The Exchanges

Q Now, yeah. And I know you say it's not a duel, but what goes through your head before you say, hey, okay, we're gonna, we're gonna go do this. Uh, you know, obviously people are gonna look at it as competitive to your old employer. Did you worry about relationships there or, um, you know, how it might be received and what has the feedback been from your former colleagues?

A You know, I obviously do worry about it. I have a lot of close relationships with a lot of people at Google. Um, and I would roughly say that feedback falls into two buckets. One set of people that go like, yeah, we understand why you're doing this and why you didn't think you could do this within Google and a different set that kind of goes, um, you know, we wish you really, you had done this within Google because if anyone could have changed what Google was, it should have been you. Um, you know, both are reasonable points of view. And there are some people that, you know, kind of don't just want to deal with it. This is all too much, um, for them. Um, and I respect those points of view, but at some level, one has to be driven by what one sees is, uh, you know, is the right long-term outcome. I personally do not think of ad supported free products as being good for consumers. Um, good for our country in the long term, um, because it is very hard for them to stay true to what you and I want as users and as customers of these, um, you know, of, of these products. That conflict of interest is just, is just really, really unavoidable. And the fact of the matter, uh, Alex, is that, um, while at one level the products are free, All the benefits of scale for products like this, they go to the creator of the product. They don't come to you and me. You know, when it comes to Neva, for…

AI assessment note: “I would roughly say that feedback falls into two buckets.”

Q able to speak with it. So you could have like anybody in the organization access, you know, whatever part of the data is, you know, available to them and actually start to have a conversation and not have to run like complex, uh, coding. Algorithms in order to be able to make sense of what's going on in the company. Is that what's gonna happen? Like give some practical examples.

A Yeah. So, you know, Snowflake is proud of its mission to democratize data access to everybody within the enterprise. There are companies like Fidelity that have made Snowflake the centerpiece of their data architecture. What we are excited about being able to do, um, is use the part of Generative AI on top of this incredible, uh, platform that's already been built. Um, it ranges from the simple, which is, how do we help you generate much better SQL queries? Um, we have something called Snowsight, um, which is where you type in SQL queries. Uh, I don't know about you, but, you know, I've spent a good chunk of my life writing SQL, um, even at, even at Neva. Um, and it's tedious. It is, uh, it is tricky to get right. Um, we want to make it much easier so people that are doing this who are typically analysts, data engineers, uh, can do this 10 x faster. Um, but even more importantly, and this goes to the point that you're talking about, um, is how do we make it easy, um, for business users, um, that don't necessarily understand the ins and outs of the schemas and the tables and stuff like that, To be able to ask business questions, um, and for Snowflake to then automatically decide, is that an existing dashboard? Is that a SQL query that's been run before? Do we need to write something new from scratch and visualize it? It is that ability to offer up this data, and this is everythi…

AI assessment note: “for business users... To be able to ask business questions”

Q reading the article. So does Google prioritize those? Like, you know, obviously it's not going to make me click more often if it's in the page versus in the, you know, the, the blue link and the description. So does Google prioritize that stuff as well? Uh, looking at the page and saying, okay, you know, this is more clickable because there's panic here and let me show it earlier.

A There's never any, uh, explicit intent for things like that. It is, um, what are the metrics that you optimize for? It comes down to things like if you optimize for click through, you get a certain set of results. Uh, similarly on social media, um, optimizing for things like total time spent, uh, the YouTube team has long had goals. You know, this is years and years and years ago, and people like this year have talked about it. Um, they optimized for watch time. Um, and why do you optimize for watch time? Because the more watch time that is, the more ads revenue that there is. Um, and when you optimize for watch time, you end up recommending things that are more and more sensational. Uh, so it becomes like, what is the core model? An ad supported model inherently drives, um, attention inherently drives towards the sensational. Um, to us, part of the benefit of the subscription model is that it can focus a lot more on what is authoritative, what is higher quality information for you. Niva doesn't care if you go to a, um, a retailer versus if you read a review site versus if you place a price alert on, uh, you know, on, on the engine and come back four days later. They're all perfectly great outcomes for us.

AI assessment note: “There's never any, uh, explicit intent for things like that.”

Q is another bot, people will type things into a chat bot, uh, that they will never dream of typing into a search engine. So, I mean, tell us a little bit about what you saw on the other side. Like what, what do people type into these bots? And then how does that, is it even search at that point? Like, how does it change what we, what we see?

A I think it's a very different product. And I think it's fascinating to watch Pi, to watch Character.ai and all of these people, um, Create products that are very different from search. And even in the context of search, the kind of questions that you would ask of it, um, have changed in a big way. Uh, the one example that I'd like to, uh, you know, give people, but there are many such examples, um, is, uh, Jason Calacanis, uh, as you know, runs like another podcast. Um, and the question that he asked Neva, um, was, Hey, how are the Knicks doing? And he was, this was early this year and he was offended that, uh, um, we gave him, uh, a summary of articles, uh, from like late December, because that was the best that the search engine could find in terms of how the Knicks were doing. Obviously the season had changed. Um, once you get used to the idea that you can just say things, I think the set of, uh, questions that you can ask dramatically, uh, change. Um, you will ask a lot more subjective questions. Remember at the end of the day, like the search engines of today are quite limited. If you ask it a deep, complicated question, um, you get a bunch of like gobbledygook pages. Um, and so I think that is, that don't, don't really have a whole lot to do with the question that you asked. Um, and so I think we ask a lot more subjective questions. What do you think this article says? Th…

AI assessment note: “You will ask a lot more subjective questions.”

Q to build a model like ours, but it won't work. And now with things like DeepSea, Kimi K-II, we've seen people able to catch up on that front. So it's being pushed by Google on one hand, the open source model builders on the other. Help me figure out how OpenAI can, can continue to lead this, this race if it can, or is it just one in the pack?

A I mean, I think the fact that it has become, OpenAI has become the Google of choice when it comes to chat for most of us, that's actually a durable advantage. And, uh, I, I, I, I use it quite often for all kinds of things, including solving problems in the real world, my coffee machine not working, um, or I can't open my gate anymore. Like, the amount of use that you can get is pretty remarkable. I think that lead is real. On the other hand, something pretty simple, like, not simple, it's hard, faster image generation, or more accurate image generation, which is what Google pioneered. With Nano Banana, it's actually having a profound impact on things like their usage, and OpenAI was late to the game just for that one feature. You think, come on, it's a small feature. How much can it matter? It matters. People like being able to create things. It just tells you that, yes, competition is actually very fierce, and, uh, big companies Generally have a lot of birthing issues when it comes to new things. It's just, it's a matter of how they work. First of all, they don't often have a clear perspective of what amazing means, um, in a new area and, uh, what they struggle with, even if they can understand amazing is fading out a path to that amazing. One can argue that, uh, XAI, for example, has actually produced what is widely acknowledged to be a world-class model that is out there. Bu…

AI assessment note: “OpenAI has become the Google of choice when it comes to chat”

Q 31 in over in a year. And of course we factor in COVID, but this party can't keep going on forever. So, or maybe it can. Um, so I guess like, yeah, the core question I want to get to you is, How long is this going to last? Do you see this eventually, you know, coming to a ceiling in digital advertising and what are the implications of that?

A At one level that's going to happen. When digital advertising is a large is like most of all of advertising. And I think from that perspective, um, we have not quite hit the ceiling yet. We have hit ceilings in a number of areas, other areas like, uh, smartphone sales. Uh, you know, that year on year, it's not really growing significantly. Um, and so there are other changes. I think it's just that the move to online advertising, Is part of the way there. So I would say there is some, you know, some more space, um, over there. Uh, but, uh, all of these converts to the level of GDP growth, which all of us know is nowhere close to that.

AI assessment note: “we have not quite hit the ceiling yet.”