Our guest this month is Monish Gandhi, founder and CEO of Gradient Ascent AI. You'll learn about the spread of AI in our everyday lives, investing in AI, profit generating outcomes and using data creatively.
For more information about Fusemachines, please visit https://www.fusemachines.com
For more on Gradient Ascent AI https://www.gradient-ascent.com
Music: Welcome to the Show by Kevin MacLeod
Link: https://incompetech.filmmusic.io/song/4614-welcome-to-the-show License: https://filmmusic.io/standard-license
So anyone that's listened to any of my past conversations knows that I sadly, always like to start off with the same question. But I like doing that because I always get very, very different, wildly different answers. So Monish what I would love for you to share your path, how you got to be where you are, typically, these are in person, or I'm in your office, that's how you got to be sitting in the chair that you're sitting in, although, clearly, I imagine you're home.
I am home. But you know, everyone's journey, as you said, is, is different. And I've been very fortunate to get to do what I really love to do. Passionate about sort of machine learning and AI for a long time. You know, my final year engineering project, many, many years ago, with friends was actually, you know, we got a pool table, we put a camera on top and a projector and as you played, you would give you sort of the next best shot, right? So I've been passionate about this for a very long time. But at the time, you know, AI wasn't as big as it is now. So majority of my career has been in enterprise software, really helping people buy or also build good software, right. So I've been a product manager, team lead, sort of dev lead. But a few years ago, it sort of became clear to me that, you know, now people are once again starting to look at machine learning and AI and starting to be excited about it. So if I really ever want to do what I love to do, maybe now is the time and in through my work, I had sort of seen the challenge of companies struggling to figure out how to start with AI, right? Like, what does the team look like? What do the skills look like? What does the data look like? Like how do we actually ship something like this, we really sort of focused on how do we help companies build great products that are data powered, that are AI powered that help them differentiate in the marketplace, or help them automate? And, and really try to get it to the market in a differentiated manner? So how do we help them thrive in our AI powered world, right? Like, how do we how do we do that for companies? Whether they have the skills in house or not? So how do we help them get started? Right? So that's really been a big motivator for us.
I love that I love that we'll dive into I want, you know, a few questions in terms of how you help folks and where you see it going, typically, but dive into some of the issues that that folks run up against, because I imagine a lot of companies or a lot of a lot of individuals can resonate with with that. So I'm gonna love to deep dive into that. And, you know, one thing that I think kind of struck me as we started our initial conversations was the similarities in terms of kind of our team, our focus, but specifically our missions. And we've got very prominently featured, you know, kind of everywhere, the concept of democratizing AI. And it's something it's, you know, it's funny, I joined initially, and I'm like, well, that's great, that's kind of marketing speak. But as, obviously, I've been here for a few years now. And it's just absolutely the truth. What does that what does that phrase or that concept, the ability to democratize that type of technology mean to to you and your team?
For sure, and and, you know, while we don't use that phrase, because it is often more frequently used, but the idea of, you know, help companies survive and thrive and grow in our AI powered world, sort of that democratization really is at the heart of what we believe too and so, so our thinking about, it has always been sort of my core thesis has been, you know, AI will be everywhere, right. And when I, when I use the word AI, I really am talking about data science, machine learning AI, right, so I'm just gonna word AI as a bit of a bit of a shorthand here. And that it will take over in some ways, on all aspects of our of our lives, right, whether that's home, whether that's cars, whether that's our work, whether that's variety of products, we use processes that are going to get automated, it will impact all the various aspects of our society. And, and what was sort of weird to me or what was in some ways scary to me, is that a lot of the AI technology, the core technologies, while they're open source, a lot of people use, you know, API's that are controlled by, you know, few companies. Right? So, you know, in single digits here, right? And I think that's not good, like those types of oligopolies around our data around these types of transformational technologies is really not good. So so our, our our sort of way of working on this has been different from Fusemachines, but it's really been about how do we go to even smaller companies, whether they're tech companies or not, and say, Can you own your own Um, piece of AI technology, right? Are you in control of your own destiny? Are you protecting your data? How are we setting you up for success in this increasingly AI powered world? So our focus around democratization has really been about, you know, are we helping businesses control their destiny? Right. So, so, really, that alignment of ideas, and obviously, you're looking at it very differently.
You know, it's, again, it's interesting, again, with all the complexities that you bring up, and the there's there's quite a bit of challenge, a lot of confusion. There's folks that are in leadership that don't necessarily understand this. But I still feel with all of that companies are diving in, or at least dipping their toes. And I'm curious why you think that is, as opposed to a few years ago, and that it does seem as though as I'm paying attention to the news, there's really cool things with generative AI and just companies kind of coming up from all over the place the clients that we're speaking to you. We've got experience in this. But again, you kind of think, well, this is interesting, this is you wouldn't expect this type of a company to be the at the phase to be looking into this. Now, why do you why do you think that that is that it might be a bit of an inflection point?
I think that AI is now getting into more and more hands, whether that's through their phones, whether through their cars, whether that's through something like Alexa or their Roomba, into more and more homes, right? And then this next part is, you know, media movies. And then news, as you said, whether you're talking about ChatGPTs, or self driving cars, or you know, when AI does something bad, right? More and more people, it's on people's minds that this notion of hey AI is going to be in all aspects of our lives. I think people are starting to see that, right? And as customers demand, as competitors do things, as partners sort of talk about it, I think more and more businesses are kind of saying, Okay, wait, are we the laggards here? And I still don't think that's the case, but I think it will very quickly become the case where majority of the companies have some, whether they're building it themselves or not, they're at least using these technologies to you know, make their make their business better.
You know, and it's funny, I had a bet with my brother in law several years ago, because I think at that time, I just believed everything out of Elon Musk's mouth, in which I thought that my I think he was 10 years old, my 10 year old nephew would never drive a car because I was convinced by the time that he you know, six years into the future, my family, my sister, and my brother in law would have would have bought a Tesla, and it would have been to that phase and and I've gotten beyond egg on my face with with that. Absolutely. Fair enough. I mean, I didn't, I didn't purchase the full self driving. So at least I can put myself in the camp of not having to not having paid for it but I on on that vein you think, you know, for some companies it might be now, some companies might think it may not be the time and when you What do you think are some of those some of those reasons that it might not be a good fit for some companies?
And so I think the the aspect that people sort of miss about AI a lot is that until you do it, it's actually hard to know if it will be possible or not. Right? So if we think about the way software has been for the last, whatever, 2030 years, right? The question was always, you know, question was never can we do this? It was how long will it take? How much will it cost? You know, so the feasibility of what you wanted to do was never in question. Whereas AI almost always gets used to do things where it's like, until you try it, you're not going to know, and it might take, you know, three months to solve a problem. Or it might take three years to solve a problem. Or it might take 30 years to solve a problem. And self driving cars are a great example. Right? Like, in some ways, I think there were examples of you know, young hackers, basically creating a, you know, self driving car up and running within months, right, then to get it to be something that you would trust day in and day out. Right? You have scenarios. We're not there yet, right? And the flip side of that is something like say ChatGPT or DALL-E, which is another OpenAI product, that helps you create images. There's Stable Diffusion, Midjourney. There's a bunch of these where you provide a text and it will generate an image. And, you know, if people had asked even two three years ago, right before the first sort of DALL-E came out, how close are we to being able to do something like this? Most people would have said, you know, a long time and suddenly we've kind of figured out how to do it. So so there is there is this, you know, unknowability about when are we going right? So that presents this tension of, you know, we've been talking about self driving cars for 10 years, there's still and this ability to create incredibly beautiful AI art. We very rarely talking about it, and suddenly it's in everybody's hands, right. So. So that's what's fun. But let's sort of if you think about that, from a business perspective, when you're considering investing in AI projects, you do want to sort of go in knowing that, hey, some of these are still experimental, they're hypothetical, and sort of know, what does your sort of investment model look like? It doesn't mean you don't try these things. It's just that, you know, you kind of have to know, where are you on that feasibility sort of, sort of, curve or model? Is this something that you can do in three months? Or are you willing to invest for three years? Because its core key strategic sort of IP for you?
I want to dive into that feasibility. Because again, I have conversations with family, I, me being me, me being considered an expert, because I'm the one that knows how to install print drivers. And those are the types of conversations I'm in now I actually have an expert on the other side of, of the table. What what's exciting, Fair enough. Fair enough. Well, what is what is exciting you right now, what is the art of the possible right now? What do you what are you? What are you excited about in terms of? Yeah, you know, what I think we're gonna get there.
Yeah, and, and so, you know, we've been following the I like applications of AI, where they let you do things that you couldn't do before, right? So as opposed to sort of automating something really enabling things that you couldn't quite quite do before. Now. Now, you can argue sort of the line between automation versus innovation, and it's very much a fine line.
So I like that distinction. That's an interesting way to look at some of the stuff that's coming out.
Exactly right. And so so you could kind of think about something like, you know, generative art, generated art as an innovation thing. But if you are a designer, then it's very, it potentially is a little bit scary, where it's an automation thing, right? I was just kind of reading this story a few days ago, where, you know, some designers spent many, many hours designing, I believe it was an art for a book cover. And the client kind of went, Oh, we're not going to pay you for this, because this looks like aI generated art, you didn't actually do something. Right. And that's, that's, that's kind of scary. But anyway, your question was, what's exciting? I mean, the ability to do that is very exciting. I think where the challenges are, is how will they actually get used. Do we as business do we as society understand the both the good and the bad implications of it enough to really sort of treat it with potentially some care that that it needs? What feels like in the last six months is a lot of just very rapid deployment and usage without potentially a lot of sort of thought into should we be doing this right? Like, is this? Is this the thing we really shouldn't be trying to do?
the Jurassic Park quote, as to you were too busy as to figuring out whether or not you could you didn't think about whether you should?
Exactly right. And I've seen sort of a bunch of examples. On another paper, we recently seen about, essentially, ability to fake someone's voice with just one or two voice samples. Whatever you need to be able to say, historically, that took a lot of data. Now it takes almost no data. And so could someone kind of call me on the phone. You know, I'm just kind of saying, Hello, who is this?
This is my colleague, Chris. We've just leveraged some of our tools, a couple of screenshots. And it's fine. This is that's what's going on here today. But all jokes aside that's not far away.
Exactly. Right. So So I think, on one extreme, we have very low hanging fruit where we still haven't deployed AI. And at the other extreme, we're kind of saying, can we start to deploy AI in these potentially very interesting scenarios. Right. So So I think the idea idea that there has been sort of this notion of, oh, you know, software developer jobs are safe for designer jobs are safe for creative jobs are safe or leadership jobs are safe from AI, and now I think we would have to kind of say, you know, maybe never say never right that in terms of AI is never going to to impact this, and when I sort of talked to my little nephews and nieces about what they should be, because a lot of them went, I want to learn programming. And I'm kind of wondering, wait a minute, what does programming look like, in 510 years, and it's incredibly hard to predict. So it's not just designers, for there's been a lot of innovation. It's been in things like programming to write like AI will come from
generative coding was something it's funny, because we're all I think I've seen memes online, which individuals are laughing at the artists are laughing at the musicians or what have you. And then the coders are like, wait, wait, wait, wait a minute, you're coming, you can do this as well. That's, that's something that's that's in, you know, we talked about the the arts and things that may not be I don't want to say not useful, but things that you know, towards beauty and culture, but coding the ability to, to, to have AI generate software would be, I imagine pretty powerful.
Yeah, and it's, it's already in some ways out there, right? Or at least the beginning trends of that it's actually deployed by even large companies like Microsoft, right? Like they have products where there is an AI that's helping you write your code or actually writing it for you. And sort of one, one sort of idea is, you know, what, what does this do to our broader work environment, society? What are the ethical implications of this? And and I'm by no means an expert in this, but I'd at least say that, I don't know if enough questions are being asked, and enough thought is being put into, as you said, I don't think people asking should we do this, it's a lot of like, I can do this. So let's go do this.
Right, right. And it's always, it's always tough for legislation to keep up and no one's silly, and not quite the same thing. But I remember I was visiting family in Los Angeles, and there was just these scooters everywhere, not AI by any means. But regardless, just the point of the technology develops so quickly, and there was no legislation around it, and the people just leaving them left, left, right and center. And now with things like with with the generative AI and new technologies that are coming up, it's interesting how it happens so fast, and we don't know what to do with it. And lawmakers are left in the lurch trying to catch up.
I mean, I would say even AI, people are all catching up, because just the pace of innovation and the breadth of innovation, right? It's hard to kind of picture government and lawmakers keeping up. And I mean, the fact is, even when they kind of try, I don't think that they are, they're kind of thinking it through. I've heard and I haven't sort of, but but I was recently reading that China has implemented laws around like, you know, who should be using AI generation tools, and any time that that AI generated artwork and things like that are used, they need to disclose that right. So so I mean, I think I think different countries are approaching this differently. Canada is obviously working on some regulation. So it's the US so is the EU. And in some ways, I think that the regulations updated pretty quickly. Either it's too restrictive, it's not recognizing that this is still an area of innovation, but at the same time, in trying to sort of protect the threat, it goes very much either in one direction where it doesn't recognize it, another direction, where it's just kind of true, you know, painting everything with a broad, broad brush, because, you know, a company deploying an AI system, a computer vision system, that's helping them figure out whether the widget that they're making is good or bad, you know, quality control, right, of course, whether something like that should be regulated is not clear to me, as opposed to something like AI generated art or text or code. That's probably different, like so. So we need to treat these things differently.
It's interesting that someone like you would even I wouldn't say struggle, but there's some thought being put in and folks in this space that are actually working day to day in AI don't have all the answers. It's a bit, it's a bit concerning, you would hope that those those in charge of regulating would be pulling in, you know, the right, the right brain trust in order to make those types of decisions. I've definitely you've seen on the internet folks that were regulating the internet in the early days and didn't understand, you know, basic browsers or things to that degree, it's it can be a little a little frightening in terms of that.
I mean, there is an argument to be made that the lack of regulation of early internet is what really let it flourish, right. It is one of the it is one of the very few areas where there is little to no regulation which is allowed it for greater innovation. Now, I'm not saying that's the right path. This is definitely beyond my skill, sort of how to regulate, but as someone who likely the regulation will apply to AI, I do believe there's a lot of questions. And there's a lot of unanswered or non answers to those questions.
On that end, like what do you think folks or leadership companies folks in your position are struggling with in terms of, or the folks that you work with, I should say, that are struggling with in terms of adopting AI? I know I have my questions, and wonderfully, but I can lean on a set of, you know, a team of experts to kind of calm my fears and get us where we need to get to. But what do you think you're the what have you seen in terms of struggling with adoption?
I mean, there's many factors, right. So even the things we just talked about, you know, do I know, will it work? Right? Should I be investing in this? Should I be building this in house? Or should I be just using an API? Really? What is my, you know, what is my strategic decision here? You know, how do I try this so that I know whether it's going to work for me or not, you know, customers, I mean, there's a lot of fun out there, there's a lot of misunderstandings out there. So I think there is a little bit of a fog there that people need to clear. So I do think people struggle with that. And, and my recommendation, there tends to be, you know, I think the only way to kind of clear the fog is to let the sun sign in, which is, you know, gain the skill set, whether by learning about it, or by talking to people about it, whether bringing someone on within your organization who can help you kind of get through this. I think I think sort of that skill set aspect of it is a big one. And I think for people who are struggling, I really need to get started in some ways. I really kind of talked about the fact that yes, you want a solution or an AI system that's either going to sort of generate revenue or reduce costs or reduce risk or something like that. I think the first goal should be to really maximize learning, right? It is a very new area. I want to be clear, it's not that complicated, right? Like it is data science, not rocket science, right. I do think that that that approaching it at least for the first little while, it may be weeks or maybe months, but but a little while, with the goal of let's maximize learning in this area, as opposed to try to maximize profits, I think goes a long way in leading to a longer term success because it is a —
Do you find it that most people easily buy into that concept of learning first, and we'll get to revenue? But we need to focus? Or is that a challenging conversation that you're having? I imagine in terms of the conversations we're having with large enterprise companies, that might be a tough one we're not doing this for fun, might be something that I'd be met with? How are you finding that conversation?
It's almost always a very difficult conversation. I don't know if anybody really fully buys it. But we've seen it again and again that that approach leads to a better outcome. So what I kind of like to think about is, when you approach it from a maximize learning perspective, what you end up doing, at a very basic level, you can do a lot of things, at some very basic level, you kind of look at all the data that you have. And in some ways you have to kind of fall in love with that data, right? Like you really have to spend time understanding what it is, and what you have, what you don't have, and where the gaps are or insights that lay within it. And those are the ones that drive the value. So in some ways, this kind of boils down to the notion of trust the process, right? So the process we're talking about here is this notion of maximize learning, faster feedback loops, fall in love with the data. And if you kind of really stick to that and trust the process, it does lead to innovative and by innovation, I actually mean valuable, right sort of profit generating ideas and opportunities, versus just kind of trying to chase that. It does feel a bit counterintuitive.
No, I adore that.
That's the innovative process, right? You have to trust the process, as opposed to sort of chase the outcome.
No, I think that's such an excellent way. You know, if you're dealing with leadership that might not have, you know, an engineering background, if you can, I think that's so succinctly put in terms of falling in love with the data and trusting that process we'll get there but we do really need to understand and build that foundation, such that you can enter you can I mean, you can phrase it in X number of ways in which it will be receptive to that type of, of a persona, but I but it's the truth. I love that that that concept of falling in love with the data.
And so, you know, early on when you kind of asked me, Hey, what should the companies be doing? I kind of talked about skill set, right sort of clear the fog and learn what this is all about. These last points we've been talking about, I really kind of like to sort of bring the other point, which is the mindset, right? Do you have as an organization as aleader, the right mindset to be successful about this, because, again, if you approach this as a typical sort of traditional project, or even traditional software projects, you may succeed. But more than likely, I don't think you will get the optimal result. Right. So I think recognizing how AI development or AI products or solutions are different from typical software goes along the way. And having that mindset, it's much more iterative, much more experimental. And experimental doesn't mean it's gonna fail a lot, right? Like, it doesn't mean you're going to waste a lot of money. I think you have to be deliberate in how you experiment. Right? And then minimize your downside risk. I mean, this is, this is what what what we like to work with our customers on is how do we get you started with success? Like, how do we get you started, right so that your first project, first idea that you work on, leads to a good outcome.
I love that and it's in I imagine, you know, folks might kind of be like me, in that they've, I've come from a software background, by no means was a developer, but I was kind of on the implementation side and, and follow those processes. And it's, it's funny you're right I kind of have to unlearn a lot like a decade's worth of process. What do you mean? Why should software software, I mean, I guess at the end, there's some nifty algorithms happening. And some models, but really software is software. And that's, that's not the case in terms of from the conceptualization phase, all the way to production, to get into production. That's interesting.
For sure, there are a lot of similarities. I mean, there are more similarities than there are differences. But think about the differences and treated just as a, you know, as if you're building an app or a website or an enterprise software, it likely leads you down some weird sort of decisions, right. And this is where you kind of see companies have, you know, hired a bunch of PhDs given them a floor and say build some AI coolness, and struggle to to ship right, like, how do you actually get something valuable in your customers' hands? Right? That that that process needs to be there needs to recognize that it's different. It can't be, it can't be treated as it's so special, that you're going to, you know, let you do whatever you want. And it can't also be just be like, Oh, it's the same as before, it really kind of is different in its own way.
Right. It's an interesting balance. I imagine again, I keep thinking about our clients and the decision makers and leadership in terms of on the the data science side of the data and engineering side and the and the AI side, as well as leadership, you do need to have that balance of that understanding and everyone being kind of on the same page and allowing people to do what they're supposed to do. But yeah, that it's funny, I think it's, it's almost a less of an asset to come from kind of that old world of software. You're right. I know we've experienced this, and you know, you you're speaking with the leader that I've got 15 years of X experience in software, that might be you might be kind of, you know, stuck in the muck of and not thinking about it in the way that you're thinking about it. I hadn't heard it put that way, it's, it's very clear, I'm just thinking about this, these types of things that I'm talking to my team in which they're having initial conversations with some folks, that might be a bit challenging on the enterprise side, I think that's, that is, that is a really cool way to look at it. So we've spoken a great deal in terms of how to be successful. I think in terms of mindset, I love the phrase falling in love with your data. I'm such a nerd, which I think is and then trusting that process in addition to that, and that and that mindset, kind of on both sides of the organization. But I know you must have had challenges that you've overcome when working on either research internally or with with it with client projects. I'd be so curious to know, what are the what are the what are those challenges that that you and your team have been able to get through?
Yeah and I mean, there is a very sort of common, obvious, unsurprising sort of answer to that, sort of, you know, every single project, right, there's 100% hit rate, has had a data problem, right. No data, insufficient data, unlabeled data, quality, missing data, inaccessible, sort of just a variety of problems related to data.
But if you don't have the data, you don't have the data like, how do you how do you get around that? I'm so curious.
Yeah. And I mean, there's a bunch of ways, right? Can you buy? Can you can you generate this data? Can you infer this data? I mean, this is where the creativity, comes in right? Where are you able to, you know, bootstrap your way to some product or some initial solution, which then lets you gather more data, right. So I think there's a bunch of ways to solve these types of problems. But what I would say really is, it's best not to have this problem. And I think people sort of thinking about data as an asset. You know, people kind of keep thinking about their people as an asset, their customers, their assets as an asset. But I think if you start thinking about your data as an asset and treat it as sort of the high value commodity that it is, I think you would not have these challenges. So I would say, broadly speaking, and I would include us in this right, like, we all sort of under invest in our data, even though the fuel of the future.
Yeah, no, that's really interesting. I think I've had conversations in which at least with some larger organizations in which they're trying to monetize their data, they don't know where to start, you know, what's exhaust? What's What can you what what's valuable? And that that's been a really key conversation with that we have.
That's exactly where I would kind of go like where I would go to what you just said, which is, I think that's the mindset conversation we were having. Right? The goal that they're starting with is I want to monetize my data. As opposed to kind of going back to the process, which is let's understand the data. And then see what the data tells us. Right, as opposed to us kind of saying this is what we're going to do, and then hoping the data fits, fits into that.
Right, right. No, I guess I can be guilty of it. Absolutely. That is interesting.
I mean, I mean, it's the natural way to approach this right. And it's not a bad way to approach it either. It's just a when it's not working. We find that the other way, is a little bit weird. But more powerful.
Yeah, well, I would hope that folks in that position like me would just listen to the other person on the other side of the data. And again, like you said, Trust, trust the process. So again, you've clearly got clients that have fallen in love with the process, have fallen in love with the data, have overcome challenges. What are some any examples whether either you've come across in other, you know, you've read about or not necessarily within your, within your, your world? But like, what are some examples of projects that weren't successful, they just they just didn't get it? What happened? And anything that you've, you've seen recently?
Yeah. And I think what it what it ends up being almost always is the is the lack of clarity around problem definition. Right. So what we're trying to do, right, so So because I think I think that lack of clarity then shows up in how you make decisions throughout the development and say, sort of launch process. Right. So either the sort of the problem ends up being a bit premature, right. So there's sort of a are you coming at the problem from the data or are you coming at this problem from from intuition. And so I think, sort of knowing what you're trying to accomplish and why you are trying to accomplish, or I suppose I should clarify, not knowing those two things. That really ends up being the root cause in many cases, aside from the data, right, so we've talked about enough about data. So I'm not I don't want to talk about data anymore. If you don't think about the data. That ends up being often the root cause of things aren't going well.
Right. Right. I think that's that's absolutely good advice. I've, again, in terms of I like to book in these conversations kind of in in a similar fashion, again, for the same reason that I do get, I do get differing answers. And you're, you're a bit of a serial serial entrepreneur, maybe maybe fair enough, not not your first rodeo. I'll put it that way. But I would, I would love to take the opportunity. And I think some of the folks that listen are just hoping to be in your position someday or maybe in early stages of that path. Any advice you would give to some of those aspiring founders?
I certainly do not feel I can give advice to anyone. I mean, I think that's a real —
We will argue about that later. But please indulge me.
But I mean, what I will say is, from my perspective, right, so it's not necessarily advice as much as I think, why, why we do what we do, right? Yeah. It's, you know, I really get to love to do what I do. And I think some people say, you know, don't turn your hobbies into, into something you work on. And don't, don't, don't work on things that you love, but I think it works works, at least for me. And to like, I mean, have fun with your people, right? So take care of people, caring about your team, your customers, right? Caring about the community within which you operate, particularly your team goes a long way, right? Because I think I think without them, you can't really do especially the type of work that we do. We need we need people who are significantly smarter than me. You want to have fun with them, right? Otherwise, what's and you want to be learning from them? Otherwise, what's what's what's the point? So So I would say take care of your teams and and yourself.
Yeah, no, absolutely. When you talk about taking your hobbies and turning into you know, and that becoming your career, I imagine starting off with an AI driven pool table is is a pretty great way to it's a good is a good example of walking the walk and talking to talk. Lastly, on this again, I'm wondering, I appreciate your time today so much and I've taken up too much of it, what is on deck what is what is upcoming for gradient ascent AI in the coming year, anything interesting that you you can share with us?
Yeah, I mean, you know, we, we continue to want to solve interesting problems with our customers and especially problems that are at the intersection of data and math and design, right? Because that's, that's really where we thrive. But one of the things I'm very excited about this years, sort of, you know, is working more and more with partners, right? So partners such as yourselves, right, so. Being part of larger teams, and, and collaborating more with other people. So I'm pretty, pretty excited about that.
That's fantastic. You know, Monish, I super appreciate your time I was dying for this conversation. I think in our early early conversations, as we first met, I'm like, Oh, I'm gonna need to pick pick this brain in particular. And And typically, I think, any of the last podcast, I cheat, I just pick individuals that I just really want to talk to you for for a decent amount of time and learn a little bit from so I can't tell you how appreciative I am of your of your time today.
No and I'm excited about the time we spent in getting to getting to sort of discuss and debate these ideas. And I mean, there's, you know, we're still at the beginning of this so so it's great to learn and share.
Yeah absolutely. And I imagine we will continue to do this I imagined not our last conversation by by any stretch and all the best in the new year and hoping to many more conversations just like this.
Thanks tomorrow. All right. Thank you, sir. Enjoy the rest your day.