Stanford MS&E435 Economics of the AI Supercycle | Spring 2026 | Infrasctructure, Enterprise AI, SaaS

Apoorv Agrawal 0:09 You know, I thought where we would start, Ali, is, you know, a lot of talk right before you joined about there's-- world's moving fast.

Ali Ghodsi 0:18 Mm-hmm.

Apoorv Agrawal 0:18 xAI, Cursor, OpenAI fighting Anthropic. Y- you know, you guys have done such a great job of stacking, you know, going from a data business to a lakehouse business to now an AI business. Just state of the union.

Ali Ghodsi 0:29 Mm-hmm.

Apoorv Agrawal 0:29 View from the top. What are you, what are you seeing? Like, frame the landscape for us. Um, what are the biggest things that you are thinking about?

Ali Ghodsi 0:36 Mm-hmm.

Apoorv Agrawal 0:36 And you know, I've got a bunch of questions that we can talk, but I thought we'd just open it up to, like-

Ali Ghodsi 0:40 Mm

Apoorv Agrawal 0:40 ... what is the biggest thing on your mind as you-

Ali Ghodsi 0:41 Hmm

Apoorv Agrawal 0:41 ... as you think about AI?

Ali Ghodsi 0:43 Yeah. I think that, uh, you know, I think you guys can chill out. Don't be stressed.

Apoorv Agrawal 0:47 [laughs]

Ali Ghodsi 0:47 You know, I think times are crazy, and, uh, I think it's not warranted, basically. And I think the stress, uh, makes people do stupid things and chase just, uh, you know, whatever happens to be the crazy thing that everybody's talking about on Twitter. Uh, I think it makes people have tunnel vision and not work on the right stuff.

Apoorv Agrawal 1:05 Yeah.

Ali Ghodsi 1:06 Uh, and I think that's what I see with, like, the current generation. Like, every year we have interns coming to Databricks, and the interns, I do always, like, a session with them-

Apoorv Agrawal 1:13 Yeah

Ali Ghodsi 1:13 ... an hour or 90 minutes or something, they can ask questions, and last two years have been just insane. Before they would ask for, like, good career advice, and you would give them good career advice.

Apoorv Agrawal 1:22 Yeah.

Ali Ghodsi 1:22 Now they're, like, 22-year-olds who are like, "Oh my God, should I, like, start, start my own company and be a CEO? Or if I, like, delay that by six months working on something, have I ruined my career and life is over, and, you know, AGI is gonna happen, and I'm gonna miss the boat?"

Apoorv Agrawal 1:34 [laughs]

Ali Ghodsi 1:35 "And, like, what am I gonna do?" I'm like-- So I'm, like, just tr- trying to tell people, like, "Calm down. Take a deep breath."

Apoorv Agrawal 1:40 Yeah.

Ali Ghodsi 1:40 Things take time.

Apoorv Agrawal 1:41 Yeah.

Ali Ghodsi 1:41 You know? So that's what I would say. I would say, actually, I think also in Silicon Valley, if you think about it right now, what's happening is, I think, uh... And you might disagree with some of this, so feel free to push back.

Apoorv Agrawal 1:51 Sure.

Ali Ghodsi 1:51 You guys can-- might disagree too. You can push back as well. Uh, but there's this quest for superintelligence, which I think is unwarranted.

Apoorv Agrawal 1:57 Yeah.

Ali Ghodsi 1:58 Because first of all, they're not even defining what superintelligence is.

Apoorv Agrawal 2:01 Mm.

Ali Ghodsi 2:02 But it's this kind of, like, god-like, you know, it's like, I think people reading Kurzweil and, you know, it's takeoff singularity, you know, this thing that comes and, like, you know, s- recursive self-improvement and, you know, cures all the diseases and GDP jumps by, like, ten percent-

Apoorv Agrawal 2:17 [laughs]

Ali Ghodsi 2:17 ... and unemployment goes to 20%, and there's no more jobs, and, you know-

Apoorv Agrawal 2:20 Fearmongering

Ali Ghodsi 2:21 ... UBI to everyone, and so on. No, I think they believe it. I think it's not needed. I think we already have AGI, so we already have artificial general intelligence.

Apoorv Agrawal 2:29 Mm-hmm.

Ali Ghodsi 2:29 Uh, you know, um, how-- Okay, this is always equally fun. How many people think we have AGI already? Okay, it's always the same. It's always, like, 10%.

Apoorv Agrawal 2:38 [laughs]

Ali Ghodsi 2:39 Okay. Uh, how many of you think that a lot of people that you interact with are not as smart as the smartest models that you use?

Apoorv Agrawal 2:45 [laughs]

Ali Ghodsi 2:48 Okay. Now let's start all over. How many of you think we don't have AGI yet? [laughing] By the way, it always works. It's like-

Apoorv Agrawal 2:57 [laughs]

Ali Ghodsi 2:57 You know? See, it's like the hypnosis is working. For some reason, they've gotten the whole world to believe we don't have AGI.

Apoorv Agrawal 3:04 Yeah.

Ali Ghodsi 3:04 But it's like you just answered it, that you have it.

Apoorv Agrawal 3:06 Right.

Ali Ghodsi 3:06 You know?

Apoorv Agrawal 3:07 Right.

Ali Ghodsi 3:07 But yet, nah, no, uh, you wanna move the goalpost. By the way, I was at the research lab in 2009-

Apoorv Agrawal 3:12 Of course, yeah

Ali Ghodsi 3:12 ... at UC Berkeley called AMPLab. Was probably the biggest, uh, most active kind of important AI lab of its time-

Apoorv Agrawal 3:19 Mm-hmm

Ali Ghodsi 3:19 ... in 2009. And, um, you know, the, you know, the god of AI was working in that lab, which is Michael Jordan. His name is actually that, so he was like the Michael Jordan of AI.

Apoorv Agrawal 3:29 [laughs]

Ali Ghodsi 3:30 Uh, and, um, back then, our definition of AGI, artificial general intelligence, um, you know, we've hit that.

Apoorv Agrawal 3:38 Yeah.

Ali Ghodsi 3:39 Like, anything we imagined would be AGI, we already hit that, and those are all the leading AI researchers in the United States. Many of them were working in that lab. But I was just-- I wanted to see, like, if I'm just full of it, so I went and asked some of those people that were there at the time, and I asked them as, "Hey, do you agree?" And they all said, "Yeah, according to that definition in 2009-"

Apoorv Agrawal 3:55 [laughs]

Ali Ghodsi 3:55 "... for sure we've hit that." But, you know, and then there's always some, you know, stupid but. We move the goalpost, or we wanna change it, or we wanna have some other definition, or it- it's-- or the AI at some-- there's some example that, you know, it couldn't count the number of Rs in strawberry or something, so therefore, we don't have AGI. Um, we already have AGI, okay? It's already smarter than many of the people that you interact with. That is general intelligence. It is artificial. It's not exactly a human. It's not the way human brain works. So we already have that. So in some sense, uh, you know, blowing a lot of money on GPUs and data centers and all of that kind of stuff is not really needed.

Apoorv Agrawal 4:28 [laughs]

Ali Ghodsi 4:29 Okay. Then there is, at the same time, so you asked for the state of the union.

Apoorv Agrawal 4:32 Yeah.

Ali Ghodsi 4:33 On the other hand, you have, like, the MIT tech report that says that 95% of the POCs are failing.

Apoorv Agrawal 4:37 Right.

Ali Ghodsi 4:38 Uh, it's kind of right directionally. I don't know if the 95% might be wrong. Maybe it's just 75%. Who knows?

Apoorv Agrawal 4:42 Interesting.

Ali Ghodsi 4:43 Uh, but if you go inside of an enterprise and-- or inside of an organization, you go into any company, and you look at how they're using stuff, the reality is that there's no, like, lots of agentic coworkers running around doing all the work, you know, blending-

Apoorv Agrawal 4:57 Yeah

Ali Ghodsi 4:57 ... with humans. That's not happening, okay? It's just humans shuffling TPS reports.

Apoorv Agrawal 5:02 [laughs]

Ali Ghodsi 5:03 Okay? It's like Office, the QA, the movie is like-- Office, the movie-- the Office Space, the movie is still, like, how the world runs.

Apoorv Agrawal 5:10 Yeah.

Ali Ghodsi 5:10 That's the reality.

Apoorv Agrawal 5:11 Yeah.

Ali Ghodsi 5:11 It's just the truth. Like-

Apoorv Agrawal 5:12 Yeah

Ali Ghodsi 5:12 ... even inside AI companies, that's how they run.

Apoorv Agrawal 5:14 [laughs]

Ali Ghodsi 5:15 It's like they, they like to think, but they're hiring salespeople from old school companies, and they're running things in old school ways, and-

Apoorv Agrawal 5:20 Yeah

Ali Ghodsi 5:20 ... I don't see, like, that futuristic thing. So then what's going on? We have AGI, but on the other hand, uh, none of this is working, and no company is using it.

Apoorv Agrawal 5:28 Yeah.

Ali Ghodsi 5:28 What the hell is going on?

Apoorv Agrawal 5:29 Yeah.

Ali Ghodsi 5:30 Uh, I think it's very simple. Um, if you don't get all the context that exists inside of these organizations and how humans work and everything, all the context we have in our heads, if you don't get that to the models and the agents, they're gonna do lots of stupid mistakes, and they're useless. And that's what's happening right now. The models just don't have, or the agents don't have the context that humans have inside of organizations, therefore they're useless. They do stupid mistakes because they don't know all the stuff that we know.

Apoorv Agrawal 5:55 Mm-hmm.

Ali Ghodsi 5:56 You know, inside of every company, there's always, like, this one guy or this one gal-

Apoorv Agrawal 5:59 Mm-hmm

Ali Ghodsi 6:00 ... who's like, "Oh, go ask John or Jane." Like, she knows everything, you know? And everybody's, like, tapping on that person's, you know-

Apoorv Agrawal 6:05 Yeah

Ali Ghodsi 6:05 ... and that's the one person you can't lose in the company. If you lose that person, the whole company collapses.

Apoorv Agrawal 6:09 Yeah.

Ali Ghodsi 6:10 That one person has-- Th- that one person exists in every department, in every company-

Apoorv Agrawal 6:13 Mm-hmm

Ali Ghodsi 6:13 ... in every organization.

Apoorv Agrawal 6:14 Mm-hmm.

Ali Ghodsi 6:14 And that one person has all the context in their head.

Apoorv Agrawal 6:17 Mm-hmm. Mm-hmm.

Ali Ghodsi 6:17 And that person, what they have in their head, is not inside of the model.

Apoorv Agrawal 6:20 Mm-hmm.

Ali Ghodsi 6:20 So therefore, the model can't operate. It just doesn't know a lot of the stuff that's sort of-- Usually, John or Jane in that company have been there for ten years, 15 years, 20 years-

Apoorv Agrawal 6:29 [coughs]

Ali Ghodsi 6:29 ... sometimes 30, 40 years. Um, you need to get that transferred to the AI. If you don't, the AI's-- doesn't matter if you get super intelligence and you can solve really difficult, you know, math questions. Um, uh, you know, and if you can get that context into the AIs, we already have AGI, and they can already crack the problem.

Apoorv Agrawal 6:47 Yeah.

Ali Ghodsi 6:47 So my, uh, urge to you guys would be, uh, you know, if you wanna have impact in the world, figure out how to get that context into the AIs, uh, inside of an or-- Like, take an organization, how do you transform how old school business is happening, and how do you get those processes into the agents? Then you will have massive impact 'cause AGI is already here.

Apoorv Agrawal 7:08 Yeah.

Ali Ghodsi 7:08 Right? That's my state of the union.

Apoorv Agrawal 7:10 AGI is already here.

Ali Ghodsi 7:11 All right.

Apoorv Agrawal 7:11 You gotta download the brain-

Ali Ghodsi 7:13 Yes

Apoorv Agrawal 7:13 ... into the silicon.

Ali Ghodsi 7:14 Yes.

Apoorv Agrawal 7:14 Get the carbon to talk to the silicon.

Ali Ghodsi 7:16 Yes.

Apoorv Agrawal 7:17 You know what? Actually, we were just talking about this-

Ali Ghodsi 7:18 And by the way, queue up your qu- your, like, pushbacks. I'm very curious to hear.

Apoorv Agrawal 7:22 Yeah.

Ali Ghodsi 7:22 I'm sure a majority disagrees. How many disagree with this? Oh, not that many. I was hop-- Okay, I'm gonna be more provocative, okay?

Apoorv Agrawal 7:29 [laughs]

Ali Ghodsi 7:30 We need, we need more pushback. Okay.

Apoorv Agrawal 7:31 So, you know, before we, before we go into AI, you know, there's a shadow of AI.

Ali Ghodsi 7:36 Uh-huh.

Apoorv Agrawal 7:36 Software is dead.

Ali Ghodsi 7:37 Mm-hmm.

Apoorv Agrawal 7:38 Software's been dead for a while. We've had this-

Ali Ghodsi 7:39 Yeah

Apoorv Agrawal 7:40 ... four times it happens. Every time it happens, it bounces back.

Ali Ghodsi 7:43 Yeah.

Apoorv Agrawal 7:44 Some macro reason, Brexit, taper tantrum, inflation. This time it's AI, and the question the class is asking is: Should we be loading up on, on software stocks? So, so is software dead? Is this a buy the dip situation?

Ali Ghodsi 7:59 Uh-huh.

Apoorv Agrawal 8:00 And, you know, I can't think of a better person to ask because depending on the day you ask, there's like software's, you know-

Ali Ghodsi 8:05 Yeah

Apoorv Agrawal 8:05 ... we, we are an AI company, software company.

Ali Ghodsi 8:07 Yeah.

Apoorv Agrawal 8:07 Speak about that. Is software dead?

Ali Ghodsi 8:08 Well, I think you know better. You're an investor. I'm not an investor. Also, I don't give financial advice. Uh, but [laughing]

Apoorv Agrawal 8:16 [laughing]

Ali Ghodsi 8:16 Having said that, if all software is dead, then isn't OpenAI and Anthropic dead? They're just software companies with a bunch of researchers-

Apoorv Agrawal 8:22 Mm. Mm

Ali Ghodsi 8:22 ... writing software. So those companies would be dead too, right? So they shouldn't have trillion-dollar valuations. SpaceX might make sense 'cause they make rockets, but-

Apoorv Agrawal 8:31 [laughs]

Ali Ghodsi 8:31 ... everybody else should be dead. Um, NVIDIA should be dead 'cause they just have really smart-

Apoorv Agrawal 8:36 Right

Ali Ghodsi 8:36 ... people who create chip designs, humans that use some software to create chip designs, and then they ship them over the internet probably over to TSMC, which is a real company creating actual chips. But then NVIDIA would be dead as well, so the world's most valuable company should be dead as well 'cause software is dead, right? So software obviously isn't dead, and it's not gonna be dead, and NVIDIA and OpenAI and Anthropic are not gonna be dead companies, uh, you know, because of whatever SaaS apocalypse or whatever we wanna call it.

Apoorv Agrawal 9:03 Yeah.

Ali Ghodsi 9:03 Um, but, um, I do think two things are true. I think that, um, two, uh, big changes have happened, which is one is barriers to entry, uh-

Apoorv Agrawal 9:14 Have gone down

Ali Ghodsi 9:14 ... have significantly gone down, and then switching costs have significantly gone down.

Apoorv Agrawal 9:18 Yeah.

Ali Ghodsi 9:18 So let's talk about those. Um, barriers to entry because it's easier than ever to write software.

Apoorv Agrawal 9:24 Mm-hmm.

Ali Ghodsi 9:25 Um, so that's like a new weapon.

Apoorv Agrawal 9:27 Yeah.

Ali Ghodsi 9:27 Anyone can produce software, uh, very cheaply-

Apoorv Agrawal 9:30 Mm-hmm

Ali Ghodsi 9:31 ... almost at zero cost.

Apoorv Agrawal 9:32 Mm-hmm.

Ali Ghodsi 9:32 It's not quite zero cost.

Apoorv Agrawal 9:33 Mm-hmm.

Ali Ghodsi 9:33 I mean, it will never be zero cost, but much, much cheaper than before.

Apoorv Agrawal 9:35 Mm-hmm. Mm-hmm.

Ali Ghodsi 9:35 But that weapon is available to everyone.

Apoorv Agrawal 9:37 Mm-hmm.

Ali Ghodsi 9:38 So also the people that create software now also have that weapon. It's not like only some, some new players have that. Everyone now-

Apoorv Agrawal 9:45 Mm-hmm

Ali Ghodsi 9:45 ... including Databricks.

Apoorv Agrawal 9:46 Right.

Ali Ghodsi 9:46 Like, we're a software company, but we also have that weapon, and it's an awesome weapon. We're using it.

Apoorv Agrawal 9:49 Have you, have you used that weapon and substituted any of your core software expenses like your CRM, your IT helpdesk, your office of the CFO software?

Ali Ghodsi 9:58 No, I think that's stupid.

Apoorv Agrawal 9:59 Yeah.

Ali Ghodsi 9:59 Uh, also, I think switching costs are lowered because it's easier to switch between UIs.

Apoorv Agrawal 10:03 Mm-hmm.

Ali Ghodsi 10:03 Again, humans get locked into software. They get-

Apoorv Agrawal 10:06 Mm-hmm.

Ali Ghodsi 10:06 You know, I don't know if you're-- Are you-- How many use Android? Oh, no one. Okay. Wow. How many use iPhone?

Apoorv Agrawal 10:12 You guys.

Ali Ghodsi 10:12 Okay. Wow. Okay. All right [laughing] . Um, you don't wanna switch to Android. Why? You don't know the UI. You don't know how to use it, right? It's like a different UI. You would have to also transfer all your data, your phone contacts, all that. It, it's too much inertia switching costs. That's a switching cost, right? But if in the future you're just talking to an agent, that switching cost gets eliminated 'cause you're just talking to an agent, so who cares if the agent is instrumenting your Android or your iPhone or your Gmail or Outlook or your Salesforce or the competitor-

Apoorv Agrawal 10:43 Yeah

Ali Ghodsi 10:43 ... or whatever it is. So that's like-

Apoorv Agrawal 10:44 Mm

Ali Ghodsi 10:44 ... the switching cost coming down as well.

Apoorv Agrawal 10:45 Mm.

Ali Ghodsi 10:46 Um, so yeah, I think it's gonna be more competition.

Apoorv Agrawal 10:48 Yeah.

Ali Ghodsi 10:49 Um, so I think software companies will have to run more efficiently.

Apoorv Agrawal 10:52 Yeah.

Ali Ghodsi 10:53 Uh, that I think is gonna happen.

Apoorv Agrawal 10:55 Yeah.

Ali Ghodsi 10:55 Um, but software is not the only moat. There is a good book you should read. It's called "The Seven Powers."

Apoorv Agrawal 11:00 Mm.

Ali Ghodsi 11:00 How many have read "The Seven Powers"? Okay, a bunch of people here. Okay. Yeah, so there's like-- there are moats that are not just software, right? I mean, like, uh, economies of scale. If you can do things at scale better than anyone else, uh, so that you can afford crazy fixed costs 'cause you're amortizing them away because of your scale, you know, Amazon, AWS, um, you know-

Apoorv Agrawal 11:21 Yeah

Ali Ghodsi 11:21 ... uh, then that's a moat.

Apoorv Agrawal 11:22 Yeah.

Ali Ghodsi 11:23 Uh, if you have a brand like Ferrari or Rolex, you know, that's a, that's a moat.

Apoorv Agrawal 11:28 Yeah.

Ali Ghodsi 11:28 People writing cheap software can't just come replace that brand. People care about that brand. Uh, trust. You know, like, "I'm the only one providing-- You can trust my company. We don't get hacked. We have, like, really secure software. We have special certification. Maybe we have patents." That remains a moat, that you cannot break that that easily.

Apoorv Agrawal 11:46 Yeah.

Ali Ghodsi 11:46 So, um, all these remain. Um-

Apoorv Agrawal 11:49 Yeah

Ali Ghodsi 11:49 ... there's a bunch of other ones, switching costs and all-

Apoorv Agrawal 11:51 Of course

Ali Ghodsi 11:51 ... and so on. So, um, data is a big moat.

Apoorv Agrawal 11:53 Mm-hmm. Mm-hmm.

Ali Ghodsi 11:53 If you have special data that no one else has, um, that only you have, that's a moat.

Apoorv Agrawal 11:59 Yeah.

Ali Ghodsi 11:59 Doesn't matter if they can write cheap software. So, so I think it's, uh, the answer is in between.

Apoorv Agrawal 12:04 Yeah.

Ali Ghodsi 12:04 The way I say it is if, if a company has, uh, been around for ten years and they have not innovated, if their software looks the same as ten years ago but the revenue's been going up-

Apoorv Agrawal 12:13 Yeah

Ali Ghodsi 12:14 ... they should be worried-

Apoorv Agrawal 12:15 Yeah

Ali Ghodsi 12:15 ... 'cause they have not been innovating, and it's probably easier for a company that starts today- To then with, you know, barriers of entry being lower, write software quickly, that's much better than that company, 'cause that company hasn't done anything for 10 years.

Apoorv Agrawal 12:27 Yeah.

Ali Ghodsi 12:28 They should be really afraid.

Apoorv Agrawal 12:29 Yeah.

Ali Ghodsi 12:29 Uh, and probably they don't even have the innovation muscle anymore 'cause they, you know, they're not innovating, so obviously they're not-- they don't have innovators. Um, those kind of companies are gonna be wiped out.

Apoorv Agrawal 12:39 Yeah.

Ali Ghodsi 12:40 But there's gonna be other companies that have been innovating the last 10 years and they're software companies, or companies that now get their shit together 'cause they're nervous.

Apoorv Agrawal 12:47 Yeah.

Ali Ghodsi 12:47 And they'll be fine too.

Apoorv Agrawal 12:49 Perfect.

Ali Ghodsi 12:49 So what do you think?

Apoorv Agrawal 12:50 I, I-

Ali Ghodsi 12:50 I mean, you're an investor.

Apoorv Agrawal 12:51 I mean, I think, um, there's a grade.

Ali Ghodsi 12:53 Yeah.

Apoorv Agrawal 12:53 Exactly as you said, like if I was to give the grades to types of software companies, I would say if you have got a lot of data, like you said-

Ali Ghodsi 13:01 Mm-hmm

Apoorv Agrawal 13:01 ... if you've got, you know, some cyber, like you're like in some core loop, you're probably s- most robust and immune from it. Somewhere in the middle is all the workflow software, like which has not-

Ali Ghodsi 13:10 Yeah

Apoorv Agrawal 13:10 ... innovated. The UX still looks same and you're like scrunching down on your shoulder and typing, "I met Ali Ghodsi today. These are the notes."

Ali Ghodsi 13:15 Yeah.

Apoorv Agrawal 13:15 Like that stuff's probably gone.

Ali Ghodsi 13:17 Yeah.

Apoorv Agrawal 13:18 If y- you, you were, exactly as you said, no innovation-

Ali Ghodsi 13:21 Yeah

Apoorv Agrawal 13:21 ... you were a part of the old habit.

Ali Ghodsi 13:24 But any one of those, they have customers, they have data. If they build great AI and start innovating, they can keep the-- keep, keep on going.

Apoorv Agrawal 13:31 Yeah.

Ali Ghodsi 13:31 They might have to change their pricing structure and their cost basis-

Apoorv Agrawal 13:34 100%

Ali Ghodsi 13:34 ... but they'll be fine.

Apoorv Agrawal 13:35 Yeah.

Ali Ghodsi 13:35 Uh, in fact, they have a lot of advantages against incumbents. They have data, they have customers, and they have some scale, so they have some economies of scale going.

Apoorv Agrawal 13:43 Yeah.

Ali Ghodsi 13:43 But they do have to get their shit together.

Apoorv Agrawal 13:45 Yeah.

Ali Ghodsi 13:45 And that's, you know, easier said than done.

Apoorv Agrawal 13:47 Yeah. Yeah, yeah. I'll, um, flash a chart. Have you guys seen this, uh, chart from Ethan Mollick? He talks about AI is very good at some things, and he called it the jagged frontier. This is customer support. Software engineering-

Ali Ghodsi 14:02 Uh-huh

Apoorv Agrawal 14:02 ... would be like-- this would be the frontier of like maybe software engineering, maybe this is customer support or whatever. But then there's a lot of stuff it's like terrible at. Like, like this scale or this scale or, and so on and, you know, you, Ali, you see a lot of-

Ali Ghodsi 14:17 We're here, man

Apoorv Agrawal 14:17 ... we have AGI already. [laughs]

Ali Ghodsi 14:19 We're over there.

Apoorv Agrawal 14:19 We're here. We're here.

Speaker 14:21 [laughs]

Ali Ghodsi 14:22 Um, they kind of admitted to it.

Apoorv Agrawal 14:23 That's right.

Ali Ghodsi 14:24 [laughs]

Apoorv Agrawal 14:24 That's right. That's right.

Ali Ghodsi 14:25 Reluctantly.

Apoorv Agrawal 14:26 Yeah.

Ali Ghodsi 14:26 Yeah. [laughs]

Apoorv Agrawal 14:27 So, you know, you, you've got what, like 6,000, 7,000 customers?

Ali Ghodsi 14:31 Mm-hmm.

Apoorv Agrawal 14:31 Those are 10,000 customers?

Ali Ghodsi 14:32 No, we have probably 20,000 customers.

Apoorv Agrawal 14:33 Sorry.

Ali Ghodsi 14:33 Plus.

Apoorv Agrawal 14:34 20,000 customers, sorry.

Ali Ghodsi 14:35 Plus.

Apoorv Agrawal 14:35 Yeah. As you see this, and you have-- that's a very good sample of the entire universe of what's happening.

Ali Ghodsi 14:40 Yep.

Apoorv Agrawal 14:40 So in that sample of 20,000 customers, what are, what are areas where AI is like hitting home runs and like working-

Ali Ghodsi 14:47 Yeah

Apoorv Agrawal 14:47 ... as advertised?

Ali Ghodsi 14:49 Yep.

Apoorv Agrawal 14:49 And w- what are areas where the frontier is still, uh, rough, and it's not working, the POCs are failing?

Ali Ghodsi 14:55 Yeah. Look, it's not AI's fault. I mean, most companies are somewhere here.

Apoorv Agrawal 14:59 Yeah.

Ali Ghodsi 14:59 I think we have AGI, but I think most companies, if you look at how much are they-- maybe they're having AI is helping me in some tasks. That's what most companies are doing. That's just how it is.

Apoorv Agrawal 15:08 Mm.

Ali Ghodsi 15:09 And, uh, it's because that context isn't there in the model.

Apoorv Agrawal 15:12 Mm.

Ali Ghodsi 15:13 So the model can't do it. Take support, which everybody said, "Okay, that's gonna be dead. Support is like gone."

Apoorv Agrawal 15:18 Yeah.

Ali Ghodsi 15:18 Right? Support is very hard.

Apoorv Agrawal 15:19 Mm.

Ali Ghodsi 15:20 Support are lit- literally the things that humans don't know what to do.

Apoorv Agrawal 15:23 Mm.

Ali Ghodsi 15:23 Like, they, they get stuck. So take Databricks.

Apoorv Agrawal 15:24 Mm.

Ali Ghodsi 15:24 Databricks offers support.

Apoorv Agrawal 15:26 Mm.

Ali Ghodsi 15:26 Databricks is a company that offers... It's a platform, advanced platform where you can do data science-

Apoorv Agrawal 15:31 Mm-hmm

Ali Ghodsi 15:31 ... machine learning. You can do advanced things on the platform.

Apoorv Agrawal 15:33 Mm-hmm.

Ali Ghodsi 15:33 These are smart people who make, you know, big salaries. They have education. You know, they, they have data science education. They're trying to use Databricks, and maybe they get stuck.

Apoorv Agrawal 15:42 Mm.

Ali Ghodsi 15:42 So their machine learning models doesn't have, you know, the right-

Apoorv Agrawal 15:45 Mm-hmm

Ali Ghodsi 15:45 ... uh, it's not getting the right F1 score or, you know, something like that.

Apoorv Agrawal 15:49 Mm.

Ali Ghodsi 15:49 And they're stuck, and they've tried everything. They call our support.

Apoorv Agrawal 15:52 Mm.

Ali Ghodsi 15:53 So it's pretty hard to automate that.

Apoorv Agrawal 15:55 Yeah.

Ali Ghodsi 15:55 You can't actually give it to a-

Apoorv Agrawal 15:56 Yeah, yeah, yeah.

Ali Ghodsi 15:56 None of the current support automation, we've tried them all.

Apoorv Agrawal 15:59 Mm.

Ali Ghodsi 15:59 Companies, all of them, immediate- even actually when they start talking to us, as soon as they know who we are, we're like, "Whoa, whoa, whoa, whoa, whoa."

Apoorv Agrawal 16:05 [laughs]

Ali Ghodsi 16:05 "We can't help you."

Speaker 16:06 [laughs]

Ali Ghodsi 16:06 Like, you go do the...

Apoorv Agrawal 16:08 [laughs]

Ali Ghodsi 16:08 "Get out of here," you know?

Apoorv Agrawal 16:09 [laughs]

Ali Ghodsi 16:09 Uh, so, uh, so yeah, most of the world is over here.

Apoorv Agrawal 16:13 Yeah.

Ali Ghodsi 16:13 But it's because we don't have the context.

Apoorv Agrawal 16:15 Right.

Ali Ghodsi 16:15 If the AI could have all the context of how our support engineers at Databricks operate-

Apoorv Agrawal 16:19 Yeah

Ali Ghodsi 16:19 ... then the AI could do it.

Apoorv Agrawal 16:21 Yeah.

Ali Ghodsi 16:21 It just doesn't have it.

Apoorv Agrawal 16:22 Yeah. You know, one of the things we used to say at Palantir is your AI strategy starts at your data strategy.

Ali Ghodsi 16:26 Yes.

Apoorv Agrawal 16:26 You gotta get the roads paved and-

Ali Ghodsi 16:28 Yep

Apoorv Agrawal 16:28 ... have the data flowing and-

Ali Ghodsi 16:30 Yep

Apoorv Agrawal 16:30 ... i- is that, you know, if you were to bucket the best enterprises who are, like, maybe, like, starting to head towards the right-

Ali Ghodsi 16:36 Uh-huh

Apoorv Agrawal 16:36 ... in your customer base of-

Ali Ghodsi 16:37 Yeah

Apoorv Agrawal 16:37 ... 20,000, what is common between the ones who are making it work-

Ali Ghodsi 16:42 Uh-huh

Apoorv Agrawal 16:42 ... and the Ferraris are flying, um, and, and, and, and ones where I'm, I'm guessing it's a context problem for the ones that it's not working, and what does it take to get the context working?

Ali Ghodsi 16:52 Yeah. It's very hard. It's a human problem. Like, it, it's not a AI problem. We already have AGI. It's a human problem. I, I don't see anyone really doing an e- excellent job at this.

Apoorv Agrawal 16:59 Mm.

Ali Ghodsi 17:00 You have to kind of rewire all your processes in the organization-

Apoorv Agrawal 17:02 Mm-hmm

Ali Ghodsi 17:02 ... uh, to, to be able to do it. This is, like, well known. I mean, my favorite is there's an article actually that I recommend people reading from 1990, uh, produced by actually a Stanford professor, uh, or researcher. Um, it's called, I think, um, you know, From the Dynamo to the Computer.

Apoorv Agrawal 17:19 Mm.

Ali Ghodsi 17:19 Okay? Check it out. So Dynamo to Computer.

Apoorv Agrawal 17:22 Okay.

Ali Ghodsi 17:22 And it looks at different, uh, sort of, uh, technological revolutions-

Apoorv Agrawal 17:26 Mm-hmm

Ali Ghodsi 17:26 ... and how long it ha- how long it took for them to have impact on productivity of econo-

Apoorv Agrawal 17:30 Mm-hmm

Ali Ghodsi 17:30 ... the, the economy. And it's just, you know, takes just forever.

Apoorv Agrawal 17:35 Yeah.

Ali Ghodsi 17:35 Like, when the PCs came out, the joke was the Nobel laureate economist, uh-

Apoorv Agrawal 17:38 Yeah

Ali Ghodsi 17:39 ... um, you know, uh, um, uh, Richard Tol said that-

Apoorv Agrawal 17:44 Mm

Ali Ghodsi 17:44 ... computers or PCs, you can find them everywhere-

Apoorv Agrawal 17:47 Mm

Ali Ghodsi 17:47 ... except in, uh, the productivity statistics.

Apoorv Agrawal 17:50 Mm.

Ali Ghodsi 17:51 You know?

Apoorv Agrawal 17:51 [laughs]

Ali Ghodsi 17:51 Like, it, it's like it just doesn't show up in the statistics.

Apoorv Agrawal 17:54 Mm.

Ali Ghodsi 17:54 Um, why? People were buying PCs, and they were using them as typewriters.

Apoorv Agrawal 18:00 Right.

Ali Ghodsi 18:00 So they would have people type on PCs but then print out the sheets and then put them in folders, and then have assistants that, like, index them-

Apoorv Agrawal 18:08 Yeah

Ali Ghodsi 18:08 ... and do things.

Apoorv Agrawal 18:09 Yeah, yeah.

Ali Ghodsi 18:09 So, like, you didn't see any productivity gains from that.

Apoorv Agrawal 18:11 Mm.

Ali Ghodsi 18:11 And same thing with, uh, if you look at the Industrial Revolution, same thing happened. Um, you know, we had these steam engines, and the steam factories were sort of super dense, and they were ru- running, like, with these, you know, they were called the line-

Apoorv Agrawal 18:25 Mm

Ali Ghodsi 18:25 ... uh, shafts-

Apoorv Agrawal 18:27 Mm-hmm

Ali Ghodsi 18:27 ... which were, like, these things that rotate.

Apoorv Agrawal 18:29 Mm-hmm.

Ali Ghodsi 18:29 When the electric engine came, that's the dynamo-

Apoorv Agrawal 18:31 Mm

Ali Ghodsi 18:31 ... uh, it took 40 years before they saw any productivity gains in the, in the economy. Wow Yeah. Th- check it out. Wow. This is in that article. It took from 1880 to- The diffusion took 40 years ... from ni- 1880 to 1920 when the electric engine came- Mm ... uh, to see impact. So what they were doing is they were going to these factories that already were these line shaft factories- Mm-hmm ... that were these dense factories, where you have a steam engine that's rotating this line shaft, and it's rotating these belts, and then everything is working. You have these multiple stories. Um, and all they did is just like the PC- Mm ... they used as a typewriter. They would replace the steam engine with an electric engine. Mm. And that doesn't, just like replacing the PC- Yeah, yeah ... with a, you don't get any, um, productivity gains. Yeah. It took till 1920- Mm ... but maybe it was 1915, but I'm roughly right- Yeah ... um, until they realized, wait, we have to change the whole factory floor. Mm. We have to move the factories out of the cities. We have to have, like, floor plans that are- Mm ... much bigger 'cause now we can distribute the electricity. Mm. Electricity is much more, it doesn't, you know, it's not like the, um, the torque that has, you know, inefficiency. Right. Uh, we can spread it out. We can have floor plans that are big, and we can run different parts of the factory at different rates. Mm. Unit drive versus group drive. Um, took a very long time. Yeah. That's what's gonna happen. Same, same thing now. Right. Rewiring, I know it because I have 20,000 customers, and I talked to them. I was late to this meeting 'cause I was meeting one of the CEOs of one of the big banks. Yeah. And same problem. He has the same problem. All the organizations I work with- Yeah ... have the same problem. They're like, "I'm not seeing any advan- Like, I don't see..." They're all like, "AI is amazing. It's coming." It's like, "I need it. I need to do that," but they're like, "I don't see any productivity gains- [laughs] ... in my organization. You know, what the hell am I doing wrong?" Yeah. And I tell them we have AGI, and they're like, "What?" Yeah. [laughs] Like, "That is not true. Like, we don't see anything," is what they're saying. Well, it's, it's a very tough problem 'cause you're like, "Hey, I got the brain, but I got to rebuild the human body." Yeah. The hands, the legs. Yeah. That was previously a different body. Let me give you, let's, let me give you... Yeah, let me give you an example from Databricks. So Databricks helps you get data from all the different systems- Mm ... like Salesforce, Workday, and so on. Mm. Collect them in one place. Mm. Secure it, and then do AI on

Ali Ghodsi 20:33 it. Mm. Like, you can do predictions. You can build predictive models. That's what Databricks is. Mm. So we built connectors to all these systems. Mm. These connectors are... It would take us three quarters to build a production connector. We're good at it. Mm. This is what we do for a living. Yeah. We build these connectors. Like, we can build a connector from Databricks to Salesforce, production ready. Mm. It would take us three quarters, so nine months- Mm ... to do that. Shipped, secure- Nice ... with so on. That's like, that's what we did. So, you know, w- as, uh, you know, uh, the LLMs got faster and faster and faster, I started sort of experimenting with this myself, and I was like, "Oh, I could write a connector in two days." So I went to the team that builds this- [laughs] ... and, um, and I was like, "Hey, I can do this in two days. How come it takes you guys three quarters?" They're like, "Okay, great point. Let us come back to you." So they, they went, and they thought about it, and they came back in two weeks, and they said, "Okay, you're right. Uh, it's, but you're also not right. We looked at it, and yeah, this AI is useful. We can compress it down from three quarters by one and a half month. So we can get it from nine months to seven and a half months." That's it. That's it. I'm like, "Well, I can do it in two days." [laughs] They're like, "No, no, no. No offense to you, but, you know, this is production code, and it really actually works." [laughs] "And, you know, we have, like, customer feedback, and, you know, it's, like, secure, and, you know, you wrote some toy." Yeah. God knows what that means. Yeah, yeah. No offense. You're great, but- Yeah, yeah ... you know, let us, let, let us do it now." What's the missing link? Uh, so I was like, "Oh, man, this is kind of depressing, but yeah, I'll take the one and a half month improvement and, you know, maybe it's something, but maybe I'm just stupid, and I don't get it." Yeah. Then I found another guy in the company. We went to him, and we sort of said, "Hey, can you look at this problem?" [laughs] And he's very first principle. He's a very smart guy, and he doesn't care about all this, like, you know, fluff. He's like, he cuts through the fluff, and he cut through the fluff, and he worked with a team, and they came back, and they said, "Hey, after looking at the problem, we can do seven connectors in one quarter." Boom. Yeah. Let's go. Yeah. What is the difference? So what's the difference? Okay, so what he did is he, he went from first principles with some team members, and they looked at it, and they said, "Okay, first quarter, they're just sending w- our very expensive, very smart, Stanford educated product managers out to the customers to talk to the customers and collect feedback. What exactly is your requirements?" Mm. "How do you use Salesforce?" And so on. That takes a full quarter. At the end of that quarter, our amazing, smart, uh, product managers come back with, like, a 60, 70, 80 page super nice report on exactly all the requirements. Mm-hmm. Okay, so you're blocked for a whole quarter. So for sure you can't- Yeah ... Amdahl's law. Yeah. You can't compress it below, below that. Mm. Then code writing starts, but we have to test this stuff. So testing requires you to set up Salesforce, Workday, NetSuite, but those are not software by Databricks, so we're not very good at that. That takes a very long time, and it's hard to find people to do that at Databricks. So that, a- again, is, like, a process that takes a long time for us to stand up, and it's very error prone. So we couldn't do that either. Um, and then we have one person for each connector. Mm. They go on vacation. They get sick. [laughs] You know, so on, so all of that. So what he did is he just from first principles looked at it and said, "We're gonna just rewire all of this." Mm-hmm. And lots of people didn't like this. They were unhappy about it, but he said that, um, you know, the product requirements, instead of one quarter, we're just gonna take one week and quickly write down whatever we have. Mm. We might get things wrong, but because the software is so fast to write, we can rewrite it again. Right. So let's iterate faster. Uh, the standing up the Salesforce instances, let's outsource that to firms that can do that for us. Mm. And we can just pay them a lot, and they do it in parallel. Mm. So we can shrink that as well. Mm, mm. And then one person per connector, let's change it. Let's have seven people, seven connectors- Mm ... and then they all work on all the connectors together, so we don't have what's called, you know, bus factor one. Mm. If someone is hit by a bus- Mm ... the whole project does not stop. Right. Right? Uh, so, um, so yeah, so got it- Nice ... all done into one quarter and seven, seven connectors shipped and, you know, so but this had nothing to do with- Yeah. Yeah, yeah, yeah ... uh... Like, really it didn't have anything to do with the- Yeah ... AI or AGI or smarter models- Yeah, yeah ... or super intelligence or gi- gigantic... Like, you can have the, the next, like, GPT-7 or Opus 6 would not have helped us- Yeah ... um, do this better. We needed to do those, make those changes. Yeah. And that's, like, a human refactoring problem- Yeah, yeah ... and process change. Yeah. And, um, so this is what the whole world is going through, so- 100% ... th- that's what you need to do well if you wanna, if you wanna succeed. Some are doing it better. Others are not.

Apoorv Agrawal 24:46 H- Hamilton Helmer actually talks about this-

Ali Ghodsi 24:48 Mm-hmm

Apoorv Agrawal 24:48 ... uh, quite a bit actually. So for all of you who are picking assignment option one and wanna be investors, Hamilton Helmer's, uh, book is a must read on process power. We were debating this, um, before this. If, Ali, you had, uh, $100 to invest across what Jensen calls the five-layer stack: energy, chips, infra, model, and apps, where does value accrue? If you were to put 100 bucks in the, in, in the index of energy and chips and infra and, and so on, uh, with let's say a long-term timeframe, where does... where, where would you put it? How would you allocate the $100, um, and why?

Ali Ghodsi 25:26 I'm, I'm a computer scientist. I'm not an investor.

Apoorv Agrawal 25:28 [laughs]

Ali Ghodsi 25:28 I don't give financial advice.

Apoorv Agrawal 25:30 Yeah.

Ali Ghodsi 25:30 But-

Apoorv Agrawal 25:30 But you're allocating-

Speaker 25:31 [laughs]

Apoorv Agrawal 25:32 You know, you're, you're-

Ali Ghodsi 25:32 If I have $100-

Apoorv Agrawal 25:33 Yeah, you are allocating-

Ali Ghodsi 25:34 ... it's house money

Apoorv Agrawal 25:35 ... Databricks time, right?

Ali Ghodsi 25:36 Yeah.

Apoorv Agrawal 25:36 Databricks is across three of these.

Ali Ghodsi 25:37 Yeah, yeah.

Apoorv Agrawal 25:38 You got any-

Ali Ghodsi 25:38 Um, I would just say, look, it's obvious that the applications are gonna be the winners, right?

Apoorv Agrawal 25:42 Yeah.

Ali Ghodsi 25:42 So I would put, put it in the top. Uh, it's kinda like, uh... and I'll give you some, some guesses at, you know, but who knows actually. It's very hard to predict.

Apoorv Agrawal 25:50 Yeah.

Ali Ghodsi 25:50 So you would have to kind of have a... I would go early stage, and I would have a seed strategy, and I would invest in many, many startups, and I would get most of them wrong, but a few would actually make it, and they would be the next Google or whatever. Um, but you know, in 1990... like when I did my PhD, um, in the early 2000-

Apoorv Agrawal 26:10 Mm-hmm

Ali Ghodsi 26:11 ... um, I was in the networking field. Networking was like the cool thing to do. It was-

Apoorv Agrawal 26:15 Mm

Ali Ghodsi 26:15 ... the advanced thing 'cause the internet was, like you wanna work on it. Like the internet was the big thing at the time.

Apoorv Agrawal 26:20 Yeah.

Ali Ghodsi 26:20 And you wanna... and the cor- coolest thing on the internet was, uh, networking.

Apoorv Agrawal 26:25 Yeah.

Ali Ghodsi 26:25 And the hardest problem, like the smartest math brains were working on at the time, we all knew what the future would look like.

Apoorv Agrawal 26:32 Mm.

Ali Ghodsi 26:32 The future, everybody knew what the most important problem everyone's gonna work on is the c- what's called the multicast problem-

Apoorv Agrawal 26:37 Mm. What is that?

Ali Ghodsi 26:38 ... which is... Yeah, see. [laughs]

Speaker 26:40 [laughs]

Ali Ghodsi 26:40 It's problematic that no one knows what that is today. We were, we were clearly wrong.

Apoorv Agrawal 26:44 [laughs]

Ali Ghodsi 26:45 So multicast is, you know, you wanna broadcast from one source, let's say a soccer game or a football game or basketball game-

Apoorv Agrawal 26:51 Right

Ali Ghodsi 26:51 ... to the whole world-

Apoorv Agrawal 26:52 Mm

Ali Ghodsi 26:52 ... 'cause everybody wants to watch it at the same time.

Apoorv Agrawal 26:54 Mm-hmm.

Ali Ghodsi 26:54 We didn't know how to solve that-

Apoorv Agrawal 26:55 Mm

Ali Ghodsi 26:55 ... efficiently. So all of the smartest brains-

Apoorv Agrawal 26:57 The scale, yeah

Ali Ghodsi 26:57 ... in the world were trying to work on this problem.

Apoorv Agrawal 26:59 Mm.

Ali Ghodsi 26:59 And bandwidth was scarce.

Apoorv Agrawal 27:00 Mm, mm.

Ali Ghodsi 27:01 While we were doing this... and by the way, we actually, you know, had pretty good problems, and I started the company on this.

Apoorv Agrawal 27:06 Mm.

Ali Ghodsi 27:06 Uh, and we had a great solution.

Apoorv Agrawal 27:09 Mm.

Ali Ghodsi 27:09 Unfortunately, the cost of bandwidth just plummeted, and they just-

Apoorv Agrawal 27:13 Yeah

Ali Ghodsi 27:13 ... deployed so much fiber that no one... this problem was not a problem ever, so no one needed to buy this software.

Apoorv Agrawal 27:18 Mm, mm.

Ali Ghodsi 27:18 So it was a complete waste of time. Uh, and at that time, we thought the hardest problems, the most interesting things to work on are Cisco routers, routing-

Apoorv Agrawal 27:25 Mm

Ali Ghodsi 27:26 ... BGP, border gateway protocol-

Apoorv Agrawal 27:27 Mm

Ali Ghodsi 27:27 ... internet protocol, like, you know-

Apoorv Agrawal 27:29 Mm

Ali Ghodsi 27:29 ... queuing theory, quality of service-

Apoorv Agrawal 27:31 Mm

Ali Ghodsi 27:31 ... these kind of things. Those, those are like the most interesting things you can... 'cause we had tunnel vision on the internet.

Apoorv Agrawal 27:36 Mm.

Ali Ghodsi 27:36 And the... what is the internet? Well, at the time, it was the internet protocols and those things.

Apoorv Agrawal 27:40 Right.

Ali Ghodsi 27:41 No apps really existed.

Apoorv Agrawal 27:42 Right.

Ali Ghodsi 27:42 So we were all focused on that, and today everybody's focused, I would say, on-

Apoorv Agrawal 27:46 Chips

Ali Ghodsi 27:46 ... I think, like, you know, I think chips and, you know, I think infrastructure. I think people are really right now the hot new thing is like NVIDIA, OpenAI, Anthropic, DeepMind.

Apoorv Agrawal 27:56 Yeah.

Ali Ghodsi 27:56 These are the things everybody's focused on.

Apoorv Agrawal 27:58 Yeah.

Ali Ghodsi 27:58 AGI, super intelligence, the, what I said at the beginning.

Apoorv Agrawal 28:01 Yeah.

Ali Ghodsi 28:01 But, uh, on the internet, there were like really weird things-

Apoorv Agrawal 28:05 Yeah

Ali Ghodsi 28:05 ... that took off. The really weird things that took off were like taxi business, you know, which is Uber.

Apoorv Agrawal 28:12 Uber, yeah.

Ali Ghodsi 28:12 Uh, or selling books-

Apoorv Agrawal 28:13 [laughs]

Ali Ghodsi 28:15 ... which is the lamest thing ever, but that became Amazon-

Apoorv Agrawal 28:17 [laughs]

Ali Ghodsi 28:18 ... which became AWS.

Apoorv Agrawal 28:19 Yeah.

Ali Ghodsi 28:20 You know, uh, or, uh, renting your bedroom to people, like-

Apoorv Agrawal 28:24 [laughs]

Ali Ghodsi 28:24 ... you know, that's Airbnb. Uh, and or, um, sending people short text-

Apoorv Agrawal 28:31 Right

Ali Ghodsi 28:31 ... which became Twitter.

Apoorv Agrawal 28:32 Right.

Ali Ghodsi 28:32 You know?

Apoorv Agrawal 28:33 Right.

Ali Ghodsi 28:33 Uh, these are like... and if you said them in those words in, in 2000 to people, people would say, "Uh, you're out of your mind."

Apoorv Agrawal 28:38 Right.

Ali Ghodsi 28:39 Like, "You're insane."

Apoorv Agrawal 28:39 Right.

Ali Ghodsi 28:39 "You're full of it." Uh, but that's, those were the great ideas of the time.

Apoorv Agrawal 28:42 Mm.

Ali Ghodsi 28:42 Those are the ones that became... So I think it's the same thing here.

Apoorv Agrawal 28:44 Right.

Ali Ghodsi 28:45 Uh, to throw a few of them out there-

Apoorv Agrawal 28:47 Mm

Ali Ghodsi 28:47 ... um, I think healthcare is like 17% of US GDP.

Apoorv Agrawal 28:51 Yeah.

Ali Ghodsi 28:52 Um, you know, we, we, we all still unfortunately will die, and we all care about our health and the health of our loved ones. I think there's huge... we have, you know, uh, the propensity to pay for this.

Apoorv Agrawal 29:06 Mm.

Ali Ghodsi 29:06 Like we'd pay anything-

Apoorv Agrawal 29:07 Mm

Ali Ghodsi 29:07 ... to be able to save lives or our, of our loved ones or our own lives or our own health issues. Uh, and it's not particularly well done today.

Apoorv Agrawal 29:15 Yeah.

Ali Ghodsi 29:15 And it's surprise, surprise.

Apoorv Agrawal 29:16 Yeah.

Ali Ghodsi 29:17 You know, healthcare is not, like, awesome. Um, so imagine a company that has seen a billion patients.

Apoorv Agrawal 29:22 Yeah.

Ali Ghodsi 29:22 Like I have seen 100 million patients with your kind of genetic composition-

Apoorv Agrawal 29:26 Mm

Ali Ghodsi 29:27 ... and the kind of issues that you might have in the future, and I can help you.

Apoorv Agrawal 29:30 Mm.

Ali Ghodsi 29:30 But what are you willing to pay-

Apoorv Agrawal 29:31 Mm

Ali Ghodsi 29:32 ... for me to help you with that?

Apoorv Agrawal 29:33 Mm.

Ali Ghodsi 29:33 That could be a company that's trillions of dollars' worth.

Apoorv Agrawal 29:35 Right.

Ali Ghodsi 29:36 Um, to take something out of left field that I think people think is really, you know, uh, not interesting enough, but take education.

Apoorv Agrawal 29:44 Mm.

Ali Ghodsi 29:44 Education actually in, in VC space, peop- the consensus has always been education's like a terrible investment, right?

Apoorv Agrawal 29:50 Mm.

Ali Ghodsi 29:50 Isn't it? Like we see people say always like-

Apoorv Agrawal 29:51 Historically

Ali Ghodsi 29:51 ... "Never invest in education."

Apoorv Agrawal 29:53 W- w- what's the last public market company you know-

Ali Ghodsi 29:55 Yeah

Apoorv Agrawal 29:55 ... public company you know-

Ali Ghodsi 29:56 Yeah. What's the last trillion-dollar education company?

Apoorv Agrawal 29:58 Not even 100 billion.

Ali Ghodsi 29:58 Yeah.

Apoorv Agrawal 29:59 Yeah.

Ali Ghodsi 29:59 Yeah, anything, right? Um, so but most people have kids.

Apoorv Agrawal 30:04 Yeah.

Ali Ghodsi 30:04 And you know, more kids are produced, and they do, they do need to go through, uh, uh, get, to get an education, whether people believe it or not.

Apoorv Agrawal 30:12 Yeah, yeah.

Ali Ghodsi 30:13 And, um, uh, and people do care actually-

Apoorv Agrawal 30:16 Yeah

Ali Ghodsi 30:16 ... if the education for their kids are good or not.

Apoorv Agrawal 30:18 Yeah.

Ali Ghodsi 30:18 Elections are won and lost.

Apoorv Agrawal 30:19 Yeah.

Ali Ghodsi 30:19 There's cultural issues on these things like, you know, of what you c- you're allowed to teach my kids or not, right?

Apoorv Agrawal 30:25 Mm-hmm.

Ali Ghodsi 30:25 Elections are won and lost on that.

Apoorv Agrawal 30:26 Mm-hmm. Mm.

Ali Ghodsi 30:26 Not because it's a stupid topic, because it matters.

Apoorv Agrawal 30:28 Yeah.

Ali Ghodsi 30:28 Like what are you teaching my kids matters.

Apoorv Agrawal 30:30 Yeah.

Ali Ghodsi 30:30 And are my kids being brainwashed to do the right thing or the wrong thing, or are they gonna be... do their... are they, you know, well-equipped to get the jobs of the future?

Apoorv Agrawal 30:38 Yeah.

Ali Ghodsi 30:38 Uh, I think if, if there's a company that can provide amazing education-

Apoorv Agrawal 30:42 Right

Ali Ghodsi 30:42 ... uh, using AI, um, I think-

Apoorv Agrawal 30:45 Right

Ali Ghodsi 30:45 ... people, a lot of people will pay for that.

Apoorv Agrawal 30:47 Right.

Ali Ghodsi 30:47 And if it's like proven that that does a better job than, um, than-

Apoorv Agrawal 30:52 Yeah

Ali Ghodsi 30:52 ... than, you know, whatever-

Apoorv Agrawal 30:53 Yeah

Ali Ghodsi 30:53 ... they're getting right now. Um, just two flavors of like obvious companies that I think-

Apoorv Agrawal 30:57 Yeah

Ali Ghodsi 30:57 ... will exist, and they could be trillion-dollar companies.

Apoorv Agrawal 30:59 Yeah.

Ali Ghodsi 31:00 If they do it well, they will have data moat.

Apoorv Agrawal 31:02 Yeah.

Ali Ghodsi 31:03 Um, they will have economies of scale modes

Apoorv Agrawal 31:05 Mm.

Ali Ghodsi 31:05 There's like winner takes it all kind of-

Apoorv Agrawal 31:07 Mm

Ali Ghodsi 31:07 ... uh, dynamics-

Apoorv Agrawal 31:08 Mm. Mm

Ali Ghodsi 31:08 ... in those markets. At least in countries, in geos.

Apoorv Agrawal 31:12 Yeah.

Ali Ghodsi 31:12 So, uh, so I think the value accrues to the top.

Apoorv Agrawal 31:15 Yeah. We can't wait for that to happen.

Ali Ghodsi 31:16 But I'm not an investor.

Apoorv Agrawal 31:17 Yeah, we can't wait for that to happen. Um-

Ali Ghodsi 31:19 Will you push back?

Apoorv Agrawal 31:20 No, I think, I mean, look, I've, I've written extensively about this.

Ali Ghodsi 31:22 Mm-hmm.

Apoorv Agrawal 31:22 Uh, eagerly-

Ali Ghodsi 31:23 Mm

Apoorv Agrawal 31:23 ... waiting for this, uh, what I call the blue triangle to, uh, to invert. Uh, I don't know if you've seen this, but basically this is, uh, all the money in AI-

Ali Ghodsi 31:32 Yeah

Apoorv Agrawal 31:32 ... is with one guy.

Ali Ghodsi 31:34 Yeah.

Apoorv Agrawal 31:34 That's why Jensen's so happy all the time, as you, as, as you know. [laughs]

Ali Ghodsi 31:37 Yeah.

Apoorv Agrawal 31:38 Um, [laughs] these guys are fighting for-

Ali Ghodsi 31:42 Yeah

Apoorv Agrawal 31:42 ... dollars. There's like no money there.

Ali Ghodsi 31:43 Yeah.

Apoorv Agrawal 31:44 There, there's very little money here. I mean, the people are making some money here and, uh, so we'll see. But it- but that's the bet. The bet is that this thing will look like a more sustainable-

Ali Ghodsi 31:51 Yeah, it will go that way. I mean, it's-

Apoorv Agrawal 31:53 100%

Ali Ghodsi 31:53 ... all value in Silicon Valley and in tech and in technology moves up the stack all the time.

Apoorv Agrawal 31:58 Yeah.

Ali Ghodsi 31:59 Like, and even look at the greatest companies, like, okay, the company that created the PCs, IBM, was, like, the greatest market cap, and all the value-

Apoorv Agrawal 32:05 Yeah

Ali Ghodsi 32:05 ... accrued there. But then that became commoditized.

Apoorv Agrawal 32:07 Yeah.

Ali Ghodsi 32:07 Then it became the software on top of it-

Apoorv Agrawal 32:08 [laughs]

Ali Ghodsi 32:08 ... which is, like, the operating systems and the Microsofts-

Apoorv Agrawal 32:11 Yeah, yeah

Ali Ghodsi 32:11 ... of the world and so on. Then, you know, here at Stanford actually a while back, it was, like, 20 years ago, VMware, which is how do you virtualize that software?

Apoorv Agrawal 32:18 Yeah.

Ali Ghodsi 32:18 And but that became commoditized. And then, like, you know, so it-

Apoorv Agrawal 32:20 Moving up

Ali Ghodsi 32:20 ... keeps moving up the stack all the time.

Apoorv Agrawal 32:22 Yeah.

Ali Ghodsi 32:23 That's how, that's, that's how it's gonna be here too.

Apoorv Agrawal 32:25 100%. And, you know, the big ... One of the forces that is commoditizing this, you've spoken about this, is open source.

Ali Ghodsi 32:29 Mm-hmm.

Apoorv Agrawal 32:30 Open source is, uh, getting pretty good.

Ali Ghodsi 32:32 Yep.

Apoorv Agrawal 32:32 Th- this blue line is open source.

Ali Ghodsi 32:34 Yeah.

Apoorv Agrawal 32:35 The, the gap is closing. This is, like, what, three, three, four months?

Ali Ghodsi 32:38 Yeah.

Apoorv Agrawal 32:38 This gap is now, like, a month.

Ali Ghodsi 32:40 Yeah.

Apoorv Agrawal 32:41 And but still, people are spending so much money on these frontier models.

Ali Ghodsi 32:46 Yeah.

Apoorv Agrawal 32:46 People cannot wait to get their hands on 47-

Ali Ghodsi 32:49 Yeah

Apoorv Agrawal 32:49 ... from Open- from, from Claude or, or 55 from GPT.

Ali Ghodsi 32:52 Yeah.

Apoorv Agrawal 32:52 But then there's this whole economy of, of very good open source models.

Ali Ghodsi 32:57 Yeah.

Apoorv Agrawal 32:58 What, what do you make of all this? Like, you ... On one side you've got people earning, what, 30 billion now-

Ali Ghodsi 33:02 Yeah

Apoorv Agrawal 33:02 ... or, or maybe 40 billion at Anthropic or, or ... But on the other side, this open source stuff is, like, nearly free. Obviously you've gotta pay the hosting.

Ali Ghodsi 33:09 Yeah.

Apoorv Agrawal 33:09 How do you think this shakes out? Will, will that model layer, the, the proprietary model layer accrue any value?

Ali Ghodsi 33:16 No, I think it's gonna be valuable.

Apoorv Agrawal 33:17 Yeah.

Ali Ghodsi 33:17 And I think people will want it, whether it's open source or not, let's put that aside for a second. I think there will be token factories-

Apoorv Agrawal 33:22 Hmm

Ali Ghodsi 33:22 ... which serve this stuff up. It's just like the cloud.

Apoorv Agrawal 33:25 Right.

Ali Ghodsi 33:25 I don't think-

Apoorv Agrawal 33:26 The inference, yeah.

Ali Ghodsi 33:26 Like, I think we foolish to say you all will have your own little mini data center in your living rooms, and you're gonna run, you know, your own PCs, and you're gonna-

Apoorv Agrawal 33:34 Yeah

Ali Ghodsi 33:34 ... insert GPU cards that you buy at home-

Apoorv Agrawal 33:36 [laughs]

Ali Ghodsi 33:36 ... and you're gonna run this yourselves.

Apoorv Agrawal 33:37 On your phone or MacBook-

Ali Ghodsi 33:39 Yeah

Apoorv Agrawal 33:39 ... or the edge inference.

Ali Ghodsi 33:39 Some of that might, will exist. It will come to the edges.

Apoorv Agrawal 33:41 Yeah.

Ali Ghodsi 33:41 But I do think there'll be, like, big-

Apoorv Agrawal 33:43 Yeah

Ali Ghodsi 33:43 ... centralized data centers where this happens. But we haven't discussed are they running open source models, are they running proprietary models? And, um, and here's a fun fact. So Moonshot, the Chinese company, released Qimi-

Apoorv Agrawal 33:56 Yeah

Ali Ghodsi 33:56 ... uh, 2.6.

Apoorv Agrawal 33:57 Very good model. [laughs]

Ali Ghodsi 33:58 Two day- two days ago or three days ago.

Apoorv Agrawal 33:59 Yeah. Yeah, Tuesday.

Ali Ghodsi 34:00 Uh, yeah, Tuesday. Uh, so two days ago they released, uh, three days, two days ago, they released 2.6. In January, they released 2.5.

Apoorv Agrawal 34:07 Hmm.

Ali Ghodsi 34:08 And here's a fun fact. 2.6 that they released on Tuesday is the best model ever in the history of mankind ever produced-

Apoorv Agrawal 34:14 [laughs]

Ali Ghodsi 34:14 ... frontier, non-frontier-

Apoorv Agrawal 34:16 Yeah

Ali Ghodsi 34:16 ... if it just had been released in January.

Apoorv Agrawal 34:18 Yeah.

Ali Ghodsi 34:18 The open source will be here.

Apoorv Agrawal 34:20 Yeah.

Ali Ghodsi 34:20 And it will apply pricing pressure, and this business of frontier models-

Apoorv Agrawal 34:25 Yeah

Ali Ghodsi 34:25 ... that core business of providing frontier models is gonna be economies of scale game.

Apoorv Agrawal 34:30 Mm.

Ali Ghodsi 34:30 And you'll have to, uh, do it at small margins. It's like an Amazon.com book selling business.

Apoorv Agrawal 34:36 Yeah. Yeah.

Ali Ghodsi 34:37 That's what it's gonna look like in the future. Therefore, there are not gonna be that many people doing it.

Apoorv Agrawal 34:40 Yeah.

Ali Ghodsi 34:41 Uh, it's just like an Amazon.com, and gross margins are gonna be tiny.

Apoorv Agrawal 34:45 Yeah.

Ali Ghodsi 34:45 And operating margins are gonna be small. That's my-

Apoorv Agrawal 34:48 Yeah

Ali Ghodsi 34:48 ... take.

Apoorv Agrawal 34:49 Yeah, yeah. I think so too. Um, three rapid fire questions before we, uh, wrap.

Ali Ghodsi 34:54 Mm-hmm.

Apoorv Agrawal 34:55 Your, uh, favorite AI product that you use every day.

Ali Ghodsi 34:58 Oh, I don't know. That's a tough one. Uh-

Apoorv Agrawal 35:00 [laughs]

Ali Ghodsi 35:01 Um, I mean, I use all of these.

Apoorv Agrawal 35:02 Yeah.

Ali Ghodsi 35:03 Uh, you know.

Apoorv Agrawal 35:04 Chat.

Ali Ghodsi 35:04 I actually like Cursor. I know that's like everybody loves Cloud Code.

Apoorv Agrawal 35:06 Yeah.

Ali Ghodsi 35:06 I like the diffs and how, how, how it works. So, like, on coding AI, I use, like, a combo of those.

Apoorv Agrawal 35:11 Nice.

Ali Ghodsi 35:11 I still kinda like it, yeah.

Apoorv Agrawal 35:12 Are you still using it after, uh, Elon owns it?

Ali Ghodsi 35:15 No, I stopped. No, of course. Yeah.

Apoorv Agrawal 35:16 [laughs]

Ali Ghodsi 35:17 [laughs] Yeah.

Apoorv Agrawal 35:18 'Cause you're gonna lose access to Anthropic and OpenAI tokens through Cursor, I presume.

Ali Ghodsi 35:23 Yeah. Yeah, man. Awesome. It's great. Yeah.

Apoorv Agrawal 35:25 Good. Good. Good. Good. Supporter.

Ali Ghodsi 35:27 Yeah.

Apoorv Agrawal 35:28 Um, you've been in this-

Ali Ghodsi 35:29 The, the truth is I do use Databricks' Genie products. Th- this is the truth.

Apoorv Agrawal 35:32 Yeah. Yeah.

Ali Ghodsi 35:32 'Cause, you know, it just ... Most of my ... Inside Databricks, most of my decisions are, like, numerical and quantitative in nature.

Apoorv Agrawal 35:38 Yeah.

Ali Ghodsi 35:38 Like, should we do this? What's the ROI on this? What's the cost on that? What it's gonna cost us? What's the ... So I need something that can understand numerical data and time series data.

Apoorv Agrawal 35:47 Yeah.

Ali Ghodsi 35:47 So Genie is, like, really good for that.

Apoorv Agrawal 35:49 Right. Right.

Ali Ghodsi 35:49 So that's, that's what I honestly go to quite a bit.

Apoorv Agrawal 35:51 Right. Right.

Ali Ghodsi 35:51 Quite often.

Apoorv Agrawal 35:52 Right. Um, future for Databricks. You've been at this for, for 15 years or so. Um, what is your vision for the next decade for Databricks?

Ali Ghodsi 36:02 Well, I think the cost of software is going down.

Apoorv Agrawal 36:04 Yeah.

Ali Ghodsi 36:04 And so barriers to entry and switching costs are going down, so there is an, a SaaS apocalypse of sorts, but not all software is gonna be dead.

Apoorv Agrawal 36:12 Yeah.

Ali Ghodsi 36:13 Uh, we would love to partake in that then.

Apoorv Agrawal 36:15 Right.

Ali Ghodsi 36:15 Kill some software.

Apoorv Agrawal 36:17 Right.

Ali Ghodsi 36:17 Yeah.

Apoorv Agrawal 36:17 Right. Right. Any advice for, uh, students in the room who are about to, uh, make career decisions?

Ali Ghodsi 36:24 Yeah, I think don't, don't be worried about the fear-mongering. Don't be stressed out. Take it easy. Um, I think that, uh, tho- that's g- those ... I was very stressed doing my PhD in the early 2000s.

Apoorv Agrawal 36:36 [laughs]

Ali Ghodsi 36:36 I felt like the world is ending with the internet and everything. And, uh, you know, working on this most important problem that we all knew was the most important problem, which was the multicast problem-

Apoorv Agrawal 36:43 [laughs]

Ali Ghodsi 36:44 ... which none of you-

Apoorv Agrawal 36:45 [laughs]

Ali Ghodsi 36:45 Which, uh, none of you have heard of.

Apoorv Agrawal 36:47 Yeah.

Ali Ghodsi 36:48 Uh, turned out not to be a problem. Uh, but I think one interesting thing is that in 2000 we had the internet- In 2009, Airbnb was started. Right. Okay? But there's no reason why Airbnb should start in 2009. Yeah. Airbnb could have started... I, we, I've made this argument to you. Yeah, yeah. Airbnb could have started in 2001. Yeah. There's nothing like we needed something additional- Yeah, yeah ... to happen in the world. Yeah, yeah, yeah. You know? Uh, Airbnb could have happened and disrupted hotel businesses in 2001. Yeah. Yet it took nine years for someone to have that idea. Right. [laughs] That's right. And that, that was Brian. And by the way, Brian is not like he sat there and he was taking a Stanford class thinking about like a- No, no, no. Yeah ... case study project. Uh, Brian needed, like, bed and breakfast. Right. And like he was like, "This sucks." At some designer conference or something. Yeah, he was at some conference and he's like- Yeah ... "Why is this so hard?" Like- Yeah, yeah, yeah ... "Can't I just solve this myself?" So it took nine years to come up with that good idea. Yeah. So I think good ideas are very hard to come by, actually. Right. I think humans are very bad at coming up with great ideas. Right, right. Uh, and we have, like, this tunnel vision, and we focus on the wrong problems like we did with multicast in my early, you know... My PhD was really stupid. Um, so, uh, chill out and take a long-term perspective and, uh, you know, work on the things that you think will have long-term good impact. I think Jeff Bezos did it pretty well, uh, when he was an investment banker in Wall Street. Mm-hmm. And, uh, and he said, "Hey, zooming out, what's like the big thing that's happening? It's the internet." Yeah. And then he said, "Hey, let's just make a secular bet on internet's gon- there's gonna be more and more internet." Mm-hmm. "So it's gonna slowly over time disrupt things." Mm-hmm. So then he said, said, "Okay. Can we, in the long run, probably purchasing can move more to the net. Maybe not right now." And then he started with... He was very modest, and he started with kinda the dumbest thing you could possibly- [laughs] No one... It was like the unsexiest thing, which was a complete commodity that looks identical- Yeah ... and there's no differentiation, which is books. Yeah. And he just started with that, and he just bet on that secular trend. Bet on books. And every year it was more and more right, and, you know, and now it's like everything on the planet. It's the everything store. Yeah. So kind of think long term like that, and don't be swayed by the coolest thing that everybody's like right now, um, sort of making lots of noise on Twitter on because chances are it's probably something like multicast. [laughs] Yeah. Awesome. Well, thank you so much for staying longer, folks. Yeah. Sorry, I- Thank you, Ali