The Exchanges

Q tech company within India, and you're working on trust and safety for four years. Then in April, 2017, you make a very interesting And I would say somewhat radical career shift where you end up making your move into Google's AI division. Uh, talk a little bit about your move from trust and safety to AI. What about the AI division in particular drove you to want to be there?

A Yeah, for sure. So while I was at trust and safety, I was already working on a bunch of machine learning related stuff. Like we were building fraud and risk models. These were like basic models, like logistic regression and stuff. But around 2016, I think TensorFlow started becoming huge inside of Google and Google decided to open source TensorFlow as well. So that really caught my attention. And as I was working on machine learning at payments, I just realized that this, this This thing sounds really amazing and this could actually change the way we do a lot of things. So I started looking for roles inside of like Google AI and Google research. Fortunately, there was a role. Fortunately, there was an amazing manager I had who was willing to give me a shot. So I ended up moving into Google AI and I spent about four and a half years there.

AI assessment note: “TensorFlow started becoming huge inside of Google and Google decided to open source TensorFlow”

Q Very curious to see where it's going to go. Okay. What, what are you up to with Inventive?

A Yeah. So at Inventive, we are building, uh, we're kind of using the power of LLMs for enterprise knowledge management. So one of the use cases that as, as I dealt deeper into my experience with language model, uh, at Google and outside of that as well was that Like enterprise search kind of sucks. Like even inside of Google being a search company, like we had probably the best enterprise search, but it was still pretty bad. And, uh, so we were kind of exploring, we explored a few different ideas, but we landed up on this one because we just felt that time is right to disrupt enterprise knowledge management. So, uh, we're focusing specifically on sales, uh, knowledge management. So we are building an AI powered platform for sales knowledge management. And, uh, the first use case that we are solving for is enabling sales teams to fill up RFPs and security reviews with their internal existing knowledge basis.

AI assessment note: “we are building an AI powered platform for sales knowledge management”

Q of transitioned to Lambda. Blake Lemoine at Google is testing it. And I think the public really realized how like powerful this technology could be where he goes out to the Washington post with this, you know, phenomenal claim of his belief that this is a person. I mean, bring us into your seat at the moment. How does How did you see that? How did you react to that?

A Yeah, I think that happened pretty late. I think I had left Google, uh, when those allegations against Lambda came out that it has become sentient. Um, it, like, I don't, I don't think the technology is there where, uh, we can say it's sentient. I think it was a, it was a, like, Uh, how do I say it? It was blown out of proportion. And if you look at the conversations closely, uh, in terms of how he had the conversations, there were, there's something called, um, nudging the model to say out certain things. Like if you ask the question in a, in a specific way, the answer, the model will answer in a specific way. So there was a lot of, uh, steering the model that was going around as well when those conversations were released. So yeah, I, I just felt it was like blown out of proportion. I don't think we are anywhere near sentience at this point of time in artificial intelligence.

AI assessment note: “I don't think the technology is there... It was blown out of proportion.”

Q they wanted people to search. There are some search elements. People have said that like, if Google released it, it would popularize searching this way and they don't have a good business model. So I'm curious if you could weigh in on that. And then also like Bing hasn't gained any market share at all against Google. Since it came out, so did people sort of overreact to this thing?

A Yeah. I mean, I can't tell you the number of emails or messages I got from like varied people in their tech industry, essentially from journalists to like reporters, researchers, and so on. Like, Hey, is Google going to lose its market share in search and blah, blah, blah. And, uh, my take on it was like pretty straightforward from the starting. It was that, Hey, like Google has probably equal, if not better technology at hand. Uh, which I believe was Lambda at that point of time. Uh, Google has a distribution advantage as in like, there are four billion people in the world who use Google products and it's very hard to change user behavior for something as fundamental as Google search. And, uh, yes, there is a novelty, uh, with respect to Bing and all of those things, but I, I never felt that it's going to be like a major difference for either Google or Microsoft. So I think the latest reports, if I read them correctly, was like Bing maybe gained, like, one percent market share, uh, before and after ChatGPT. So did it get something, but it's, yeah, it's not bad. It's like probably, yeah, it's not a lot. So I guess it was, they were definitely, like, more panic in the ecosystem, especially from investors. Uh, but I guess, yeah, now the, now it's pretty clear in terms of where things stand.

AI assessment note: “they were definitely, like, more panic in the ecosystem, especially from investors.”

Q we're going to speak with them about what exactly happened. I think I have a clip from it that I can play here. This is with So our first episode, which is going to be running next week, is with Gaurav Nemada, who is the first product manager on Google's Lambda chatbot. What are some lessons learned for Google? Looking in the rear view there. Like, how should Google change?

A They need to. Go back to the experimental cycle, like over the years, Google has become more and more conservative about doing things. They care a lot about PR, public relations. They care a lot about how their image is shown in the media. And I feel that, at least in my experience, that played so many projects inside of Google. It was like the PR was always top of mind for leaders. And on the other side, like open AI, like they don't give a shit about PR or like for the most part they don't like they're like, okay, this is what we think is right. This is how we think is a reasonable way of putting it out. They be well, they become vulnerable. They put it out and then they kind of work with the community with respect to that.

AI assessment note: “They need to. Go back to the experimental cycle”

Q Okay. And so what were some of the things that you started to see within MENA that made you believe that this was gonna be something different from what we had seen in the past?

A Yeah. So at that point, like in one of my, so I used to drive product for a couple of research teams. One of my other teams was working on intent detection models, which is like the first step in this assembly line. And then this email comes along that I was telling you about, I play with this technology and it's like, I can ask, it can frame the question in any way, or I can ask it about any generic thing and it would respond versus an intent model will essentially fail if it was not trained on a particular set of data. So that was kind of a big moment for me. I feel that, Hey, you can, I like really ask anything in a way like this model understands that was, that was kind of the switch that went up in my mind at that point of time. And that's why I got really excited about it. That, um, honestly, I did not do it about it.

AI assessment note: “I can ask it about any generic thing and it would respond versus an intent model”