Introduction - Gaurav Madani
🫛 This edition of AI, Software, and Wetware features an audio interview on ethical AI with Gaurav Madani, a 🇨🇦 Canada-based entrepreneur and a co-founder of real estate AI startup Leni.
Leni (Leading Engine for NextGen Intelligence) is an AI super agent purpose-built for the real estate industry. Gaurav spends most of his time educating investors about how they can use AI to improve multiple facets of their business.
In this interview, we discuss:
How he and his co-founders are tailoring Leni to deliver the accuracy needed for decision-making by real estate investors
Experimenting with newly released AI tools out of curiosity, such as for combining classical Indian music with other forms
Automations he’s built around AI tools to help manage his daily workload
Witnessing AI tools that push their own foundation models over Leni and over other LLMs
His views as a capitalist about the 3Cs of creative rights (consent, credit, and compensation)
Bernie Sanders’ video on his chat with Claude, and why Gaurav believes governments need to act quickly to protect companies and people
and more. Check it out, and let us know what you think!
This post is part of our AI6P interview series on “AI, Software, and Wetware”. Our guests share their experiences with using AI, and how they feel about AI using their data and works.
This interview is available as an audio recording (embedded here in the post, and later in our AI6P external podcasts). This post includes the full, human-edited transcript. (If it doesn’t fit in your email client, click HERE to read the whole post online.)
Note: In this article series, “AI” means artificial intelligence and spans classical statistical methods, data analytics, machine learning, generative AI, and other non-generative AI. See this Glossary and “AI Fundamentals #01: What is Artificial Intelligence?” for reference.
Interview - Gaurav Madani
Karen: I’m delighted to welcome Gaurav Madani from Canada as my guest today on “AI, Software, and Wetware”. Gaurav, thank you so much for joining me on this interview! Please tell us about yourself, who you are, and what you do.
Gaurav: Thank you for having me, Karen. I’m Gaurav. I’m one of the three co-founders of Leni. Leni is basically a super-agent that is purpose-built for real estate.
A little bit about me. I grew up in India, started my career at Goldman Sachs, hated it quite fairly quickly. Did a little bit of copywriting for Bollywood award shows. Even 12 years ago, they weren’t paying right, as well. So I moved on and then joined this Indian startup called sato.com. I was doing sales for them, essentially. And along with a couple of folks, helped launch their business in New Zealand where I was at for a few years.
And then I moved to Toronto, Canada to pursue my masters. That’s where I met my co-founders. Both of them were asset managers. The original idea behind starting Leni was how can we better organize data so that investors have real time visibility into their portfolio performance? We happen to get lucky by being in the right place at the right time, because when AI came about in late 21, early 22, we had already normalized and organized a lot of good data sets and we were able to define models quickly, and exponentially increase the value we were delivering to our clients. Because now, instead of simply looking at dashboards and trying to make a sense of what’s doing well, what not so much, with regards to asset performance, our clients were able to have natural language-like conversations with their data.
So think about experiencing the delight of using something like a ChatGPT with your proprietary data sets. And also get some proactive alerts on things you need to watch out for with regards to your asset performance and all of that. So yeah, that’s a little bit about me and Leni.
Karen: Okay, great. So tell us a little bit about your level of experience with AI and machine learning and analytics. Sounds like you’re using it professionally, but do you use it personally, or did you study the technology before you got into Leni?
Gaurav: You know, despite being a co-founder of an AI company, I’m very conscious of the space changing so much that it’s hard to say if you’re an expert or if you’re adequate. I don’t come from an engineering background. But having said that, I’m surrounded by some amazing engineers. So I learn constantly in terms of how I use AI now. My role is not very different from that of someone who works in early-stage sales. So we have done some automations with regards to how our top of the funnel gets activated. But for the most part, I have tried to keep it simple, in the sense.
When I’m attending meetings, then obviously I have an AI notetaker, and then those recordings automatically get transcribed on Zapier and then are made available in a format that I have pre-programmed on my Slack. Very simply, what it allows me to do is at the end of the day, after many calls, I have a list of emails and action items that I promised to people, that I can send right now. I still maintain the human-in-the-loop and try to personally read everything before I’m getting it out. But obviously I’m also very cognizant of the fact that these tools, by being good, are creating habits inadvertently in all of us. And therefore, sometimes you can just feel like, “Hey, let’s get this out there”. There is a bunch of stuff outside of this, obviously, that we do at the company level with regards to how we build and ship our product. But for me personally, I think it’s limited to transcribing meetings, sending out follow ups, research, and those kinds of simple things. Yeah.
Karen: Do you use AI tools at all outside of your work with Leni, in your personal life? Or have you tried any of the image creation tools or videos or music or anything like that?
Gaurav: Yeah, actually, it’s interesting you mentioned music. A couple of friends of mine and I, we were just trying to see if we can create interesting combinations of AI music over the weekend. Like, say, taking a classical Indian style of music, Karnataka, and then infusing it with jazz. Obviously, for the most part, all of us are limited by our own imagination.
And I think AI has replaced search for me. So anytime I’m looking at anything, AI voice has become a big thing for me. Earlier I was very big on typing and a lot of my job is essentially writing, be it emails or articles that educate our prospects about the company. So voice is a big way in which I’m using AI.
And in the personal capacity, I think it’s mostly around this general search that I would undertake. Work itself can be fairly all-encompassing, so there’s not a whole lot of personal stuff going on outside of it, but yeah.
Karen: Yeah. You mentioned music. I think that’s an interesting area. I had done some deep dives into AI for music a couple years ago. There were a couple startups in India, one of them that stood out was Beatoven, which is interesting because they sourced their music samples ethically. And so, the music that’s generated with Beatoven, you don’t need to feel guilty about using. They had quite a grounding in the Indian classical music. So I’m curious which tools you tried.
Gaurav: We were essentially using a couple of LLMs itself in order to produce music. But the ethical conversation is something that we did get into in terms of, like, what does this mean? I’ll tell you how this came about. Interestingly, we saw a few playlists pop up on our recommendation engine. And it was fairly obvious, just because we are in this space, that we were able to tell: this looks like AI-generated music. And I wonder how long this is going to stay on the channels before it gets flagged. Creation has become so easy. And in fact, some companies I noticed are also putting out actual people who are playing AI-generated music at concerts.
For the most part, in these cases, I have been a consumer and really haven’t spent as much time thinking about these things for the most part. We are dabbling in any and every new tool. That comes about more from a place of curiosity than anything else, because we are in this space.
Karen: I’d like to hear a specific story on one of the ways that you’ve used a tool that had AI or machine learning features, both what worked well for you, and an example of something that didn’t work so well.
Gaurav: Right. I constantly toggle between the LLMs, but I won’t call any of the LLMs a tool as such. Of course, I use Leni a whole lot. And there’s bias there. And the other thing that I have personally dabbled with is Zapier, mostly because I don’t think it’s too dissimilar to, say, if someone is now using connectors on Claude or Codex. But basically what you’re trying to do is tasks or actions, right? And in any capacity, if you were doing say, A, B, C, and D and invariably that A, B, C, and D was on different systems, then ideally, you can get an agent that is able to mimic your human action and do those processes and give you an output.
A specific story for me would be: typically Read.AI is my note taker. And now what happens at the end of a call that I’m on is the detailed transcript gets pushed out into Zapier. It runs through two instances of custom-built GPTs and Claude Cowork instances that I have built out. And it essentially generates a list of action items and a follow-up email, based on the conversation and my style of writing. And a private Slack channel that I’ve created has those notes input available within five minutes of me ending the call. Now what I do after that essentially is simply copy-paste that email.
Because there’s a lot of tools available, I use superhuman as my email client. It’s a native AI tool. So essentially, I’m inputting all my emails over there, rewriting it in my voice very quickly; also getting context based on all the historical interactions I’ve had with this person, if there’s anything for me to note, and then sending this out. Simultaneously, there’s another integration that ensures all of these details also get replicated in Pipedrive, which is a CRM, so that anyone else on our team is also aware of these conversations and they’re getting associated with the specific deals we are pursuing. Yeah.
Karen: Those are some good examples of ways that you’re using it to help you to be more productive. You mentioned that you use AI as, kind of, a substitute for search nowadays. I’m wondering if you’ve had any spectacular or interesting failures in that regard, where it came up with something in search that didn’t pan out. You hear a lot of stories about, you know, what some people call hallucinations.
Gaurav: All the time.
Karen: All the time! Okay.
Gaurav: I’ll tell you why this happens as well. And this is a unique use case to us because Leni, in some cases, is an enterprise software that becomes an alternative to someone building a bunch of connectors on Claude and trying to do it themselves, right?
And invariably now, when I’m using ChatGPT or Claude and I’m uploading my transcripts and trying to work out “What exactly are buyers looking for? How can I best position our platform?” I have noticed multiple instances where the LLM is actively subduing the capabilities of Leni the agent, and showcasing its own prowess in a more favorable light.
Karen: Oh, wow.
Gaurav: Yeah, yeah, yeah. So I’m very aware of, if you were to try to get ChatGPT to answer a certain question about Claude, it is likely to be more aggressive and more out there, as opposed to if you were to press it on the same question on its own models, right? And I’m fairly certain there are enough instances that happen, and it just comes down to you keeping an eye out, and being very careful about how you prompt the LLM. Because at the end of the day, it’s a generic model that’s fully loaded with so much context and information, right? Unless you’re able to isolate context and information into a unique containerized instance, then you are likely to experience hallucinations whether you like it or not.
Karen: I know some of the tools have talked about trying to reduce that. Claude, I think, has a way now for people to specifically request for the sources that were used. And Perplexity has gone a bit down that route as well. ChatGPT, not so much. I’ve heard more complaints, I think, about that than about some of the other tools, about hallucinating or misreading.
Gaurav: We also speak about verifiability – to your point, it’s actually extremely important. And of course, in personal context, maybe the stakes are slightly lower when you’re asking for a restaurant recommendation or, say, trying to book flight tickets. But the clients that we service, for instance, are using LLM to make decisions on multi-million or billion-dollar investments, right? And you can’t have an engine that is simply designed to please, when you’re trying to get something to work. You need to trace back an answer to which report is it coming from, what is the maths that is being done on this answer, and whatnot.
So even Leni, much like Perplexity, is agentic in nature and is programmed architecturally to understand all of these nuances when it comes to the real estate data sets. And how do you ensure that there’s reliability, that precision and accuracy when it comes to each output that’s being put forward?
Karen: Yeah, that’s really important and it’s good that you’re taking those precautions. I heard something just the other day that with Claude, there is actually a way that you can configure its settings to tell it not to make something up when it doesn’t know. And my question was, “Well, why isn’t that the default?”
Gaurav: Exactly.
Karen: You know?
Gaurav: Exactly.
Karen: It was a deliberate design choice and you really have to wonder about if someone thought through the implications of that. But I thought that it was interesting that, “Yeah, here’s a way you can make it not make things up.”
Gaurav: Yeah, yeah, yeah. And surely, everyone’s smart enough to know what they’re doing, and they’ve chosen to do what they’re doing anyway, right?
Karen: Yeah. Well, it’s a default in the tool, that it will make up something plausible if it doesn’t have an answer. And it will sound confident in doing that, which is, I think, the more worrisome part: that you don’t really know, necessarily. And when they introduce ads, that may be another factor that could be influencing what you get in a way that you maybe don’t always realize. Is that a concern for you at all, with regard to how it integrates with Leni?
Gaurav: I would have some separation in terms of Leni versus how I think about it personally when it comes to ads. And personally, I don’t care as much, just because I feel all of us are being programmed all the time anyway. And this is a new medium. Corporations have a way to get to us. And of course, now it’s become pervasive, in the sense it is so ingrained in our deepest thoughts that it has greater power over influencing our actions. That goes without saying. Even as someone who is acutely aware of this, I know that whether I like it or not, I am being programmed or engineered to make certain decisions, based on everything that I’m consuming.
For Leni, I think the conversation becomes more around anti-competition or what is fair use in terms of, say, if a user is objectively asking for the recommendation on an engine, then is the engine actually providing objective, fair evidence about what is the right solution for a certain requirement? Or is the engine indexing its solutions over others? And that’s, again, at least as far as I understand, not unique to Leni. Because there are three foundational model companies at this point, at least in this part of the world that have a substantial user base. When I use Gemini versus Claude versus ChatGPT, given my understanding of the unity economics and the fact that everyone is right now bleeding money, invariably, whether we like it or not, they aligned towards increasing their revenues and profitability. And we all know what that drives corporations to do.
Actually I’ve written about it recently. While it’s amazing that all of us are having these conversations, truth be told the ownership sits with the governments because they’re the ones who can really do something about it, you know?
Karen: Yeah.
Gaurav: I don’t think I expect these companies to act in favor of the average individual, even if their marketing campaigns speak to these stories or visions, just because their incentives are not aligned, and it would be naive on our part to believe they’re acting in our best interest. Yeah.
Karen: Have you noticed any instances where it’s trying to do math, but it makes mistakes on the math? That’s one thing I had heard about the earlier versions, that they were very much not good at doing calculations. And here you’re looking at numbers and evaluating return on different potential investments and such.
Gaurav: That was one of the big reasons behind putting Leni together. Of course the foundational models are getting better. But the thing is these are language models, deterministic in nature. So they are reasonably predicting what the next token output or the answer needs to be. And that’s where you need to configure scaffolding and provide context that allows these models to, I suppose, increase the accuracy with which they’re giving you outputs. And of course now, another good thing is there are multiple third-party benchmarks that determine the quality of output as well.
So very often, when you look at, say, a bullshit benchmark – there are bunch that even we benchmark ourselves against and publish on our website. Essentially you are trying to run similar use cases with similar data sets, on, say, a generic LLM versus a purpose- built agent to see what the accuracy of output, the depth of output, the quality, the speed, and all of this looks like. Yeah.
But they’re improving. I think they’re better at doing maths. But now the ask is, can it generate models? And then the ask would be, now can I generate a model that’s exactly as per my investment thesis? And then there’ll be a point of failure there. Yeah.
Karen: So you’re making good use of tools in a lot of areas. Are there any things for which you would not ever use an AI tool? For instance, you may use it to help you write messages to prospects, but would you use it to write a personal message to a family member or a friend? Or are there other cases where you’d say, “No, I’m not going to use AI for this, ever”? Or are there really no such examples that you can think of?
Gaurav: When it comes to writing, I still think it’s a tool that’s helping me frame things in a better manner. One of the things that has unfortunately happened is my ability to construct sentences is deteriorating now because of AI, right? And I think we are okay, because fortunately, I grew up in a generation where at least I had traditional schooling. And at least for the first few years in my career, I was still writing without being prompted in terms of how to write. I think the problem would be a lot more grave for kids today who probably don’t have that ability, right? We have seen issues with folks who, say, grew up in an age of social media from the get-go. And fortunately, I’m happy I did not have to do that either.
In terms of, where would I not use AI? The only thing that I’ve not touched is, say, banking. Because I’m, like, “Don’t give access to any of the accounts.” But outside of that, I think, of course, sensitive information. So you don’t want to upload your social security, passwords, and things like that, wherever you can, whatever you are sensitive to.
For the most part, I think it just comes down to judgment, in terms of information that you’re okay with an engine having. I don’t think any engine is personally targeting an average individual. So I do think, on the other side as well, a lot of fear can be unfounded. Sure, operations are trying to enhance profits, but I don’t think individually anyone’s being targeted.
And then, of course, it opens up another Pandora’s box about data breaches and frauds and things like that. Those kinds of things are always happening. And now AI has created a new way for it to happen, and therefore we’ll have to think about security guardrails and all of those things. But there’s tremendous productivity as well to be unlocked. So both of those things go hand in hand. Yeah.
Karen: Yeah, we connected on LinkedIn, so obviously we’re both active there. Have you any concerns about LinkedIn having use of your data? There was a bit of a flap of not quite two years ago about them saying, “Okay, we’re going to opt into using everything you’ve ever put into the site as fodder for training our machine learning models.” Did you opt out? And what are your thoughts about that?
Gaurav: I opted out there. Again, how I see it is: invariably, if you’re trying to profit off it, then maybe I, as someone who is creating the foundational data at the back of which any of this is being done, should be entitled to some form of profit share or revenue share. If you can’t tell, I’m a capitalist, by the way. So, yeah, I did opt out of it. I think if any of these companies have to use our data, then it would be nice for them to outright also have a model, much like, say, YouTube creators or any creators where we get paid for what they’re using. Yeah.
Karen: It sounds like even as a capitalist, or perhaps because you’re a capitalist, you feel like they should be either getting our consent, or giving us credit, or compensating people when their data is used.
Gaurav: Yeah. And I think both things can go hand-in-hand. It’s not either though, right? I’m also now building a company and of course we are dealing with enterprises. So we have to go through hundreds of security audits on data and whatnot. But what I was saying was, these are decisions I think users are also making for themselves. Because now, whether I like it or not, every time there’s a software update, I don’t read Apple’s privacy policy. I just opt into it because I don’t have a choice. LinkedIn gave us a choice; I acted on it. So, yeah. Yeah.
I still think the folks who are quiet in these conversations are the governments, because they are the ones who are essentially allowing all of this to happen. And they can be a lot better in terms of how companies are treated, how they are taxed, how individuals are voted, how individuals are taxed. So the onus is on elected representatives who are in charge of creating policies. At the end of the day, I have an opinion. You have an opinion. But policy is what is going to move the needle, right?
Karen: Yeah. And that’s one of the concerns I’ve heard some people express is that the technology is moving so fast that governments, even when they are aware of the issues and trying to keep up, there’s a big lag. And so regulation’s always going to lag to some extent. But when the technology moves this fast, it’s even harder.
Gaurav: Yeah, and there was an amazing 10-minute video where Bernie Sanders talks with Claude, and some good outcomes come off it. I think it was published over the weekend itself. Now it’s unfortunate, even, that it does not get a lot of popularity or mass views. But I actually think the technology, while it’s moving quickly, can be reeled in. And from a policy standpoint, I mean, I don’t want to sit on this podcast about AI and bash the government, but innovative, I think I’ve been clear about where I stand on this. The governments always move slowly. And I have a lot of frustrations around, like, I sell into large enterprises, they move slowly. Governments are so much worse. So, yeah.
Karen: Yeah, I am glad you mentioned the video with Bernie Sanders talking to Claude, because that did not get onto my radar at all. I don’t spend a lot of time online looking at YouTube, but I missed that one. So now I’m going to go look for it, and we’ll put the link into the interview. [Link]
Gaurav: It’s just 10 minutes. It’s really clean and nice. I think it is a conversation that everyone should see, because Bernie Sanders is a people’s champion, right? And I think he poses some pertinent questions. Questions that all of us have. Because of the fact that I’ve now been building Leni for the last few years, I don’t think of AI as the average individual, whether I like it or not. Because I’m so inside it right now as it’s being created. But for someone from the outside, who is not quite sure what to make of it, is maybe using AI once or twice a week, I think it’s a good conversation just to see and decide for yourself, in terms of where you stand.
Karen: Yeah. That’s great. So when you’re working with building Leni, are you getting data when you engage with a customer? Are you working with that enterprise or customer’s specific data, and how are you navigating that?
Gaurav: So we work off of containerized models. What that allows us to do is, within the environment that we provide to our clients, they’re interacting with models there. And from a security standpoint, one of the required protocols is we are a SOC Type 2 certified company. So we go through third party audits to showcase how we got enough data based on org guidelines, so no one outside an organization has access to that company’s data. All the learnings that get derived from that data, all the feedback loops that get created, only benefit the client, but not other clients, per se. And the data, they are opting in and providing it themselves because at the end of the day, they are the ones who need to create reports off of it, understand patterns, strengths, and things like that. We are merely providing the infrastructure that allows them to do all of these things.
The alternative is – which is already happening, and I don’t know how some people feel about it – an analyst or associate at your company is just going and uploading it into the cloud and just trying to do the work that took two hours in five minutes. So that security breach anyways happened in hundreds of organizations, and we are a secure alternative. So yeah.
Karen: That’s what they call shadow AI, where there may or may not be a corporate policy about use. Or if there is, it’s not clear to people on how they’re supposed to use it in their day-to-day jobs. And so people end up trying tools that maybe they shouldn’t, but it’s not necessarily very clear to them what they should or shouldn’t do. Definitely some breaches happen that way.
Gaurav: And to your earlier point about things moving fast and policy making, taking some time to catch up, that’s also happening at these companies. There are a lot of companies right now that even I talk to, without taking any names, where the leadership is aware that an associate is generating this report quickly on an LLM. They are trying to figure out, okay, we know our company is doing this, but in the meantime, what can we do to get this to a place where our team can engage with it in a secure environment? And of course, there’s copilot and things like that, but the thing with copilot is people are not really enjoying the quality of output. So while it’s providing security, it’s failing on other parameters. So there’s that and lots of trade-offs.
Karen: Yes, yeah, definitely lots of trade-offs. So you mentioned breaches, and we’ve talked about them just a little bit. Are you aware of any times when your personal data has been involved in a breach? Has that happened, or have you been able to avoid that so far?
Gaurav: I know that my phone number is out there for sure, because I work in a sales role, right? So I’ve had an instance or two where people have called me saying, “Hey, I got a call from this number requesting some information.” And I was, like, “It wasn’t me”, you know? So, yeah. But nothing beyond that.
Of course, there are regular phishing attempts that all of us, I suppose, get on our phones. Hopefully, at least so far, nothing that I was not able to immediately identify.
Karen: Okay. Well, that’s good. Let’s hope it stays that way. I’ve probably gotten a notification on a breach at least once a year for the past, I don’t know how many years now. Not something with a bank account – fortunately I haven’t had any of those issues. But things like my phone number leaking out from the time when I rented a real estate property 20 years ago. That data is very old, but now it’s being scraped and used for new purposes. There was one that was a medical organization and some of their data got breached. It seemed like there’s one every year down here. You’re lucky if that’s not happening to you.
Gaurav: Yeah. Yeah.
Karen: I hear a lot of stories from people that have had more serious issues.
Gaurav: No, I know of one friend, not exactly a close friend, but a friend of a friend who did fall for one of these classic phishing attempts and lost some money in the process. But for the most part, I think now these phishing attempts start with a notification about a data breach, right? And that’s when you get alarmed and if you are not familiar, right, you get sucked into it, then yeah.
Karen: Right. And there are all these companies that offer to try to protect you from it. But when I was talking to my husband, he said, “You know, you have to give them all of this sensitive personal information for them to protect you. How do you know you can trust them?”
Gaurav: Exactly.
Karen: So how do you win?
Gaurav: Just don’t give any information out.
Karen: Yeah. So it’s an interesting trade-off: what protection they can give you versus the risk that they’re going to end up compromising something of yours.
Gaurav: Yeah.
Karen: Do you know of any companies that you gave your information to – social media or something – that you knew that they might use your information or would use your information for training an AI system?
Gaurav: Social media, I give out everywhere. And by the way, LinkedIn is the only social media that I’m active on. I mean, I do this for work, but I was one of those guys who got on Facebook only in, like, 2018, because my university needed me to have a Facebook account. I actually am fairly private about what I share, and not a whole lot of my private life is out there on the internet. So whatever you see is very intentional and business related and that I don’t mind putting out there.
Karen: You mentioned Apple and devices and not reading terms and conditions. They’re actually usually pretty dense. I usually skim through them now just because that’s something that I tend to look for, but they’re hard to read. They pop up on your phone and it’s, you know, 20, 30 pages of scrolling, and it’s not useful informed consent at all.
Gaurav: That’s what I’m going to do. If I don’t agree with it, it’s not, I’m not going to update my software and get stuck. It’s not like I really have a choice. It’s an illusion of choice, at the end.
Karen: Yeah. Well, last question. So we’ve seen public distrust of AI tech companies has been growing recently, especially as we find out what they’re doing with our data and looking at ways that they are, in some cases, not taking any pains to protect our data. And as you mentioned, our governments are temporarily letting them get away with that. But nonetheless, I think it’s healthy in a way that people were trusting them less because we need to be more cautious about what we do and what we share. But from the point of view of a company becoming trustworthy, in deserving our trust, your trust or my trust, what’s the one most important thing that you think companies would need to do to earn and keep your trust, if that’s even possible? Some people say it’s not possible.
Gaurav: There’s something I think about all the time, right? And opinion, again, this can be subjective and it means different things to different people, but I think for any company to gain my trust, the company needs to do what it broadly advertises it’s going to do. So what I mean by that is, say, if I’m buying an Apple iPhone, I want the phone to work well. And if I’m using a particular agent, then I want that agent to reliably output what I’ve tasked the agent to do. And that is a function of usage and trust.
And the second thing, of course, is when it comes to data usage. Like I said, I’m less concerned, but others might be extremely concerned about what’s being done with their data. If companies can make it easy for people to do two things – firstly, understand in simple words, what are you doing? And stuff like putting it in like dense language that’s written out over 30 pages. And the second thing is: make it easily accessible for the person to opt in or opt out.
Beyond that, at the end of the day, I do think innovation unfortunately does require data. So if you want to benefit from the output that these models are giving you, then you would have to provide it. But I think as long as there is little or zero ambiguity between the transaction that’s taking place between the company and the user, then that would go a long way in establishing trust. But the first thing, I still think, is: let the product do what it needs to do. And then I think the policies follow.
Karen: Okay. That makes sense. Well, thank you so much for taking the time to share your thoughts about AI with me and with my audience. Is there anything else that you would like to share? Any upcoming events, any closing thoughts, anything like that?
Gaurav: Thank you so much for having me. If your audience, or members of your audience do work in real estate, feel free to check out Leni and how Leni helps real estate professionals. We have a lot of cool product updates and announcements in the works throughout summer. Yeah. And it was nice talking to you. Thank you for having me.
Karen: Great, great. Thank you so much.
Gaurav: Alright, cheers Karen.
Interview References and Links
Gaurav Madani on LinkedIn
Leni on LinkedIn
Gaurav Madani on Substack
About this interview series and newsletter
This post is part of our AI6P interview series on “AI, Software, and Wetware”. It showcases how real people around the world are using their wetware (brains and human intelligence) with AI-based software tools, or are being affected by AI.
And we’re all being affected by AI nowadays in our daily lives, perhaps more than we realize. For some examples, see post “But I Don’t Use AI”:
We want to hear from a diverse pool of people worldwide in a variety of roles. (No technical experience with AI is required.) If you’re interested in being a featured interview guest, anonymous or with credit, please check out our guest FAQ and get in touch!
6 'P's in AI Pods (AI6P) is a 100% human-authored, 100% reader-supported publication. (No ads, no affiliate links, no paywalls on new posts). All new posts are FREE to read and listen to. To automatically receive new AI6P posts and support our work, consider becoming a subscriber:
Series Credits and References
Disclaimer: This content is for informational purposes only and does not and should not be considered professional advice. Information is believed to be current at the time of publication but may become outdated. Please verify details before relying on it. All works, downloads, and services provided through 6 'P's in AI Pods (AI6P) publication are subject to the Publisher Terms available here. By using this content you agree to the Publisher Terms.
Audio Sound Effect from Pixabay
Microphone photo by Michal Czyz on Unsplash (contact Michal Czyz on LinkedIn)
Credit to CIPRI (Cultural Intellectual Property Rights Initiative®) for their “3Cs' Rule: Consent. Credit. Compensation©.”
Credit to Beth Spencer for the “Created With Human Intelligence” badge we use to reflect our commitment that content in these interviews will be human-created:
If you enjoyed this interview, my guest and I would love to have your support via a heart, share, restack, or Note! (One-time tips or voluntary donations via paid subscription are always welcome and appreciated, too 😊)
