• 4:49Jason Calacanis?

  Summer Fridays. Conde Nast, the same thing, but they had to get in at 8:00 a.m. So they had to like put in at least a half day at work and, you know, they were very pampered. Everybody had an assistant. Nobody was really working. You know, it was, they were selling $20,000 handbags. So let's just leave it at that. That's the equivalent of 37signals. They're Chanel in this equation. They're not grinding it out versus 20 competitors who are going to eat their lunch.

• 5:13Speaker 00

  Yeah. Question about this because that's kind of what I thought you were going to say about this, which is this applies to a certain class of companies, not most startups and probably not places that a lot of people are going to end up working at because 37signals is only so big. But don't we want to create an economy in which people can work a little bit less? I feel like we talk about this usually as a crisis, like if we work less than 40 hours, we're all going to hell in a handbasket. And I'm just kind of curious like if that's the world we're trying to build, Jason, from your view. This week in startups is brought to you by Perspective AI. Surveys, they never capture what customers are really thinking. That's why we use Perspective AI.

• 5:56Speaker 00

  Real insights straight from your customers and the first two months are on us. Just go to getperspective.aion/twist. Devstats. Devstats translates complex engineering metrics into a shared language everyone at your company can understand. Get 20% off by going to devstats.com/twist. And Uber AI Solutions. Bad data, bad AI. Your AI is only as good as the data it learns from. Uber, that's right, Uber AI Solutions now works with enterprises around the world to source, label, evaluate and scale real world, high quality data for every industry, everywhere, so that you can focus on building the next big thing.

• 6:40Speaker 00

  High quality data equals smarter and faster AI. Go to uber.com/twist.

• 6:49Jason Calacanis?

  All right, everybody. Welcome back to This Week in Startups. It's Wednesday. It's October 29th.

• 6:54Speaker 00

  Phew.

• 6:55Jason Calacanis?

  Yeah, time continues to march on. Lots in the news. I noticed OpenAI did their restructuring. I noticed we're going to have nuclear power plants to power all these AI

• 7:08Jason Calacanis?

  data centers and that

• 7:12Jason Calacanis?

  they're not hiring young people and they're laying off everybody. So

• 7:15Speaker 00

  Yes, correct.

• 7:16Jason Calacanis?

  Just a little bit in the news this week.

• 7:19Jason Calacanis?

  But I guess at the start of the show, I hate to talk about myself, but we have a little announcement.

• 7:24Speaker 00

  We do, and it involves, which Jason, one of your favorite countries in the entire world, although for a slightly different reason this time.

• 7:31Jason Calacanis?

  Yeah, so pretty excited to announce that we will be bringing our Founder University to Japan. As many people know, I have a love of Japan, the people, the nation, the culture, and the skiing. And I go every year to do a little pilgrimage and get some of that J-Pow, the Japanese powder. We will be kicking off on November 14th. I will go to Tokyo to announce our 10-week program. And this will culminate with an immersion week in San Francisco and a demo day. So all the great startups and founders that launch from the Founder University will come to San Francisco and spend a week with us. And

• 8:16Jason Calacanis?

  it'll be great because I've always had great admiration for the startup ecosystem in Japan. They are incredibly innovative. They make very interesting products, both hardware, software, and have for many decades. And so we're going to build that bridge between Silicon Valley and Japan and look for great ways to

• 8:40Jason Calacanis?

  work together. Now, I need an analyst, a researcher and analyst to come work here at Launch. So if you know somebody who speaks Japanese fluently, they could be Japanese-American, they could be a Japanese expat living here in the States, who wants to split their time between two amazing cities and hang out with me. Doing it, it's Austin and Tokyo is where you will be living. And yeah, you got to be fluent in English and in Japanese. You can just apply by emailing japan@ Launch.co. We don't have the .com yet, but launch, we'll keep trying for 10 years. [email protected]. And just tell me a little bit about yourself. My team will review the

• 9:24Jason Calacanis?

  applications. You could be right out of school. You could just be highly technically savvy. We'll train you up on what we do in terms of mentoring and finding resources for and supporting founders. But we need to have boots on the ground for these programs. And as you know, I am going to be in Riyadh next week for the launch of Founder University in Riyadh. So the sun officially never sleeps on the J Cal empire now. It is literally now, I think we're going to be in three very disparate time zones.

• 9:57Speaker 00

  Very, very disparate time zones. I will say, you know, I always thought we needed to replace the British Empire, Jason. I just didn't think that you were going to try to do it single-handedly.

• 10:05Jason Calacanis?

  With AI, like everything else, the old saying is so true. Garbage in, garbage out. Without the right training data, you're not going to get great results. But my pals at Uber, they're now working with enterprises all over the world to source, label, evaluate, and scale high-quality, real data for your startup. Uber is one of my favorite companies in the planning because they demonstrated their ability to scale and to build great products. First in the rideshare market, then they figured out Uber Eats, and then they moved on to autonomous driving. And all of this was only possible because of a deep understanding of how to collect, label, understand, and analyze data. Now with Uber AI solutions, you can put the team and tools that turned Uber into one of the world's

• 10:47Jason Calacanis?

  best companies to work for you, partnering with top talent and experts from Uber's global network. What better partner to help you scale your business than the company that organizes over 36 million rides every single day? Book a demo right now by going to uber.com/twist. That's U-B-E-R dot com slash twist.

• 11:11Jason Calacanis?

  And it's very exciting because we're also going to be bringing the Angel University there, which is my course, based on my book, where I just help people learn about angel investing. And we're doing all of this with the Japanese External Trade Organization known as JETRO. And so they are dedicated to promoting trade and investment between Japan and other nations. You may have seen our amazing, extraordinary, fantastic President Trump has done, was just in Japan. And so I am turning into Trump Jr. And I will be in Japan as well, doing business

• 11:51Jason Calacanis?

  deals. But apparently, and I'm not sure where on the docket, this story is, but we're definitely going to be talking about it. We're going to be building more nuclear power plants thanks to this relationship with the Japanese and their trade organization.

• 12:07Speaker 00

  Yeah. So the US is going to underwrite about $80 billion worth of new nuclear reactors with Westinghouse. And Japan comes into this because, if you recall, Jason, the current trade agreement between our two nations involves some Japanese investment into the US. So some of that capital is going to flow into these nuclear reactors. Now, it's not all just gifts from the government. In exchange for capital from the US, the US gets a 20% share in any dividends paid out by Westinghouse that exceed 17.5 billion. And there is an equity stake possible if essentially Westinghouse is worth at least 30 billion, then it has to go public, and then the USG gets a stake as well. We don't know exactly how much money

• 12:47Speaker 00

  Japan is going to be putting into this transaction. But it is, I think, good to see aggressive investment in nuclear power here in the States. I think we all are very aware of how much more power we need. Though I will say, just as a caveat, this does not solve our grid and transmission problems, but it does at least start the ball rolling towards more power here in the States.

• 13:08Jason Calacanis?

  Yeah, this is a little bit of a complicated deal. I saw Bloomberg was like, this is convoluted, and I just read their story on it. We are going to underwrite this, we being the US government, but we're going to be underwriting it with some amount of the capital coming from Japan. And in return for their participation, I'm unclear as to what they get, but the US is going to get a 20% share in any dividends that come out of Westinghouse above 17.5 billion. Yeah. And then, if this does come to pass in the next three years, and Westinghouse's valuation, which is a private company currently, I understand, if that becomes worth 30 billion,

• 13:53Jason Calacanis?

  the company is forced to IPO, and the government, our government, US, gets an equity stake. And you remember, I guess it was Besen who talked about at the AI summit we had over the summer with All In, that he had just closed up this 550 billion dollar investment package. And when I was talking to him, I talked about this on All In, I talked about it here, he couldn't shut up about nuclear. And so, this is another one of these win-win-wins. You can disagree with many things in the 47th administration, or the 46th, or pick any administration. One thing that they're getting right, I believe, is investing in infrastructure. Because if we're going to win the AI race, we're going to need power.

• 14:36Jason Calacanis?

  And if we don't get this power, yeah, we're going to need lots of it. And if we don't get this power dialed in quickly, from all different sources, we're going to have a real problem with PR. Yeah. Because, energy prices keep going up. Energy prices are not insignificant if you're in a household that makes under, I don't know, 100,000 or 200,000 in household income. You know, that electrical bill for 500 a month, or 800 a month, or 300 a month, I mean, it adds up to thousands of dollars a year. And if it goes up 10%, 20%, people are going to feel that. And they're starting to, and it's creating a negative perception

• 15:20Jason Calacanis?

  of AI. Because the public is now starting to put this together, Alex. You're building AI solutions, great. They require a lot of power, great. My cousin who's an Uber driver and my brother who's doing DoorDash while they're in college and my other second cousin who's in college for computer science can't get a job and the Uber drivers are going away because they'll be in Waymos and the DoorDash has their own delivery robot. And the reward we get for everybody losing their jobs.

• 15:53Speaker 00

  Higher prices.

• 15:55Jason Calacanis?

  Is higher prices. This is gonna be, I mean, if you want to talk about a revolution and the beginnings of one, when you see Mamdani winning, when you see the cost of living going up, it creates resentment and then that resentment will happen in the, hopefully, you know, when people go vote, but hopefully they don't vote for socialism, we don't kill capitalism. But I would be lying if I didn't say I was concerned with the game on the field.

• 16:24Speaker 00

  Yeah.

• 16:25Jason Calacanis?

  But I am excited about this because nuclear is just such a clean, great solution. If you told me you're gonna put a nuclear power plant within 50 miles of my home, I would be like, okay. And when I said that on Twitter, somebody said, hey dipshit, here's a nuclear power plant within 50 miles of Austin, you're already living within one. I said, okay, so I guess I'm, I guess that's consistent.

• 16:46Speaker 00

  That's how comfortable I am, I didn't even know.

• 16:48Jason Calacanis?

  Well, actually, that's, that's the point. It's safer than anything else out there, yeah.

• 16:54Speaker 00

  It's incredibly safe. I'll live next to a wind farm, a solar farm, a nuclear power plant, a future fusion power plant, bring it on, anything but coal, pretty much, and I'm, I'm good to go. Just for context though, folks, we have seen tech companies themselves line up deals with other nuclear power companies. A number of deals in the AI hyperscaler and energy space include Google Plus Next Era, Meta Plus Constellation, Amazon and X Energy, and Google and Kairos. So we're doing this at the national level, we're doing this at the corporate level, and I'll just say, Jason, 10 points to the Trump administration for doing this, minus 10 points for killing off wind and solar projects, why can't we have all of them? That's my.

• 17:29Jason Calacanis?

  The good news is, the, the, not wind, wind is a bit expensive right now, but solar is so cheap that even taking away the subsidies, which there's an argument for taking away subsidies from a lot of these things.

• 17:42Speaker 00

  Sure.

• 17:43Jason Calacanis?

  You know, if the free market is working and you don't need to accelerate it, in the case of AI. I mean, this thing's regulatory bound, local, there's a lot of, these things are not like putting up solar farms, let's just leave it at that.

• 17:57Speaker 00

  Yes, for sure.

• 17:59Jason Calacanis?

  For sure. And I wanted to just point out, I asked producer Claude, just tell us, remind us, the 10 largest economies in the world by GDP, and their populations and their GDP per capita. And it's really interesting. People forget just how amazing Japan is in this ranking. Japan is still the third largest economy in the world at 4.2 trillion, behind the 27 trillion for the United States and 18 trillion for China. And then you have Germany, India, United Kingdom in fourth, fifth, and sixth, and then France, Italy, Brazil, Canada.

• 18:38Jason Calacanis?

  And so it is a, not a giant company, right? Not a giant country at 125 million, but that's a pretty impressive GDP. And thank you to our friends over at Claude.

• 18:52Speaker 00

  Yes, and you can't.

• 18:54Jason Calacanis?

  Go ahead.

• 18:55Speaker 00

  You can't do everything we're doing without a paid plan over at Claude from Anthropic. So if you want to get 50% off your first three months of Claude Pro, just go to claude.ai/twist to get started. We live and die by Claude. Thank you, Anthropic. On that point, though, Jason, I was really curious about how Nvidia's market cap, which we'll talk about in a second, compared to the aggregate stock markets of different nations, how many more countries is Nvidia worth more than? And just totally, you know, just found out this morning that it turns out Japan has the third largest stock market by capitalization at 6.25 trillion dollars, only falling behind the US and China. So it really is an enormous economy and an enormously wealthy nation.

• 19:37Speaker 00

  I want to bring up something else, Jason, that's a little bit off from where we just were on the energy side of things, and instead talk about humans. There's been a theme on the show for, ever since I got here, that companies are either not hiring or reducing their team sizes. And so when we think about taking care of employees, we don't talk about attrition that much. It's just not. The thing we're kind of tuning for, I think, as an industry. But our friends over at 37 Signals, DHH and Jason Fried, are talking today quite a lot about their use of sabbaticals. This is a six-week break for workers in their third year at the company as part of their overall kind of like kind setup for their staff, including four-day work weeks from May through September. And this struck me as just out of the blue because we're talking about laying off 14,000

• 20:21Speaker 00

  people at Amazon, not trying to keep people. So, one, are you in favor of sabbaticals of this type? And two, are they just drinking different Kool-Aid than everyone else?

• 20:31Jason Calacanis?

  One of the great challenges of running a startup is getting your business and engineering teams on the same page. It's almost impossible to build the sleek, effective product your customers need if these teams aren't communicating well. If this sounds anything like your company, you need to check out DevStats. DevStats integrates your dev work and your business goals into a shared language that everyone can understand. One team using DevStats was dedicating half of their dev time to low-impact bug fixes and other internal busy work, and none of this was improving their product or, more importantly, delighting their customers. But after using DevStats, they were able to revise their allocations and workflow and start actually moving the company forward with impactful updates.

• 21:16Jason Calacanis?

  So it's time to start focusing your team on the work that actually matters. Check out DevStats today and we're going to give you 20% off plus access to their dedicated Slack channel. Just go to devstats.com/twist for your 20% discount. That's D-E-V-S-T-A-T-S dot com slash twist.

• 21:42Jason Calacanis?

  So they are a highly profitable software company that has existed for a long time with a loyal base of users and it is pretty reliable. Those revenue streams, people who use Basecamp, they just love it. They're never going to stop using it. Okay, period, full stop. So when you start a company and you don't have a bunch of investors, which they don't, I think they bought out Bezos, who was their only investor back in the day.

• 22:10Jason Calacanis?

  They can build the world they want to live in. Now, you remember, they had a whole DEI woke thing happen at their company where people wanted to bring all their opinions to work and debate it on Slack, just like at Coinbase. And we had this sort of pendulum swing the other way. Hey, if you're at Coinbase, you're here to do one thing, you know, provide crypto solutions to the world. If you're at 37signals, they took a similar approach. Hey, if you want to be here, you're here to make great productivity software, full stop, and that's it. And then here's an incredible pay package. I think both companies just said, hey, if you want to opt out, we get it. So during that culture war, which we forget about, but I guess that was 2017 to, you know, COVID, basically, they went through that.

• 22:54Jason Calacanis?

  Then whoever's left, what's happening today is companies are getting more efficient. If the company gets more efficient and you own a business, you really appreciate stability. You really appreciate the fact that Alex shows up three days a week and has a suit and tie on, or I didn't tell you about the ties. He has a pressed shirt on. Don't make me buy ties. You never know. So you start to appreciate that, right? And then when you lose a couple of video editors, you know, because they want to go make the next great movie, or they want to be a screenplay writer, you're like, oh my God, I own this business and I'm going through this again. I got to hire that person. Oh, you know, start the process again. It becomes a distraction from growing your business.

• 23:39Jason Calacanis?

  So these kind of offerings seem crazy if you were running a large company, and they would be. They would create entitlement, they could create abuse on the margins. However, with a small company, you can make a lot of decisions like this that are quixotic, but you can be 20% inefficient over the summer or for four months. You can be, if six weeks is like giving, over three years is like giving people an extra two weeks of vacation, essentially. If you've built a company that's resilient and you have plenty of people and you're overstaffed by 20%, it doesn't matter if people go on sabbatical for six weeks. So you can build companies like that.

• 24:20Jason Calacanis?

  Now, if you tried this strategy and you were up against OpenAI and XAI, you would fail miserably. So you do need to understand the game on the field. If you had venture capitalists and you tried this... And the company wasn't growing. They'd be like, hey, maybe people should come in on Fridays since we're not growing. But we will see in the United States, I predict.

• 24:39Speaker 00

  Okay.

• 24:40Jason Calacanis?

  Uh, by the end of our lifetime, the emergent of, the emergence of a four day work week. I think, uh, you know, a four day, you know, not this like five, six hours that people are working and going and getting matches and yoga, whatever, like, you know, the full on 10 hour day, four 10 hour days might actually, for some people's rhythm, be better than,

• 25:01Speaker 00

  Oh, yeah.

• 25:02Jason Calacanis?

  five, eight hour days or five of these like six hour days where you do a life in the thing. So I think it's worth noting here because you have to know the game on the field and the context. The context is highly profitable company, extremely stable revenue, no investors, two owners. They can do what they want and they can pursue whatever they want. I remember when I worked in New York, uh, I would go to Chanel and we managed their IT systems, uh, like the luxury brand and they all were allowed to leave work at 12 o'clock or 12:30 on Fridays during the summer.

• 25:35Speaker 00

  Summer Fridays, yeah.

• 25:36Jason Calacanis?

  Fridays. Conde Nast, the same thing, but they had to get in at 8 a.m. So they had to like put in at least a half day at work and, you know, they were very pampered. Everybody had an assistant. Nobody was really working. Uh, you know, it was, they were selling $20,000 handbags. So let's just leave it at that. That's the equivalent of 37 signals. They're, they're Chanel in this equation. They're not grinding it out versus 20 competitors who are going to eat their lunch.

• 26:00Speaker 00

  Yeah. Question about this because that's a kind of what I thought you were going to say about this, which is this applies to a certain class of companies, not most startups and probably not places that a lot of people are going to end up working at because 37 signals is only so big. Um, but don't we want to create an economy in which people can work a little bit less? I feel like we talk about this usually as a, as a. A crisis, like if we work less than forty hours, the we're all going to hell in a handbasket. And I'm just kind of curious, like if that's the world we're trying to build, Jason, from your view.

• 26:33Jason Calacanis?

  Depends on the culture you're in. America and Japan and China want to be at the top of that GDP list. And if you look further down, France and Italy, you know, they are known for maybe not wanting to move up that list, but wanting to have a better quality of life. So you can pick the culture you're in. And then if you were to look at the outlier companies, the greatest companies in the world, if you pull up a list of by market cap, it's no coincidence that the top companies by market cap come out of the United States, China. And then once in a while, you know, when you get down past twenty five, you got a Ramco on there from Saudi,

• 27:16Jason Calacanis?

  and you've got LVMH from France, but you don't really have many outside of.

• 27:21Speaker 00

  No.

• 27:23Jason Calacanis?

  When's the, I mean, Ramco and LVMH putting.

• 27:26Speaker 00

  Ramco is number eight.

• 27:27Jason Calacanis?

  Right. And then so not counting China and US companies, you have a Ramco at eight.

• 27:32Speaker 00

  Mm hmm. And then you do have one company from South Korea, Samsung. You have ASML.

• 27:37Jason Calacanis?

  What number is Samsung?

• 27:38Speaker 00

  Samsung is twenty two and ASML is twenty five and then.

• 27:43Jason Calacanis?

  Where does ASML look at it?

• 27:45Speaker 00

  ASML is right here. It's Dutch, Jason.

• 27:46Jason Calacanis?

  Dutch, right. They make the fab machines famously for TSMC.

• 27:51Speaker 00

  And then the first country from France, number thirty three, LVMH, just besting Procter and Gamble here in the US. So I don't disagree with any of that, but I'm just, I guess, as we work on humanoid robotics, as you work on automation, as you work on more intelligent digital systems, I don't know. I think it'd be kind of cool if everyone could work four days a week, hard for those four days. But it, I, we haven't really managed to give humans extra time for arts and philosophy. And I think that. We should keep that in mind that that was a goal at some point, giving people more of their life back.

• 28:24Jason Calacanis?

  Well, if you think about it, YouTube and podcasting and crowdfunding and Patreon, and Patreon is actually named after patronage, is actually made a pretty big comeback and it's been done in a kind of capitalistic way. The people who do Patreons are making full-time salaries in many cases, but working the hours in which they prefer. So that opportunity is available for rugged individualists and brave people who start companies. I do think as a society, we'll see this manifest itself in two ways. The 54% tax in New York under Mondami is the largest in the nation, followed by, I think, California, just right behind it by maybe three or four less points.

• 29:07Jason Calacanis?

  And then we talked about on All In, this floating of the wealth tax. Hey, anybody who's a billionaire is going to have to pay 5% on their assets. So if you have a billion dollars in, I don't know, metashares and you were, you know, some senior executive there who hit it out of the park, you just have to give $50 million one time, which means you'd have to sell 100 million in shares to the government, I guess. Or maybe you would not have to pay tax on selling those. You actually have to pay the capital gains and then pay the wealth tax. So you get hit twice. We're gonna, there's a reason why that's getting popular and those ideas are floating and they're going to win with certain folks because they would like to see the government get bigger. And then if the government gets bigger, they can do things like provide universal healthcare

• 29:50Jason Calacanis?

  and hire more nurses and orderlies and doctors, and then you could hire twice as many teachers. There is a possibility that you would even hear tech and rich people endorse this. You might hear, in the next couple of years, if the wealth keeps polarizing like this. You can't make your product better without listening to your customers. But how are you supposed to actually figure out what your customers want? You could send out a survey, of course, but people just want to give the right answer in that case. They're not really giving you their honest opinion. Well, at This Week in Startups, we actually found the perfect answer. It's called Perspective AI. Their expertly trained AI conducts one-on-one interviews with your users based On your prompts and questions, just tell their system whatever you want feedback on in

• 30:35Jason Calacanis?

  simple language and they take care of the rest. We've learned so much about our audience since teaming up with perspective AI. Reginald, for example, is a founder who listens and wants more quick hit faster paced content and an anonymous guy from Canada gave us tons of helpful feedback about some audio syncing issues we've been having and we fixed them. This is the kind of feedback that's invaluable when you're a founder who is product obsessed like you should be. Sign up today and get started in just a few minutes at get perspective.ai slash twist and you'll get your first two months free. That's get perspective.ai slash twist.

• 31:13Jason Calacanis?

  Where the top half owning equities keeps doing so well, you will hear some folks who are notable and affluent and rich say, you know, charge me a little bit in taxes and why don't we double the number of teachers and have them work four day work weeks. That would be a way to deal with unemployment. So if it is true that, you know, the robots are going to replace a lot of people at Amazon and Amazon's laying off 30,000 white collar executives, UPS is, you know, letting go of management as well. And the robots take the jobs of drivers and delivery individuals or factory workers. You might see some folks who have benefited from that say, yeah, charge me a little bit more in taxes and let's increase the number of teachers, the number of people who work

• 31:55Jason Calacanis?

  at parks, increase the budget for that because I really don't want the guillotines. That's bad. The guillotines are bad. I mean, I was, I really don't care if I paid a little more in taxes or not. It doesn't really change my lifestyle. I think it's like one of the things rich people need to also consider with when it comes to tax cuts like.

• 32:18Speaker 00

  Yeah, yeah, I think, I think that's fair. I am legitimately concerned about the health of labor and capital relations in this country over the next five or 10 years. Not, not just because of one mayoral election in one American city, but because I think that the people who don't have money are doing much worse than people with money think they are. And I think that's going to lead to some surprises politically down the road. And I think we should get ahead of them.

• 32:43Jason Calacanis?

  Absolutely. Yeah. I mean, it's happening right now. This is the conversation. I did a tweet today where the cost of living has gone up. For everybody. Yeah. But what I tried to explain to people is, you know, now, having moved from the bottom third to the top third of income earners in the country, and perhaps maybe more, which is embarrassing, and I got lucky 12 times, so, you know, and a lot of things broke in my favor, like being born at the right time. Putting all that aside, I do think what I learned is, rich people and affluent people, they don't think about their cost of living,

• 33:25Jason Calacanis?

  and in fact, they think about the opposite, Alex. They think, what could I do to upgrade my cost of living? Yeah. Okay, maybe I should get a new car. Maybe I should remodel my ski house. Maybe I should refurnish something. Maybe I should redo my wardrobe. Maybe I should get a chef to come twice a week to feed the family. They actually think the opposite. So, one group of people is sorting a search result when they're shopping by lowest to highest price, and you always wonder, who's doing high to low, right? The high to low is there. Why is the high to low there? Well, some people, when they're buying a sweater or a shirt, I used to buy my shirts, and I'd be looking for the cheapest white shirt. When

• 34:10Jason Calacanis?

  I started looking at shirts, I said, let me sort it from the top. I want to know what the best is and why. So now I'm looking at $800 Brioni shirts. Okay, that's too much. Oh, Eton's 300, that feels like the right thing. I used to pay 100 bucks a white shirt. Now I'm doing 200 or 300. They're looking to upgrade their lifestyle. So that's important for founders to understand. There's always some group of people who are willing to spend, but I think the bottom half, and this is where the job displacement, if it is not managed properly, and we can manage it, we have managed it before, if it's not managed properly, and we keep doing tax cuts on the rich, and we keep doing things that stimulate the stock market. You know, this is going to increase that chasm between the two groups of people.

• 34:51Speaker 00

  Yeah, it's worrying. Jason, I also have some data here for you on this, just to kind of illustrate the point. If you look at the approval rating of the president, just to bring in the kind of current temperature of the folks out there. One thing you'll see is that his net approval rating is most negative on inflation and prices, which I think just goes to show how stretched the average consumer is. And here is the data, if anyone wants to get a closer look at it. If you're on the audio version, this just tracks changes in polling numbers for different vectors across the US market. Inflation, prices, jobs, the economy, taxes, immigration, and NatSec. And they're all trending down, but inflation prices and jobs in the economy are where Trump has lost the most ground. I think people are mad.

• 35:32Jason Calacanis?

  Inflation is, yeah, negative 33, it looks like 34% net approval. So that's when you take the approve and the disapprove and you subtract them from each other and you get, you know, people are disapproving on inflation and prices, when the truth is, inflation has been modest, you know, like 3%, 2.8%. It hasn't been crazy. And if inflation isn't crazy and people are not feeling good about it, you have to wonder why. Well, there's also wages and wages haven't been booming. They've been muted for many years, in fact, many decades, wages haven't grown.

• 36:11Jason Calacanis?

  So the confounding part about inflation is it's not enough to have inflation not increase, because the prices of a steak are jaw dropping even for me. I look at a steak now and I'm like, wait, what? It's $40 for this Whole Foods steak, it's $50 for this tomahawk, it's $60 for this tomahawk? Am I at a restaurant? I gotta cook this myself for 60 bucks? And it's like, yeah, steak's really expensive now. It's not the, and if it doesn't go back down, people are having a permanent price increase, which is, you know, the reason why this is polling so bad.

• 36:53Speaker 00

  If you wanted to just put that into chart form, Jason, here's a chart from Fred. This is how consumers feel inflation, which is here is the COVID, post-COVID inflation bump, and then here's inflation since then.

• 37:06Jason Calacanis?

  Economists will say, oh, it's down to 3% gain, but the working person just feels the total gap here in price change. So that's just a reminder for everybody. But look who has joined us. It's our friend Shishir. I was gonna say from Grammarly, which is what your company was called the last time we had you on the show when you bought Superhuman. But Shishir, I think you have a brand new corporate identity to talk about.

• 37:29Speaker 02

  We do. We do. Nice to be back on the show and yeah, a couple of brands that I know you all know well and love well. So we're excited about that.

• 37:39Jason Calacanis?

  Of course, I'm huge fans of Grammarly and was the first investor along with Dharmesh, who Raul told was also the first investor. He told us both were the first investors. I don't know who got their wire in first, but Dharmesh probably got in first because he was probably doing it from his personal account. I have to do it through a whole series of...

• 38:01Speaker 02

  Dharmesh is fast.

• 38:02Jason Calacanis?

  He's fas, yeah. Our friend who's the co-founder of HubSpot, of course, and a great angel investor. But you have the amazing Coda, which we built the twist500.com list on, the top 500 private companies. Superhuman, which my whole team uses day in and day out and we pay for. And Grammarly, which my whole team uses, which we pay for. So I think we might be getting some relief here. You have three different brands. Tell us what you're announcing today.

• 38:28Speaker 02

  Yeah.

• 38:29Jason Calacanis?

  Vis a vis the brands and the pricing.

• 38:32Speaker 02

  All right. So we're launching three big things today. We're launching a new company name. We're launching a new product. We call Superhuman Go. And we're launching the new suite. So each of them are separate, but they do tie together in an important way. The company name, we're changing from Grammarly to Superhuman, a much...

• 38:53Jason Calacanis?

  Now, that's a real decision you had to make.

• 38:55Speaker 02

  Very big decision.

• 38:56Jason Calacanis?

  Because Superhuman's obviously a more expansive name. Okay, we all want to be superhumans. Great. It's a very aspirational, awesome name. When I heard that name from Raul, when he was starting the company, I loved it. And I am a name snob. I love that name. Now Grammarly, though, is one of those great iconic names, and you have tens of millions of people using Grammarly.

• 39:19Speaker 02

  40 million daily active users, yes.

• 39:20Jason Calacanis?

  Yeah, I mean, like 40 million people who love Grammarly and they, it is what you say it is. It helps you have perfect grammar and spelling. It makes you a better writer.

• 39:32Jason Calacanis?

  But you want to do more as a company, so you are keeping that name, but you're going to try to also have Grammarly fall under the Superhuman umbrella. This is, uh, must have been a big debate?

• 39:43Speaker 02

  It was a huge debate. Yeah, it was a huge debate. I mean, I think there are obviously companies that, as their product suite expands, we've seen other companies do it. It's how Google evolved in the alphabet over time. As you know, I used to work on the YouTube team, and we started building a set of brands there. I think similar, similar here. Um, so I think the idea of a corporate brand being different than the, than the product brands, I don't think is, uh, is that surprising. I think, given the breadth of things we're doing. I do think it was a very hard choice in the name itself. Uh, I think this particular transition of pulling a sub-brand up to be the parent brand is not a very common one. Uh, I was watching Twitter this morning for what people were commenting, and somebody made the list of all the companies that have done that before. It's not a very long list of, of...

• 40:27Jason Calacanis?

  Oh, really?

• 40:28Speaker 02

  Uh, yeah.

• 40:29Jason Calacanis?

  Yeah, what was on the list? Any notable ones on that list?

• 40:31Speaker 02

  The most famous one was probably SBC and AT&T. Um, where the SBC bought AT&T and then turned AT&T into the list. Uh, another one was Ziff Davis. Uh, J2 bought Ziff Davis. And I think, you know, there's obviously, uh, a, a brand that felt like it appealed more directly. And there's some similarity here, that obviously the, the, the, the biggest product in the set is, is Grammarly. It's definitely the one, uh, that touches the most, um, the, the broadest audience. But if we thought, as we thought about a proper corporate brand that can house the breadth of ambition, you know, Superhuman was the obvious, uh, obvious choice. And it's sort of reflective of, uh, you know, in my mind, I, I was drawn to it for two reasons. It's breadth, uh, but also it's, it's message. I mean, I think the breadth is obvious.

• 41:14Speaker 02

  Superhuman can, can apply to so many different products. The empowerment message or the, the alignment to our mission, I think, is, is more interesting, uh, because the word we were drawn to, I think a lot of people are drawn to the super part of the word, but we, we were actually drawn to human. From our perspective, we're one of our unique approaches across the whole company, but Grammarly is our roots, is that we are the ones that make people more human. Everybody else is trying to get AI to replace humans. We're going the other way around. You know, when you use Grammarly, we'll, we'll give you lots of assistance. But at the end of the day, you send that email, you post that article, you submit that essay. It's still you. We're just here to help you. So from our perspective, I got this line from one of our customers.

• 41:58Speaker 02

  We spent the last 16 years turning people into super writers. Now we can spend the next few decades turning them into super humans, better versions of themselves.

• 42:06Speaker 00

  Okay, so Shishir, if the, if the focus here is on empowering humans, comma, you also just launched a new AI agent tool companion thing. Seems to be a slight bit of tension there.

• 42:18Speaker 02

  No, actually, I think it's, I think it's very aligned. So, right, so second big announcement, we launched a new product. We're calling it Superhuman Go. It's launching today, you can go give it a try. It's our new AI assistant product. It's a platform that brings your network of proactive and personal AI assistants directly embedded into your work. And the way to think about this is it really is taking the core technology layer of Grammarly, pulling it out as a platform for all agents. So I think last time I was on, on the show, I talked a little bit about that. That idea that the, the core ability that Grammarly has is to bring AI directly into your surface and work right where you work. And all of a sudden, we can do that with a much broader set of use cases.

• 43:00Speaker 02

  If you picture Grammarly, it's almost like, you know, you have this, this English teacher on your shoulder that is helping you with your grammar and spelling everywhere you go. Now, similar to Grammarly, but with a new set of such agents, as we call them, they can all sit on your shoulder too. And so if you're a salesperson and you're writing an email, you can now not only have Grammarly help you, you know, tune your grammar, make sure you sound upbeat, but you also have your sales agent that can go make sure you pull all the right information from your CRM. You have your product marketing agent that makes sure you get product information. And your calendar agent tells you when you can actually meet the person.

• 43:34Speaker 00

  Shishir, are, are you guys building all these agents? Or are you going to expect the customers to make their own agents? Or is it going to be kind of one agent that just does everything?

• 43:43Speaker 02

  So we are building a set of agents, the most important of which is Grammarly. And we're going to try to make sure there's a great set of first-party agents from us. We're opening up the platform so many other people can do it. Today, about a dozen other partners also launched their agents with us. Some really exciting ones in that set. And we're enabling customers to build their own. So if you have something that is going to connect to your own internal system or so on, I think that's a very common use case. And in that way, we can open up this platform, this ability of bringing AI right to where you work, to any of these products and tools.

• 44:18Jason Calacanis?

  Yeah, that seems really wise to me. The thing I love about Grammarly is that it persistently follows me around my desktop. It's on my keyboard, on my phone. And it feels like it's starting to learn and become a little more personalized, which is to say, the name Calacanis, it should understand after a certain amount of time. And in a company, and I talked to Raul about this incessantly with Superhuman, was multiplayer mode. And multiplayer mode is so powerful because in Superhuman, when you send an email thread, sometimes it's like, oh, this is a customer support thread. I really should get the customer rep CC'd on this. So now I'm adding people. Oh, and I should have the head of customer support do this. Oh, and I should also have the sales manager.

• 45:02Jason Calacanis?

  Now you get three people, and then everybody's email box is flooded, and everybody's day is trying to get to inbox zero, which Superhuman helps you do. But wouldn't it be better if you could just start a chat alongside the email? So now you can just hit the little chat button and say, who's managing this? In our case, who's the primary contact for this startup? And are they aware that they had an acquisition offer? And then I can just CC one person. That person can CC somebody. And now we've got a thread going alongside the email. And you don't have to forward the email because the email's right there. So now you've lowered the amount of emails, and you've created a parallel universe. Those kind of parallel universes kind of speak to what you're doing with agents as well,

• 45:44Jason Calacanis?

  because the agent can then, it would know, oh, in the coded database, the primary contact for that startup is this person.

• 45:52Speaker 02

  That's right.

• 45:52Jason Calacanis?

  Or here they are in your HubSpot, or here they are in your Salesforce. And this is what everybody wants.

• 45:58Speaker 02

  Yeah, maybe, I mean, just breaking those two thoughts apart, I mean, I think that the collaboration part, the feature you mentioned, we call that Shared Conversations and Superhuman Mail. Super popular and just so people can see all the value of it. Not only does it help you pull on the right people in the conversation, those people are not on the mail thread. So I think you've always had the problem of, you know, you send a mail and then you forward it to a group of people and then you get a reply from the customer and then they have to get remembered to forward it again. So you sort of get this side channel alongside the conversation. I use it all the time with customers, with candidates, with,

• 46:31Jason Calacanis?

  Yep.

• 46:31Speaker 02

  you know, with support issues, so on. The other thing you said that I think is really important is what do people really love about Grammarly? It's that it works everywhere. And we, we work in a million unique applications. That's every web app, every desktop app, every mobile app. You know, many people know us from our Chrome extension, but we actually work in everything. We'll work in, you know, Word and Slack and we'll work on, like you said, on the keyboard. We can, we can work in your phone as well. And so that ability to bring that set of context to all your applications, I think is very unique to our approach. And I, to be honest, I don't think there's a lot of players in the market that are doing anything close to that.

• 47:08Jason Calacanis?

  No, it's kind of operating system level information on the user, which basically boils down to Google with Chrome OS, Windows and Microsoft and macOS at Apple and you guys.

• 47:18Speaker 02

  Yeah, I mean, interestingly, I think it's deeper, right? I mean, I think the, the operating system can do a lot of things in your windowing system and so on. But just as an example, we've done the hard work to integrate right into your Google Doc. So we show up with an underline under the perfect word. When you're writing in a desktop app like Slack, we can show up right in the middle of it and tell you that you're about to schedule this, this meeting and actually three of the people can't be there. Or you're talking about this customer and they're actually, their contract is up next week, not this week. And so you, you can, you can bring all that information, all this from context. You know, even the operating system players aren't really getting there.

• 47:53Speaker 00

  Yeah, so this is the question that I wanted to ask you, which is you talked about in your post announcing the change that you built with Grammarly essentially a mass transit system with only one car Grammarly on it, and now you're adding more cars to it. Once you have the tracks in place, you can put as many trains on them as you want to share. So what are you going to buy next?

• 48:12Speaker 02

  I don't know yet. Let's see. I think there's a lot of the, and I actually, maybe I should answer that more directly with the company's going through a massive transition. Obviously, the multi-product transition is a big one. It leads to us having a suite of offerings that you can buy in a bundle. We'll come back to that, I'm sure. The, but it also means we're a platform. So there was a number of companies that we want to bring on this platform. And the de facto case is going to be we work with you as partners. And there's now dozens of them. You know, maybe one good example, you guys, I think, are probably familiar with Kim Scott at Radical Candor.

• 48:44Speaker 00

  Yeah, she's been on the program. I knew her in her Google days.

• 48:46Speaker 02

  Yeah, so she wrote this really popular book. She got, I told her about what we were doing here. She got really excited. The team built an agent. It's a, it's a pretty obvious idea in retrospect. I mean, if you've read the book, you know, the book is fantastic, but what do you do with the book? You take the book, you stick it on your shelf, and you forget about it. You know, what if your Radical Candor agent could actually follow you everywhere? And, and really help you with your messaging and following the principles of the book everywhere you go. So that's a great example of a company we can now work with. We're not going to buy Radical Candor. We are definitely going to partner with them and work with them. And so I think that will become our, our default motion is to partner with companies. There are cases where we want to buy and own products. I think those are really cases where we want it to be part of the core platform.

• 49:29Speaker 02

  And the reason why it was so important to own Coda and so important to own Superhuman Mail is that they're the surfaces you actually work in. And I'm sure we'll do others. But I think the predominant motion for us will be actually partnership.

• 49:43Jason Calacanis?

  What I love about this is right now I'm playing three different people in your organization. I pay for Superhuman, I pay for Grammarly, and I pay for Coda. Happy to do so. It's thousands of dollars a year each, but I get plenty of value for those dollars. But here, you can get Superhuman free with a little bit of Coda and Go and Grammarly. You can pay for Pro, and that's only 12 bucks a month. And you get Grammarly, Coda, and Go, or you can get Business, and I guess I get everything there.

• 50:17Speaker 02

  And Mail.

• 50:18Jason Calacanis?

  And Mail.

• 50:19Speaker 02

  Right.

• 50:19Jason Calacanis?

  So this is like a really interesting offering because I don't know if I'm going to save money or not. If somebody's paying for all three, do I save money or do I net out?

• 50:29Speaker 02

  You'll definitely save money. I mean, I think the, if you just, and you're probably on the equivalent of the business tier on all products. And so you'll, you'll end up saving money. And the principle of it is to allow the products to spread and try to make sure that if you're a super fan of one of the products, you can easily get access to the others as well. Bundling is a topic that I've been somewhat known for and spent a lot of time on. I wrote, I wrote a paper about it a while ago called "Four Minutes of Bundling", ended up joining the Spotify board to help build out the Spotify bundle. I think there's a, there's a lot of different core concepts there. But the, the main idea, you know, if I just start with, for example, if you happen to be a Mail user, or a Grammarly user, or a Coda user, you know, the, if you're a super fan of one of them, you might find interest in the others.

• 51:15Speaker 02

  But we, we don't have to convince you as much to try them. Because you buy one, you kind of get a taste of all of them, and it eases the ability for people to get to a multiproduct state. So that's very.

• 51:25Jason Calacanis?

  What do you think about these automatic transcribers for Zoom calls, etc.? You don't have one in the current offering, I don't think.

• 51:35Speaker 02

  We don't, although the Fireflies team did build an agent for this launch, so there's a, there's a Fireflies agent.

• 51:40Jason Calacanis?

  Oh, cool.

• 51:41Speaker 02

  There'll be more. I, I think there, it's a starting point for what they'll be able to do with it. So I do think it's an obvious platform offering. I think the, the. And I'm really excited about that space when I think that there's an enormous amount of knowledge that is not written.

• 51:56Jason Calacanis?

  That's why I was asking, yeah.

• 51:57Speaker 02

  Voice. And so I think that's really important. Uh, I, you know, personally, I think the biggest problem with them is that they all forced me to change the rest of my tools too. That when I, when I use one of these note takers, it goes and sticks the notes somewhere else. And you know, if I happen to be someone who is trying to take notes in, uh, you know, a Coda or a Google Doc, or maybe I'm in Salesforce or Greenhouse, or I'm just trying to take my notes there, you know, none of these tools really let me do that. Um, and so I think we provide that, you know, what we call the I super highway or this, this set of, this set of, uh, um, uh, tracks to be able to get those agents to work right where I work. So that, that's what we'll do in this category.

• 52:37Jason Calacanis?

  Yeah, it looks like I'm spending two grand a month, 24 grand a year on all three of these. Uh, maybe I save a little bit and that's all good. Um, and you reward your loyal customers. Yeah. Right now I had a, uh, air traffic control issue. I would pop on a Zoom or on a podcast. Granola would say, I'll record. Zoom would say, I'm recording. And then Notion launched, uh, their own recorder. And it was like, um, and we're like, well, we're paying for that as well. So I banned Granola. I was like, guys, I, I can't pay for three different products just because one third likes the interface. If you really want to pay for it, pay for it yourself.

• 53:11Speaker 02

  Yeah.

• 53:12Jason Calacanis?

  But we got, we got to standardize on something here. And, um, then I guess the last piece, the last tower to fall for you is to start thinking about, uh, Zoom, uh, or huddles on Slack and then Slack or Teams itself. You're building this team product. I was always talking to, uh, Raul about when would we get a calendar? And, uh, you know, Zoom meetings and Slack inside of superhuman. We got the calendar, but we don't have, uh, audio, video conferencing, nor do we have, uh, you know, messaging and a Slack-like competitor.

• 53:50Jason Calacanis?

  What are your thoughts on those last two?

• 53:51Speaker 02

  Yeah, I mean, I think if you think broadly about our our stack, you know, the bottom of our stack is now this new product, Go. That's our platform for running agents everywhere. On top of that, we're building a set of agents. So that's the grandly being the most important one. We're building a number of other agents, including a set that connect to basically every application you can think of and opening up that marketplace. And then in a small set of cases where we believe that the surfaces are not really built for agents, we're going to building the first party surface as well. So in every case, you know, you can go use the existing tools, but we're trying to build the the premium version, the deserved version for each of these, each of these categories. And we started with two entry points. One, where people do their productive work.

• 54:36Speaker 02

  That's where Coda fits in. And that's, you know, you produce your artifacts, whether it's your documents or your tables and spreadsheets and presentations and so on. And the other one was in communication and we started with the superhuman mail product. So these are the two that we that we worked on and we'll branch from there. And I think Coda already represents a pretty broad set of productivity tools. And you'll see us add to that set as well. And similarly, for the superhuman mail product, it represents, you know, the the base of communication. And it's already started broadening from there. You could you've already probably seen on the calendar side, for instance, we're awfully close to covering all the use cases that you would need a separate calendar product for. And so we're working our way methodically through through the other ones. But we want to, you know, in cases where we think a first party surface can do a better job,

• 55:22Speaker 02

  we're going to make sure that we have the best possible offering for it.

• 55:26Jason Calacanis?

  Amazing. Do we have any questions from our live audience, Marcus? Were there any questions from the live audience?

• 55:33Speaker 00

  Yeah, one that I saw was, according to the superhuman mail website, you can spend forty bucks a month on the business plan there, Shishir, and thirty-three bucks a month for the entire suite. Are you not giving away too much value slash revenue? That's kind of the tone of the question.

• 55:48Speaker 02

  First of all, just to correct them, the the pricing is forty dollars a month is a monthly price, thirty-three dollars a month is is the if you pay annual price, the the pricing is about the same. But it does mean it's what I like to call One of N bundle, so maybe just a small amount of bundling theory is that when you think about a bundle that has, say it has, you know, a hundred different products in it, you could say, if you look at the bundle, you can assess the value of the bundle based on how many products in the bundle you need to be a super fan of in order for the bundle to make sense. Maybe to give you an example, the Adobe suite, I think it have like eighty different products. But if you, they're each priced at thirty bucks each, but at eighty seven dollars, you get all of them. And so it's what I call three of N bundle. That's how, that's how I think about Adobe.

• 56:31Speaker 02

  We decided in this case to start with what we call a one of N bundle. So if you are a paid and happy user of Superhuman Mail, then for the same price, you will also get two, now three other products alongside that. Same the other way around. So the Coda, if you're a loyal Coda customer and you're paying for the Coda team tier, it's about roughly the same price. You can now say, well, now I can deploy Mail Grammarly and Go to all my users as well.

• 56:58Jason Calacanis?

  Here's a fun question from Andre. What do you think about training users to write in the style of a specific thinker like Duski Efski or Kant? Is it Kant or Kant?

• 57:08Speaker 02

  Kant.

• 57:09Jason Calacanis?

  Kant.

• 57:11Jason Calacanis?

  Through sentence suggestions with L L Ms, it's now feasible. What do you think about, yeah, helping people? Yeah, you, you, there's actually, before Grammarly, there was

• 57:23Jason Calacanis?

  Write Like Hemingway. There was like a Hemingway app that was like, hey, here's how to write like Hemingway, use like five word sentences, which can get a little annoying.

• 57:32Speaker 02

  I mean, I think, I think.

• 57:33Jason Calacanis?

  No semicolons.

• 57:34Speaker 02

  I think it's a really interesting idea, actually, but I think, I think the way they phrase it almost felt like it's just fun, but there's actually a lot of utility to it. I mean, I gave an example earlier with, with Kim Scott and Radical Candor. If I want to write more like how Kim thinks I should write, I want Kim sort of sitting next to me with a marker and saying, here's how you should write differently. If I go in a different direction and think about schools, for instance, we have a, a big partnership, for example, with Arizona State University, and they're building a whole set of agents on, on this platform as well. And for them, this is about bringing your teacher to you. So you, you want your professor to be next to you and not, not in this, you know, today's interaction where they give you homework and then three weeks later, they mark it up and give you feedback.

• 58:14Speaker 02

  What are they giving you feedback? Write as you go. So it's, it's not so much about, you know, I think the right like, like Kant so on is, is like, that sounds like fun, but what does your English teacher really want you to do? Or what's your history teacher want you to do? What's your chemistry teacher want you to do? So I think there's a number of cases for that, that, that allow that to go from fun to utility.

• 58:30Speaker 00

  All right, we haven't.

• 58:31Jason Calacanis?

  Okay, you give the other one, yeah.

• 58:32Speaker 00

  Well, we have another question here from Zier App, and they're curious about customers that are afraid to share their private emails with others. I wanted to ask about this too, which is just the privacy element of this. You're bringing in even more of my information into your, into your domain. What are you doing to make sure that it's both safe from leaking and also safe from potential abuse inside the company?

• 58:52Speaker 02

  Yeah, I mean, I think it's a, I think the trust is at the core of everything we do. All, all three and now four products start with trust, security and privacy right at the core. I think Grammarly has, has built a lot of technology around this over the years to make sure that we're anonymizing, privatizing, so on, all the data so that it can simultaneously be personal for you, but not exposed to, to anyone else. Same with Coda and Superhuman, we know that that's, that's a core of what, what everybody expects.

• 59:23Speaker 00

  Okay.

• 59:24Jason Calacanis?

  But there's such an awesome, I guess there's, there's two levels to this. There's the awesomeness of having the whole organization flowing and management understanding what's happening in the organization. So if you live in a non-siloed organization with an open forum plan and everybody hears everything and everything's transparent, you know, like a Netflix or a, you know, an Amazon, a modern company might be like pretty upfront about those things. Maybe people don't care about it as much in a siloed organization. Maybe people are very concerned about that. But there is a whole nother category of software that I feel like you're going to collide with, which is employee management software or optimization software. Workforce analytics software is.

• 1:00:09Speaker 00

  Do you mean the stuff that tracks us on our work?

• 1:00:10Jason Calacanis?

  Spider, yeah, basically. So things that, you know, make sure that people don't, I don't know, download a database or use their computer for something they're not supposed to. But these things also, and I talked to one of the providers, like, could it like suggest to people like, oh yeah, you're talking about this, these three people in the organization know about that customer. You should talk to them. There's some way for it to, you know, kind of make you aware that, oh, you work in, you know, IBM's office in Albany, you really should be talking to the people in Fort Lauderdale because they're doing something, you know, where they're doing an integration, you know, with a trade organization and they probably would have the right answers for you. So.

• 1:00:55Jason Calacanis?

  Have you started to think about that?

• 1:00:57Speaker 02

  Yeah, absolutely. And I think as we think about agents in particular and building this agents platform, it's a really important tenant of, like you said, there's a trade-off there because you want to bring people information that might be usable for them, but you need to understand and respect the permissions underneath them. The agents platform is actually being built on top of a component we pulled out of the Coda product that we call Coda Packs. And Coda Packs were and have long been the platform for integration that Coda, probably the thing Coda is most known for is that Coda integrates, there's about eight or nine thousand of these packs that have been built, everything from Salesforce to Jira to so on. There's eight or nine hundred of them that are public in the gallery and those were all built with this,

• 1:01:39Speaker 02

  you know, I think we started that project seven, eight years ago now. So we've layered on many layers of privacy and security in that. So they can sync and understand the permissions of the underlying data set. So that makes sure that it doesn't send those anywhere else. It's also a set of controls for your IT team to go and layer on top of that and say, even if those permissions are there, we've decided that only the finance team should have access to this. We want you to respect that boundary throughout the platform. So I think that the core of each of these products is being built with, you know, and each of these teams has been in front of some of the toughest organizations in the world and, you know, gotten, you know, all deployed with thousands

• 1:02:18Speaker 02

  of, you know, every IT team and security team has scrutinized them and we're bringing them together into one safe environment.

• 1:02:27Jason Calacanis?

  All right, listen, continued success. I'm a shareholder via my super human share. So I am rooting for you. The logo is great, by the way, over your right shoulder. People who are watching can see the logo and I really love what you do with the logo because it's a person, but it's also the navigation.

• 1:02:47Jason Calacanis?

  Yeah, it's like one of those little arrows. That's like the cursor, but with the head on it and it kind of points you in the right direction of what to do next. So. How much does it cost to do this kind of like branding madness? You spend a hundred grand on this?

• 1:03:01Speaker 02

  We did a lot of it in-house. But I will say, one thing to note about it is, you know, all three companies, Grammarly Code and Superhuman, had a letter as their logo. So picking a character was actually a really good, really big choice. And you'll actually see it in the product. Uh, one of the things we love about it is that it animates very well. So this, you know, it feels like a person, feels like your assistant that's living, living next to you. And I think that allows it to feel active, allows it to feel like it's working alongside you. So I'm, I'm really excited about the logo.

• 1:03:29Jason Calacanis?

  Hey, chime in on the, uh, the big debate of the moment, which is, um, we're seeing places like Amazon, UPS, uh, you know, maybe some big layoffs on all sides of the business, white collar, blue collar. And then, you know, people talking about massive efficiency gains. So obviously, uh, companies are doing more with less internally. Uh, and having run companies and worked at these companies, what are you seeing game on the field wise? Is it the case where it's easier for you to empower your team with tools first and then hire second? Uh, how do you think about this static team size that everybody's, seems to be deploying in the world?

• 1:04:08Speaker 02

  I think it's more a shift in skill sets. I don't think it's so much about the overall team size. I think it's a different, difference in what you need. Maybe to rewind time back to when I started technology, I started at Microsoft. That was the first company I worked at. At that time, your average development team was one PM, two developers and four testers. That was the, you know, four out of seven people on the team were testers. You won't find any Silicon Valley company where over half your team is testers anymore. Cause we automated a lot of it. You also wouldn't see any Silicon Valley company with less people than Microsoft had then. Cause we created more opportunity and more things to build. And you know, in my mind, there are certain roles that are going to feel less necessary and that's okay. Um, but there's going to be others that we're just going to push human ingenuity in other, another direction.

• 1:04:53Speaker 02

  So I, I, I think these are all small shifts. Uh, Amazon is shifting a few employees. They also employ, employ over a million people. Uh, and I suspect they're going to continue to grow. Um, but as they do it, they'll shift around the skill sets they need.

• 1:05:08Speaker 00

  Actually, one last question before you go, which is, we were talking about sabbaticals earlier this year, and Jason says that six-week sabbaticals every three years, as 37 Signals do, are a luxury for companies that are profitable and not growing too quickly and don't have a lot of investor pressure. I'm just curious if your newly reconstituted company will offer sabbaticals, or if you're going to keep the nose to the grindstone?

• 1:05:28Speaker 02

  We do. I don't, I don't think the idea of sabbaticals is that controversial, actually. I think people do get burnt out. I think that's okay. Some people never take them, some people do. And, you know, I've generally found that people take them, come back rejuvenated. I think that's okay.

• 1:05:45Jason Calacanis?

  Yeah. Dara moved it from five years to eight. I think he was maybe, you know, having inherited a workforce, maybe a workforce that got bloated, he wanted to maybe trim back all of these entitlements and get people back to in-office. You believe in in-office or remote? What's your take on that?

• 1:06:04Speaker 02

  We, we're hybrid. We do a lot in-office, but we're spread all over the world. Actually, the company started in Ukraine, so that we have a, we have a big team there, and so we spread, spread that direction as well. I mean, I deeply believe in creating in-person experiences. Some of that's by being in-office. We do a lot by bringing people together at regular intervals all year round. The one thing I'd say about it is that you want people to work hard. The way you do it is by creating an inspiring mission that they want to be a part of. That's, that's what motivates people.

• 1:06:32Jason Calacanis?

  First and foremost.

• 1:06:33Speaker 02

  Yeah, I don't think you do it by checking the time clock more regularly. While you want to make that possible, you, I think people want to work with other people. That's why we do in-office. It's not because, you know, that's the only way to get them to work. You get, you know, people do great work when they believe in what you're building.

• 1:06:48Jason Calacanis?

  Yeah. I, I think I have found with young people, I don't know if you have like a training program if you hire people, you know, out of school, but we do, and it really does help with that high bandwidth in-person culture, at least at a venture firm, investing in companies, to be in-person, to have these discussions, you know, to have the, let's go walk for a coffee and talk about this startup, or this decision we're making. Did we make the right decision to not take our Pro Rata or not? You know, like, these are, for young folks, I think, you know, first five, 10 years in the workforce, getting in-person is just so essential for the mentorship, and that's I think that might be where this kind of changes is when you have kids or you're further

• 1:07:31Jason Calacanis?

  along in your career and you have your choice of where to work. Hey, listen, everybody go check out superhuman.com and make sure you're using Grammarly and code is the sleeper hit. If you want to be super productive, it's not an ad. I just love these three products and it just happens to be that I loved two of them and then they bought my favorite product, Superhuman. So I am so excited that you're all you're shepherding all this so well and I'm excited to see all the new products, continued success.

• 1:07:59Speaker 02

  Hey, thank you.

• 1:08:00Speaker 00

  Thanks for sure.

• 1:08:00Jason Calacanis?

  Talk to you soon.

• 1:08:01Speaker 02

  All right.

• 1:08:01Jason Calacanis?

  Cheers.

• 1:08:02Speaker 02

  Bye.

• 1:08:02Speaker 00

  Every time he comes on, I learn a couple of things and he always challenges at least one thing that I'm thinking and I appreciate that.

• 1:08:09Jason Calacanis?

  It is really great to see how how he thinks about the brand and it takes a certain fearlessness to say, yeah, we need to, we got this incredible brand that 40 million people are using, but that brand is a feature of this bigger vision now.

• 1:08:30Speaker 00

  Yeah.

• 1:08:31Jason Calacanis?

  Let's keep going, you know, let's keep turning over the cards and see where this hand leads us. What else do we have in the talk?

• 1:08:38Speaker 00

  Next up on the talk today is Nvidia reaching a $5 trillion market cap. Just kind of a news update for everybody here. We're going into, Jason, as we talked about on Monday, the big technology earnings cycle. Later today, we'll hear from Alphabet and Amazon and Microsoft, I think. But in the wake, in the middle of all that, Nvidia reached $5 trillion, which is probably the single most valuable company of all time, which I thought was incredibly impressive. We've talked also on the show, ad nauseam frankly, about there being an AI bubble, risks thereof, are we spending too much? But it seems that the public markets are completely and entirely unconcerned. I'm curious if you're surprised.

• 1:09:16Jason Calacanis?

  Um, you know, if you look at the price earning ratio of Nvidia, it is reasonable, I think 25 to 34 looking based on their projections of where they're going. So it's not a situation where You have to suspend disbelief like Palantir and it's got a, you know, and the extraordinary price earnings ratio. Or, you know, Tesla right now where you're like, well, Optimus and self-driving are going to dwarf the car business, which makes total sense logically. But this is one where, listen, they sell a product at a high margin. People need them and there's a global market for this product. And it's hitting on all cylinders.

• 1:10:00Jason Calacanis?

  They can't make enough of them and there aren't.

• 1:10:05Jason Calacanis?

  There are, there's a very diverse group of buyers for this. You have cloud providers.

• 1:10:10Speaker 00

  Yes.

• 1:10:10Jason Calacanis?

  You have governments. You have, you know, you have different government agencies even like not just one government like UAE buying a bunch or, you know, China buying a bunch, but you have this specific group of intelligence. You know, the education group, whatever, they're going to buy 100,000 of these, 50,000 of these. There will be competition and at some point the margin will be eroded away. Like happens in almost every business. And then the question is, can they have such an elite product that they can command? A decent market and the other fear people have is the overbuild. Those are rational fears to have.

• 1:10:50Jason Calacanis?

  But if you look at the price earnings ratio, there's still room for that company to grow is I think the consensus, which is why it keeps growing.

• 1:10:58Speaker 00

  Speaking of which, Justin, here you go. This is from Y charts. This is the forward price earnings ratio for Nvidia over time. And it's now up to, I don't know, 46, but that's no higher than it was back in January of 24. So.

• 1:11:10Jason Calacanis?

  And 46 is the trailing. I'm guessing this is the trailing PE.

• 1:11:15Speaker 00

  This is, this is forward.

• 1:11:17Jason Calacanis?

  This is forward for next year.

• 1:11:20Jason Calacanis?

  Okay. Yeah. So that's pretty high.

• 1:11:23Speaker 00

  Here's the, uh, trailing is.

• 1:11:26Speaker 00

  Fifty-nine.

• 1:11:27Jason Calacanis?

  Yeah, so it's pretty rich, actually. It's double the rest of the market. So it's pretty rich, but it's still high growth. If you look at their revenue chart and their revenue growth, it's pretty extraordinary how they're growing. And they have years of orders in that order book. I think everything is pinned off of OpenAI right now. Can OpenAI get to some level of revenue that makes logical sense for this build-out?

• 1:11:58Speaker 00

  Yeah.

• 1:11:58Jason Calacanis?

  You can't spend a trillion dollars and make a hundred billion. That doesn't work. You could spend a hundred billion and make a trillion.

• 1:12:07Speaker 00

  Sure.

• 1:12:07Jason Calacanis?

  You can't spend a trillion and make a hundred billion. And they're making, whatever, fifteen billion? I'm not sure exactly where it's at now, the runway, twelve, fifteen billion.

• 1:12:15Speaker 00

  Thirteen billion was the last number that we heard.

• 1:12:17Jason Calacanis?

  Yeah, so we're in the billion-a-month club, but they're spending a lot more than that. So they're deeply in the red.

• 1:12:25Speaker 00

  Yeah, the lead is that Sam Altman, Sam Altman says 1.4 trillion and 30 gigawatts of capacity over the next couple of years is what they're pipped to pay for.

• 1:12:34Jason Calacanis?

  I don't believe that number. I think that's the number that's not believable.

• 1:12:38Speaker 00

  Is that like a Trump, Japan's gonna give us $550 billion number? Kind of one of those like headline numbers that we kind of all go like, it's probably a fraction of that?

• 1:12:45Jason Calacanis?

  Well, I mean, in the Trump case, eighty billion of it seems to have shown up for these two reactors. So yeah, maybe half shows up. So I think, yeah, maybe in his case, of the 1.4 trillion, it's 400 billion, but I don't think it's gonna be more than that. That seems like a,

• 1:13:08Jason Calacanis?

  it seems like a marketing slash competitor psy-ops to me. It feels like more of a psy-ops to say you're spending all that much and that's your plan so that other people don't try. That's like directed at Zuckerberg, or it's directed at XAI, or it's directed at Microsoft, or Apple. Like, we're lunatics. We're gonna spend money before we have any revenue. And this is why I think the companies that could make a case to be more involved here are the ones who already have money printing machines. Because once the sovereign wealth funds, once Nvidia stops investing in OpenAI, once...

• 1:13:48Jason Calacanis?

  You know, um.

• 1:13:50Speaker 00

  Softbank says no more.

• 1:13:52Jason Calacanis?

  Yeah, at some point, people have bought enough shares in that company, they're going to want to wait and see if the revenue can catch up. So if they went from four or five billion last year to 12 this year, you know, and they're doubling revenue, well, doubling revenue from 12 to 24, and then from 24 to 48, and then from 48 to 100, it's not actually enough.

• 1:14:16Speaker 00

  Not for their infrastructure spend. But what's interesting is you said that you think that maybe saying that number just scared people off. I thought you were going to say they're announcing that number to try to bait their competitors into overbuilding, because we also know that, you know, XAI is building a lot of capacity right now, and NeoCloud's building a lot of capacity right now. So I thought it was the other way around.

• 1:14:36Jason Calacanis?

  Yeah, I think they have a lot of cleanup work to do there. And I saw yesterday, or two days ago, that they got permission from the California Attorney General to do their big flip.

• 1:14:48Speaker 00

  Yes, they have executed their conversion. Their nonprofit is now called the OpenAI Foundation. It still is in charge of the for-profit entity, which is a public benefit corporation. The foundation will receive about $130 billion worth of equity in OpenAI at a $500 billion valuation, Jason. That's about 26%. Microsoft, in the newly rejiggered cap table and commercial agreement setup, is about 27% of the company. So between Microsoft and the foundation, a little bit more than half.

• 1:15:20Speaker 00

  And I think this is a pretty big deal for OpenAI in the very near term, for a number of reasons, but one of which is that now $22.5 billion from Softbank has been unlocked. That was on hold until they completed this conversion. And then there's quite a lot of changes with the Microsoft agreement. The gist of it is that they're still going to be close, but they're no longer married, is how I'd phrase it. But Microsoft is going to be able to hold on to IP rights a little bit longer. OpenAI will pay a little bit less to Microsoft over time. OpenAI did another massive compute commit to Azure, Microsoft's cloud, but also now doesn't have to go to Microsoft for right of first refusal. So it's kind of a slow motion breakup with a lot of goodies, but huge. We've been waiting for this for, what, years now?

• 1:16:03Jason Calacanis?

  Yeah, I can't believe they got it done. Apparently, the attorney generals in Delaware and California gave them the go-ahead. The fact that it's a benefit corporation means nothing. Just so people know, a benefit corporation is just like a fancy label put on top of a for-profit entity. And it sounds like it's a nonprofit. Public benefit corporation just means you have to state the purpose for this organization to exist. And then, when you make decisions as a board, you have to include that in your decision-making process. In other words, the mission of the company. So if the mission is to provide open-source AI for everybody, which it no longer is, you know, they would have to take that into account when they were deciding to go public

• 1:16:47Jason Calacanis?

  or hire somebody or fire somebody. We're just making big corporate-level decisions. And so, but if Microsoft owns 27% of it and this thing becomes worth, you know, gets into the trillion-dollar club or the two-trillion-dollar club, is that meaningful if they were to sell all their shares and they would make $270 billion? How much cash does Microsoft have on hand right now? I think they're always sitting on $50 or $100 billion. But this could be a windfall for them. I'm not sure how much they paid for that equity. It was tens of billions. I think they were $20 billion in. I'm not sure of the number there.

• 1:17:28Jason Calacanis?

  So I guess those are the two numbers you need to see to understand how great of a move this was by Satya Nadella. And I do believe Bill Gates said it wasn't worth doing.

• 1:17:40Speaker 00

  Yeah, he said we're lighting the money on fire. Also, I love it, Jason.

• 1:17:44Jason Calacanis?

  He still might, that might turn out to be true, actually.

• 1:17:46Speaker 00

  I think they've gotten enough value out of it already that they're pretty happy. But it's funny, oh, here, now it's finally loading. You asked me for how much cash Microsoft has on-site, on-hand. I was immediately going to go get that for you. But it's earnings day, so the Microsoft's Investor Relations page is a complete mess. But I'll get you that number right now, Jason.

• 1:18:03Jason Calacanis?

  Okay, so Microsoft, cash, and bank. I'm guessing they have.

• 1:18:08Speaker 00

  Cash equivalents and short term investments were ninety four point five billion dollars.

• 1:18:15Jason Calacanis?

  Yeah, they've always had like fifty to a hundred billion laying around in cash and they throw off obviously tens of billions in profits every year. So this is not a de minimis amount. And then how much did they put into OpenAI? It's the second number in terms of, um, yeah, Microsoft investment in OpenAI. Let's take a look at that. So how much have they invested to date?

• 1:18:39Speaker 00

  About thirteen billion total from what I can tell, um, which means they would be up about ten X, give or take.

• 1:18:47Jason Calacanis?

  Yeah, it's ten X right now and pretty clear. It could get to two trillion. It could get to a trillion, a trillion five. So this could be a twenty thirty X bag around a big number.

• 1:18:57Speaker 00

  On an enormous number and Microsoft gets the commercial benefits along the way. For example, as part of the new agreement, uh, if OpenAI makes a product with a third party that is served via an API, it has to be on Azure. So they still have some pretty good,

• 1:19:12Speaker 00

  uh, like I said, goodies as part of the deal. Um, and they did also redefine AGI a little bit to make things a bit more fair and a little bit less scary. But, um, I think Microsoft's going to stay winning because they're also getting a revenue cut this whole time.

• 1:19:23Jason Calacanis?

  What does this mean for corporate governance and corporate structures? There was another organization that I think laid the groundwork for this, and I believe the lawyers who did this also did OpenAI's work, which was the Mozilla Foundation. Netscape made an open source foundation, AOL, which owned Netscape, the original browser, put, uh, Mosaic was the original. And then there was Netscape, the commercial one. They put all the, um, IP into this foundation. The foundation started making so much money because the Firefox browser had a search deal with Google and they were making tens of or hundreds of millions of dollars. And so the.

• 1:20:07Jason Calacanis?

  IRS was like, this doesn't make any sense. You're a nonprofit making all this money. Aren't you supposed to deploy it? So they started hiring more people, they're paying extraordinary salaries, doing whatever they can to come up with new projects to do. And they came up with this structure of a for-profit owned by a nonprofit.

• 1:20:21Speaker 00

  Yeah, so the Mozilla Corporation is the thing that owns Mozilla developers and does Firefox, and that's owned by the Mozilla Foundation, which also owns Mozilla Technologies Corporation, and that is the Thunderbird subgroup. So it's two for-profits underneath a foundation.

• 1:20:39Speaker 00

  I, okay, let me be honest here, Jason. In a world in which we have offshore corporate structures, and we have all sorts of fun ways to move money around, which I don't actually have that much of a problem with, why is this conversion of a nonprofit to a for-profit-ish thing such a big deal? Because I feel like we let corporations change their clothes pretty often, and let them play the game.

• 1:21:00Jason Calacanis?

  I guess the question would be, is this fair to taxpayers? And is this like a tax loophole that's unfair? So if I were to donate $50 million to a nonprofit, and then that nonprofit and the employees and the board all got equity in this new company,

• 1:21:21Speaker 00

  Sure.

• 1:21:21Jason Calacanis?

  then I was able to get a tax write-off for that donation, and then I would double-dip by having equity in the new company at some point, and become a billionaire. So that's, I think, where people would be wondering, should that be allowed? So if I were to give you an example...

• 1:21:41Speaker 00

  Yeah.

• 1:21:42Jason Calacanis?

  And maybe this exists in the world. I'm sure other people have tried similar things. But we all donate our money into a nonprofit. That nonprofit becomes an accelerator. Like, let's say I wanted to do this for the Launch Accelerator, Founder University. I make Founder University a nonprofit. And I get tax write offs by donating to it. And then I decide there's going to be an entity that invests in them that's owned partially by the nonprofit. And I raised a bunch of money for that nonprofit from other individuals, but then I own the spin out. That's where I think it starts to feel like really conflicted party transactions.

• 1:22:20Speaker 00

  Well, no conflict, no interest. But I'll just say that I think intent matters quite a lot here. I think if you start as a nonprofit and then things change down the road and you want to make a change, that doesn't strike me as as bad. If you do this entirely as a ruse to get around tax rules or to build a better tax benefit for yourself, sure, penalize that. But I think we should let companies that act with good intent make changes that they need to pursue growth. I mean, this just doesn't bother me that much. If we're going to talk about, you know, inefficient taxation policies, I have a few that are higher on my list of grievances than the potential write offs for early OpenAI donors, I guess, in that case.

• 1:22:59Jason Calacanis?

  Or the management. And that's where like Sam Altman owns was like working for a nonprofit, owns no equity, but I assume he owns 5% of the new company. What does Sam Altman own now?

• 1:23:11Speaker 00

  That we don't know. But I do know that he's keeping his nominal salary of $76,000 a year, because that's totally an important portion of his wealth.

• 1:23:19Jason Calacanis?

  I gotta think he owns 5% of the company and rightfully so, if he architected the whole thing like, but, okay, so then how does that play into it?

• 1:23:29Speaker 00

  It's thorny. I do not mean to dismiss dismiss your complaints, but I mean, the alternative is OpenAI is hamstrung and we lose as a nation, one of the most valuable companies potentially of all time, which would generate huge tax benefits for, you know, the public purse.

• 1:23:44Jason Calacanis?

  There's that's a valid argument as well. The question is then, what's the largest nonprofit in the world? And what would happen if they flipped?

• 1:23:52Speaker 00

  Gates Foundation, I presume, right? Oh, here we go. I got some data here.

• 1:23:57Jason Calacanis?

  Yeah, well, that's like one person's wealth. I'm thinking of one that raised money from multiple people, you know, like. If the American Cancer Society decided to flip and the management became wealthy because they came up with a cure for cancer and put it into a for-profit. And then the members of the American Cancer Society then went to work for the for-profit? I don't know. It's just very confounding.

• 1:24:20Speaker 00

  Yeah. I don't know Sam Alvin's take or stake in it, but the information reports that the OpenAI employee vehicle is worth $130 billion as well. The same percentage of the company that is owned by the foundation.

• 1:24:34Jason Calacanis?

  So like 26 or 27%?

• 1:24:35Speaker 00

  Yeah, that's from Amir Afroudi over at the information, who is fantastic. Shout out to him for that data point.

• 1:24:41Jason Calacanis?

  Okay, there it is. Okay, let's keep going.

• 1:24:44Speaker 00

  Next up on the docket is a story that I did not see coming at all, Jason. There's a company called One X. We've talked about them on the show before. They're humanoid robotics. They make the Neo, but we've been talking about Figure. We've been talking about Tesla's Optimus robot. And kind of from nowhere, One X came out and decided to start selling its Neo personal home robot, starting in 2026. Couple of options, you can pay 20K for it, or you can pay 500 bucks a month for it. And it's not ready yet, but it's certainly getting closer. And I think people are going to buy enough of these that it's going to be a worthwhile first launch for the company. Would you like to see a clip of this thing?

• 1:25:19Jason Calacanis?

  Yeah, sure. I mean, it's all over social media. It's a good looking robot. I think we had one of these at an all-in, maybe at the all-in holiday special. It was on stage.

• 1:25:33Jason Calacanis?

  But yeah, it's, looks pretty good. It's petite, small, I think.

• 1:25:37Speaker 00

  It is.

• 1:25:38Jason Calacanis?

  To not be threatening to humans, if that's in fact it. It's got like cloth on top of it and it's currently tele-operated, but they show it like cleaning and loading dishes and picking up packages. So I guess they believe they're going to have this in a year for 20K. The other one that everybody seems to have that's remote controlled,

• 1:26:04Jason Calacanis?

  the Chinese one that was on the Walmart site, that also was 20 grand. So it seems like these are going to be able to be built for 20 grand.

• 1:26:12Speaker 00

  The Unitree H1, and then they just came out with another one that's even more fancy. I think that's their cheaper consumery one. So there could be a two-tier pricing element here, Jason. The Journal, Wall Street Journal, did get its hands on some testing time with the Neo, actually at OneX's headquarters. They sent over Joanna Stern, who's fantastic, and, in her experience, nearly entirely teleoperated. But the idea, according to the company, is that, quote, "Neo's autonomy improves with diverse data and real-world experiences. As Neo does more chores, it will receive updates to your Redwood model, its internal AI, that increase the complexity of tasks that Neo can handle." So I think really what they're selling here is, is early access robots to people who are willing to

• 1:26:54Speaker 00

  beta-test them and help them get more data on real-world use cases in real houses to finish up the AI. And if people are willing to subsidize that, great, because I would love to have this problem fixed. But I don't think Neo suddenly solved the entire general robotics question for the home with no one watching.

• 1:27:11Jason Calacanis?

  Yeah, and I don't think, I saw Lon had put in the notes, like, does this mean there's a teleoperator walking this thing around my home? That's pretty creepy. I think the answer is no. I think the teleoperator is just to show what it's capable of now, just in terms of its movement, and eventually that'll all be done by the large language model, you know, behind it. But these things, you know, a lot of the great innovations start as toys that are supposed to do like some tasks, and they do them badly. The PC revolution was like, hey, boot up your PC, take 16 minutes to load this calculator or chess tutor, and you can play chess or use a calculator. And it was like, this is dumb. People were like, I could just use this calculator. Why would I do that? I have a chess set right here. Why would I do that?

• 1:27:56Jason Calacanis?

  If I want to play Dungeons and Dragons, I have it over here. Why would I play this Zork thing? So, and then, over time, compounding technologies, memory, fidelity, in this case, the...

• 1:28:10Jason Calacanis?

  The actuators will get better and better. These things in three years, it's not going to be uncommon to see one of these things at an airport, at a UPS store.

• 1:28:22Speaker 00

  Yeah.

• 1:28:22Jason Calacanis?

  I wouldn't be surprised if like, there's a driver in a Zooks Likes car and the robot gets out and brings the packages out and puts them on the door and the operator is just taking a picture of it working. You know, we'll see that in three years and then six years, they'll take the driver out and you'll have a Zooks car owned by Amazon with a bunch of packages in it and a robot in it. And the doors will open and it will carry your box from Zooks and put it on your doorstep. And that means it'll be free to ship things around the world or close to free.

• 1:29:02Speaker 00

  It'll go- Cause it's electric power and electric power, yeah.

• 1:29:05Jason Calacanis?

  Yeah, it's all electric and these things are going to cost but a dollar a day. I always do the math on these. If it's twenty thousand dollars and you know, there's um, you know, and you had it for 10 years, you'd have it for 3,000 days. Uh, 3,000 days, you can just very easily back of the envelope this. It's, it's not going to be an expensive thing to own. You know, a couple of dollars a day, these things will cost ultimately.

• 1:29:31Speaker 00

  So, right now, thinking about the math, 500 bucks a month to rent one of these. Okay. So-

• 1:29:37Jason Calacanis?

  Twenty bucks a day, yeah.

• 1:29:38Speaker 00

  What do I pay my housekeeper currently? About a hundred and forty bucks a week to come once. So that means I'm already paying more per month for a human that, to be clear, I love. She's part of the family.

• 1:29:48Jason Calacanis?

  Yeah.

• 1:29:50Speaker 00

  But this is already cheaper than, than she is. And I can have it all the time. It's not as good.

• 1:29:56Jason Calacanis?

  It's not gonna work right now, but-

• 1:29:58Speaker 00

  No, no, no, but if- It's already cheaper. That, that's a great place to start. If it was five thousand dollars a month, we would be 10 million miles away. But if it's already coming in cheaper than a human-

• 1:30:09Jason Calacanis?

  Yeah, these things should be, if I was working at one of these companies, I would not sell these robots. I would only sell them on consumption basis, a dollar an hour, 24 bucks a day. And if you want them, it's a dollar an hour to have one. Plus, you pay for the electric, electricity and, you know, just a dollar a day. Anything goes wrong, we just, we'll go pick it up and swap it out.

• 1:30:33Speaker 00

  A couple, a couple of things about that, Jason. So Neo, when you, when you get your Neo, comes in a weird pod and opens up and starts talking to you, which is pretty cool. There's no real setup for it. You can also, on your phone, go ahead and teleoperate into your own Neo and see out through its eyes and see what it's doing at your house. You can schedule chores for it to do and so forth. So I, I think the blocks are getting there. I think, I think we're getting close. I do, I'm just curious, who's gonna actually get the first one of these out that works? Like to a solid B plus level?

• 1:31:04Jason Calacanis?

  Yeah, the Chinese will have plenty of these operating in China. And I would say they'll probably be first because they're not gonna have the social constraint. They can just decide they're gonna use them for certain jobs. And they'll probably decide to do them for jobs that are truly arduous or dangerous. So imagine, you know, it sounds silly, but if you are fighting a fire or you had to go into a dangerous situation, these would be very good for going into dangerous situations. In fact, they have an equivalent of this, like a little robot with an arm on it that you can use for bombs, right?

• 1:31:43Speaker 00

  Oh, yes, bomb disposal robots, of course.

• 1:31:45Jason Calacanis?

  Yeah, that's a bomb disposal robot. So I think it's gonna be first for like super dangerous things that nobody would ever complain because they don't wanna do. But these are going to, in China, if they start taking factory jobs, that could be very dicey as well.

• 1:32:03Jason Calacanis?

  Very dicey.

• 1:32:04Speaker 00

  I don't think we're too far from that.

• 1:32:06Jason Calacanis?

  Yeah.

• 1:32:07Speaker 00

  All right, last up on the docket for the day. Uh, Uber's self-driving plans are accelerating. Now, Jason, we've talked about Uber working with Neuro and Lucid to build their own robotaxi fleet. Now, we know that that's going to launch next year, at least in the Bay Area. So we now have a location for it. I'm not at all surprised that another tech company is building something in its backyard. But I am a little bit encouraged to know that we have a location and a timeline. It feels like this product has gone from vaporware announcement to more serious. And we also learned today that Uber is, um, working with Nvidia's in-car autonomous vehicle technology to help more OEMs bring level for self-driving to their cars, which will then benefit Uber.

• 1:32:51Speaker 00

  It seems, and I don't want to call it too early, that after giving up on self-driving and selling off their self-driving assets, Uber has somehow managed to come back and is now like a top three global player in the space, and I don't quite know how they did it.

• 1:33:04Jason Calacanis?

  Uh, pretty straightforward. It's a fragmented market. It's a giant market. Uh, so you have geographies, you have people who are making the core technology, the sensors, the cars, and you of course have the network, and then you have maintaining all these cards. And so what Uber is using is their huge network and their balance sheet to invest in many different partners. And so this Nvidia tie-up means that they will be using their driving data from the real world, like Uber, like Tesla does. Uber will take their driving data and give it to, uh, Nvidia. Nvidia is providing a cloud platform and the hardware stack for cars. So that means if Toyota or Volkswagen need hardware or the Neuro cars need hardware,

• 1:33:49Jason Calacanis?

  they don't have to think about that. They could just, you know, bolt this on to a car. Now, bolting it on to a car aesthetically doesn't look good and it doesn't scale like building cybercabs does, but. Lucid, after bolting these things on, is obviously going to just build them in. And so now you're going to have somebody like Nvidia, a global player, who knows how to build things at scale, be able to build these and give them to any car manufacturer. Then you're going to have software plays, like Neuro and others, be able to give the software stack, and then they will be able to sell those cars to Uber, or to people who buy a flock of cars, and they'll put them into the Uber network, they'll put them into the Lyft network.

• 1:34:32Jason Calacanis?

  There's nothing that says, I couldn't buy 10 of these cars and then maintain them at my parking lot. If I had like a cafe, I take half my parking lot, I put a bunch of these cars in it, and I maintain them, I clean them, they come back, they charge at my lot, and I maintain the cars and I own them, but I put them into these networks to make money. That's really the future of this. And then you'll have Tesla, of course, doing their robotaxis, and probably a year out, I think, from taking the driver out. I think the aggressive estimates are three or four months. I put it probably at a year because of regulators. And I think they'll do it in constrained, very constrained areas, like Waymo is, so they'll have a wide area with a safety driver, and then a much smaller area, 25% of that,

• 1:35:15Jason Calacanis?

  where they know they can't screw up, or it's very low chance, because, man, a Tesla hits something, it's not going to be good. There's a lot of people who would love to write that story of a Tesla, you know, killing a puppy, God forbid. And so...

• 1:35:30Speaker 00

  But what if it's a Tesla killing a puppy while it has the anti-Elon bumper sticker on it? Ben, does that all, does that net out?

• 1:35:38Jason Calacanis?

  I mean, I think the anti-stuff is waning now. The longer you stay out of politics, I think it's like got a half-life, and so I think staying out of politics, you know, people will, the people who put those stickers on will, I think... It'll

• 1:35:53Speaker 00

  All right, um, really, really quickly here. Uh, two things before we go. One, while we're recording live today, the Fed did cut Jason 25 bips as expected.

• 1:36:02Jason Calacanis?

  Oh, wow.

• 1:36:02Speaker 00

  It's Fed Day. And then two, um, as we, uh, finish up.

• 1:36:06Jason Calacanis?

  When did that happen? When did the fuck Fed cut happen?

• 1:36:08Speaker 00

  Uh, I think about 25 minutes ago. It was at 2pm, uh, Fed meeting, uh, Fed, uh, press today. And then in Boston, right now, they are taking votes on, uh, whether or not to ban fully autonomous vehicles in the city. And I will say, Boston, I already hate you. Don't do something stupid and make me hate you marginally more.

• 1:36:30Jason Calacanis?

  I think, um, yeah, it's, um, you're going to see this happen over and over again, uh, and

• 1:36:41Jason Calacanis?

  people can do what they want in their local communities. You have to give people the ability to do what's right for their city or state, even if you don't agree with it. So, you know, I get to carry a gun on my waist here in Texas and you don't. And you live in a state where you probably don't want people carrying them. And I live in a state where people do.

• 1:37:01Speaker 00

  Uh, yeah, I, I, I'm fine with that one because I've shot a lot of handguns and.

• 1:37:06Jason Calacanis?

  Yeah. Uh, so everybody's got like a different, uh, take on this. I think this one will fall into reproductive rights, cannabis, self-driving. They're gonna just, uh, there'll be states that are very slow. I know this because they banned Airbnb in New York. They were gave Uber a really hard time in Vegas, uh, and London, and it just takes time. So it's no big deal.

• 1:37:31Speaker 00

  When are we gonna get, uh, when are we gonna get abortion and weed in Texas?

• 1:37:35Jason Calacanis?

  Um, I think they have tons of weed here. They just call it Delta nine. They have like variants. So when you drive down the road, it's like Delta nine, Delta nine. And you can get like your own variant of it. Um, yeah, it's a great question. Uh, I think. Probably not soon is what I would say.

• 1:37:55Speaker 00

  Also like, because there's bits of Texas that I like. And then I have to, then I look around and I'm like, oh God, 10 commandments in the classrooms? Come on now. And I just kind of slowly back away like a stuck cat.

• 1:38:04Jason Calacanis?

  Yeah, it's kind of like being in a time warp to like the 1980s in some ways. And then a time warp into 2050 and others.

• 1:38:13Speaker 00

  Yeah, yeah, yeah, yeah.

• 1:38:14Jason Calacanis?

  You can build your factory. You could build if you own some land and you want to build an apartment complex, they let you. And there's no NIMBY to stop you. And if you want reproductive rights, yeah, you got to get on a plane or take a bus to the state next door.

• 1:38:28Speaker 00

  Strange situation.

• 1:38:29Jason Calacanis?

  It's a very strange situation.

• 1:38:30Speaker 00

  Glory of America, 50 different states. All right. Friday, everybody, don't forget, we're going to be going through big tech earnings from the cloud perspective, trying to figure out what's going on in the AI world. So if you're curious about the state of play, we are going to have that for you here on This Week in Startups on Friday.

• 1:38:45Jason Calacanis?

  All right, we'll see you Friday. Bye bye.