The Exchanges

Q From all of your experiences, how do you feel like product is changing right now?

A That's a great question. I love jumping right in. Uh, you know, I think the way that the product is changing right now, uh, is, is In a variety of different ways, you know, you always have to understand the history to understand the present and the way that product used to be way back in the day. Is that, uh, you know, people have the role or the title of product manager and people kind of like deferred to them because of that title. And I think that what you're seeing is, is a continuation of a backlash to this idea that product people are just sort of like put into this position to make decisions about what the product should do next. When the reality is there's all this data that's available that can help you to make those judgments. They can help you to make decisions and product managers need to be really data driven. They need to have real evidence. For why customers are going to use this next feature, or they got to have real clear information about why some version of a product is better than another one, or they, or why one design is going to perform better than the next, et cetera. So you kind of think of product managers less as being sort of like just straight up decision makers, you know, making their best kind of like judgment calls, but really making smart decisions around how they use the data that's in front of them to make the decisions that they need to make.…

AI assessment note: “what you're seeing is, is a continuation of a backlash”

Q to know from your side, do you feel like The number of people that will remain in the product team will definitely diminish, and if they do, will that enable them to earn more money because they'll be seen as higher value, or do you think they'll continue to be paid what they're being paid now, excluding market fluctuations and inflation, et cetera, just because they're the ones that remain?

A Yeah. I'll cover the first part first. Do I think that the size of product teams will diminish? Yeah. You know, I think over time that's, that's certainly going to happen. I mean, let's, let's look at historical precedent on this, right? There's a lot of other, you know, we, we all look at the innovations of today and we think, oh my gosh, how's the world going to change, you know, et cetera. But the reality is there are plenty of innovations that have happened in the, in the past that at the time, you know, we look back at them now and we're like, oh, that's obvious. Or, you know, it seems pretty basic, consider it, but like, At the time, it was probably perceived very similarly to the same way that we're looking at generative AI and things like that now. I mean, you know, imagine the world when the calculator was invented and you had all of these, um, accountants that were out there doing calculator work. In fact, there was even a job called calculator, right? And like, they were like human calculators. That's what they did is it's just kind of like crunch numbers all the time, right? And they needed to do it without mistakes, you know, et cetera. And they had their various tools and eventually this new technology comes out. That sort of like replaced the need for human calculators, but it didn't sort of like take away all accounting jobs, right? Like what it did is it just m…

AI assessment note: “I'll cover the first part first. Do I think that the size of product teams will diminish? Yeah.”

Q off of a few words. Now, the quality of those things are not great, but the fact that it can be done at all is pretty fascinating as it is. That's kind of my own perspective. What do you think, what tool or tools are out there that you're excited about that you think are going to really change the game? And I don't mean ChatGPT. Something different. And how.

A Sure, sure. I mean, so there's a lot of attention right now on generative elements, the usage of AI to kind of predict the next word, to predict the next, you know, frame in a video, you know, things like that. Um, I'm honestly a little bit less excited about that from the perspective of product than I am about the analytical side of things. So this is more taking the information that you're already seeing and then making assertions based on it. Right. So if I could just sort of feed a bunch of data in and have it spit out for me, all the insights that I really need to know, account for bias in the data, ask the right questions, the data, et cetera, you know, and then kind of like deliver to me the set of insights that I need to know. And you can be even couple that. Potentially with generative, you know, things you can say, Hey, here's a new feature that you can conceive of that might be included in your product based is that by itself, not all that interesting, but if you actually linked it together with the output of the data and you said, because the data looks like this, you can see that there's an opportunity that's emerged. And this is an idea that of the next thing you could do with your product to kind of like go after that opportunity. That's kind of like in a lot of ways what, you know, at least the basic role of a product manager has sort of like, you know, been his…

AI assessment note: “I am about the analytical side of things.”

Q to help me, you know, show me how it would make sense if I were to write this in this language. It could do all of those things. Now, again, it's just a vacuum. So who knows if it's actually valuable, but on its own, it's quite interesting. Uh, what tool do you know of that, that people are using that ingest the data first before it goes to Chatwt?

A Yeah, I mean, my understanding is that there are, um, you know, that GPT-IV actually does allow you to infuse a bunch of information or context into it, but I don't know how sort of, like, reasonable, sort of, like, how meaningful that actually is. Like, what it's still doing at the end of the day is predicting the next word that you would have in a sentence, you know what I mean? It's kind of, like, generating content based on the kinds of things that it's seen historically. And it's trying to predict that and it might be able to modify it based on inputs that you can provide. But I think that that notion of generating the next most likely word Is exactly the problem, right? It, it, it means that the output's going to look like it's going to smell like, you know, content that a product manager might produce because that's exactly what it's being asked to provide is it's not saying generate for me this next great idea. It's not saying, you know, you're, you're not, you're not actually prompting it with something like that. What you're prompting it with is give me content that would pass a sniff test of making it look like it was the next great idea. And those two things are very different from one another. Right. And that's because whether it's chat GPT or any other, these other kind of like generative models, they, or, or, or AI systems, they don't have an underlying model of …

AI assessment note: “GPT-IV actually does allow you to infuse a bunch of information or context into it”