Latest / CXOTalk / Building an AI Startup: What's Different in 2025? | #871
Transcript
- 0:00We're discussing AI startups today on CXO Talk episode 871 in
- 0:07a conversation with Arvind Jane, the founder and CEO of Glean.
- 0:13He was a distinguished engineer at Google before starting and
- 0:18taking public cybersecurity company Rubric.
- 0:22Glean has raised $600 million and is valued at almost 5
- 0:28billion. Glean is an enterprise AI
- 0:30company. Think of it as Google or Chat
- 0:33GPD but inside your company. So that's what we do.
- 0:37And you are a, shall we say, born in AI or AI native company.
- 0:44So you're the perfect person to help us understand what's going
- 0:49on with start-ups, AI start-ups today.
- 0:52Absolutely. We are in fact I think the first
- 0:55company to bring the transformer technology to the enterprises
- 1:00and, and and say, yeah, it's a lot of great learnings, you
- 1:02know, starting as a native Gen. AI company six years back.
- 1:05Gen. AI company for six years.
- 1:08That's well before the marketing hype surrounding ChatGPT and
- 1:12Open AI. That's true.
- 1:14In fact, the term Gen. AI did not exist at that time.
- 1:17But the core technology that powers Gen.
- 1:19AI, the Transformers, they were there and we were using them,
- 1:24you know, in early 2019 to, you know, understand enterprise
- 1:29data, knowledge, information at a really deep level using AI and
- 1:33then making it all searchable for people inside a company.
- 1:37Can you describe the differences between AI startups and
- 1:43traditional technology startups? Every company like, you know,
- 1:48first of all, whether you're a startup or you are a mature
- 1:51company, I think AI is becoming such a core fundamental tool
- 1:55that you've got to use in, you know, in products that you
- 1:58build, because that's, that's the way to sort of stay ahead.
- 2:01That's the way to actually build new amazing things.
- 2:04So, so I actually feel like, you know, like this.
- 2:06My view is that probably all startups that you think of
- 2:08today, you know, that you can call them all AI startups.
- 2:13The but, but maybe the other way to look into look into this is
- 2:16that there are some companies, you know, that are actually
- 2:19building core foundational technologies.
- 2:22For example, they're building models, they're building
- 2:25infrastructure to actually train models.
- 2:27And maybe, you know, that's what I'm like, you know, you know, I
- 2:29would say like, you know, one set of companies, but then the
- 2:32vast majority of AI startups that are getting started today
- 2:37are companies that are thinking about solving a business
- 2:40problem, a consumer problem where they feel that AI and you
- 2:46know, the new reasoning and generation capabilities is going
- 2:49to play a big role in, in the product that they're going to
- 2:52build. The but like, you know, like us
- 2:56living in the AI world, you know, all the, all the startups
- 2:59that I, you know, engage with, like I'm not, I've actually not
- 3:01seen a single startup that is not, you know, that that's not
- 3:04actually making AIA big part of the core tech stack.
- 3:07As AI has matured, there's become a greater clarity that AI
- 3:14as you just described must support the solving of some
- 3:18business problem. For AI startups where the the
- 3:25centerpiece is that AI technology, what are the
- 3:31differences from again traditional software companies
- 3:35where yes, they're solving a business problem, but then but
- 3:38the technology is very different the underlying foundations.
- 3:43AI technology, you know, moves very fast, it changes, you know,
- 3:46at A at a very rapid pace. So when you think about product
- 3:49development, the the new AI startups, you know, they're
- 3:54they're able to you know, the first of all, they're very lean.
- 3:56They they are able to actually do a lot of things, you know,
- 3:59because software programming, building systems with AI, you
- 4:03know, these things are becoming, you know, much, much easier than
- 4:06than you know, than in the pre AI world.
- 4:09So, so I think, I think the like, again, like, you know, I,
- 4:14I, I, I struggled with your question because I, you know,
- 4:17like, like my, from my vantage point, like when you think about
- 4:21investment that is happening, you know, most of the investment
- 4:25from venture capital actually is coming into companies that are
- 4:28making AIA big part of their story.
- 4:31And, and so I can compare more with like the, let's say that
- 4:36startups, you know, that we're getting started two years back
- 4:39versus today. The key difference between them
- 4:41is, well, like thinking about this new model where you can
- 4:46truly build products, you know, which are a lot more powerful,
- 4:51you know, a lot more capable than what would what you would
- 4:53have even, you know, you know, thought of like, you know, a
- 4:55couple couple years back. But but but I search like, you
- 4:59know, I think from a company building perspective, like, you
- 5:01know, startup is a startup. You know, we are AI company and
- 5:05like you know, when I compared this with my previous startup, I
- 5:09would say that like, you know, most of what we do doesn't
- 5:11change. Like, you know, we, we still
- 5:13have to think about the core business problem.
- 5:15We have to figure out how we're going to actually solve that,
- 5:17how we're going to build great technology and and like, you
- 5:21know, the teams to to then sort of like, you know, bring that
- 5:24technology to our customers. So so personally, like I have
- 5:27this feeling that yes, like, you know, AI is becoming a big part
- 5:30of our technology stack, but fundamentally how we build and
- 5:33design companies are not changing except for maybe one
- 5:36thing like, you know, you do here.
- 5:37Sometimes companies saying that, well, like, you know, now with
- 5:40AI, you can actually have a one person company that can generate
- 5:44a billion dollars because, and then so you see a little bit of
- 5:47that, like, you know, like, you know, in the new generation of
- 5:51companies, you are seeing some of these, you know, AI companies
- 5:55like scaling up revenue at the pace that, you know, the
- 5:58previous industry of SAS companies couldn't.
- 6:00Like, you know, we have, we saw some, you know, I, I did these
- 6:03examples all the time, you know, like start up actually reaching
- 6:06$20 million, you know, in revenue two months, you know,
- 6:10like after it's got started or a company that's, you know, at 100
- 6:13million rate, like, you know, within the first year, these
- 6:16things are not possible. Like in, you know, in many ways,
- 6:19you know, in the Pai world, But given like you know how fast,
- 6:24like you know, you can actually build products and, and how
- 6:26different your products can be compared to the, you know, the
- 6:29current products in the market. It's allowing people to actually
- 6:33just fundamentally scale at a different pace and level than
- 6:37than the previous generation of start-ups.
- 6:40I just want everybody to know that you can ask your questions
- 6:44right now on Twitter. There's a tweet chat, X on XI
- 6:51should say there's a tweet chat and X chat taking place.
- 6:56Use the hashtag Cxotalk if you want to ask your questions.
- 7:00If you're watching on LinkedIn, just O your questions into the
- 7:04chat. Arvind, you raise a very
- 7:07interesting point just now, how startups, AI startups can
- 7:13generate so much revenue with relatively few people.
- 7:17What is it about AI, this technology and the nature of the
- 7:21startups that are happening around this to enable this
- 7:26phenomenon? So first you can actually build
- 7:29products that are very different in their capabilities.
- 7:36Then like, you know, then things that are out there, when you
- 7:40think about, for example, software development, there are
- 7:45companies, you know, that are offering new development
- 7:49environments, which allows developers to go five times
- 7:53faster than what they could before.
- 7:55And, and so there's a, there is such a big leap in, in the
- 8:00capabilities, you know, that these, you know, that these
- 8:03products bring, which creates that instant, you know,
- 8:06excitement in the market for them.
- 8:08Like, you know, like, you know, these, these new coding tools,
- 8:10like, you know, even even for us, like, you know, as a native
- 8:13Gen. AI company, like we, we see it
- 8:15everyday. Like, you know, people just
- 8:17like, you know, they, they want to use these tools.
- 8:19Like, you know, there is a, like, you know, there's, there
- 8:21is a big, you know, demand, you know, from the ground up, like
- 8:26there's love for these new AI products.
- 8:28That is, that's incredible. And that is what's allowing them
- 8:30to, to actually like launch a product in the market, like, you
- 8:34know, have a PSG motion and, and like, you know, there you go.
- 8:37Like, you know, people come in, they want to use these products.
- 8:40So it's, it's, I think largely it's driven by these brand new
- 8:44amazing capabilities and software doing things that
- 8:48people didn't expect, you know, it could do.
- 8:50So I think I think that's the biggest part.
- 8:52Like that's what plays that excitement and then instant
- 8:54demand. But the other thing which which
- 8:58is actually allowing these companies to have this success,
- 9:01you know, you know, in a short duration of time is that, you
- 9:05know, the software development itself has, you know, gotten
- 9:08very accelerated. When you use these new tools,
- 9:11you can build products which are a lot more capable than, you
- 9:16know, current products that we have in the market.
- 9:19And, and you can do that in a short period of time.
- 9:21Like you actually put something amazing, you know, in a month,
- 9:24you know, in two months, you know, something that used to
- 9:26take you like a year or two years before.
- 9:29So that's that's the other factor that is actually, you
- 9:31know, shrinking these these cycles for these companies.
- 9:36So you have this combination of tremendous interest and market
- 9:41demand combined with significant developer productivity increases
- 9:50as a result of the tools. Is that a correct way of saying
- 9:52it? That's right, yes.
- 9:54We have an interesting question coming from LinkedIn and this is
- 9:59from Santosh Sirivo, who says who is actually using and paying
- 10:06for AI services. AI has captured everybody's
- 10:11imagination. There is no enterprise in the
- 10:13world today that feels that well, like, you know, we can, we
- 10:18can, we can look into AI like, you know, a few years from now.
- 10:23Everybody knows that they have to act, they have to embrace AI
- 10:29today if, if you want to be, you know, if you want to stay
- 10:33relevant. And so, so starting from that,
- 10:36starting from that, like high level desire to actually, you
- 10:41know, bring AI into your enterprise.
- 10:43Now what we're seeing is that there are two different ways our
- 10:46people are bringing AI into the company. 1 is the CIO, you know
- 10:52that, you know, CIOs like, you know, obviously hold the, the,
- 10:57the responsibility to bring the right set of technologies, you
- 11:01know, to, to the entire working population of your, of your
- 11:03enterprise. So they are taking the
- 11:06leadership in terms of bringing, you know, bringing godly,
- 11:11applicable, useful products that are AI based to their
- 11:16enterprise. And that's for, for us, for
- 11:18example, at Glean, you know, you know, they are our primary
- 11:21personas too, because Glean as a tool, you know, it's, it's a
- 11:25knowledge access tool. Like people go to glean, they
- 11:28ask questions, you know, we quickly answer those questions
- 11:30for them using all of the enterprise context and data and
- 11:33knowledge. And so that's the tool.
- 11:35That's actually every knowledge worker, like, you know, whether
- 11:38you're engineer, the support person, somebody in HR or IT or
- 11:42sales, all of you have the need for that.
- 11:44All of us have questions, all of us have tasks that we think AI
- 11:47can do for us now. So it's a broad tool and CI OS
- 11:51are the ones who actually typically like, you know, will,
- 11:55you know, purchase a a company by two like that.
- 11:58But then you also have every individual functional leader.
- 12:03You know, if you are head of support, if you are the CTO and
- 12:07you're trying to actually make sure that you know, your
- 12:10engineers, your developers are productive with AI.
- 12:12If you're, if you are a sales person, like you know, the sales
- 12:16leader and you want to make sure that you're using modern ways
- 12:20to, to prospect, to reach out to your customers, to actually have
- 12:24powerful engagements with them. Function by function.
- 12:27You're seeing every functional leader in the enterprises
- 12:31looking into and evaluating AI tools and bringing them on
- 12:34board. So this is a, this is a very,
- 12:36very broad phenomena, industry wide vertical, wide
- 12:39geographical. Like, you know, I've been
- 12:42travelling a lot across the world.
- 12:44I don't, I don't, I don't see any company that is not paying
- 12:47attention to AI, any country that's not paying attention to
- 12:50AI either. So this is, you know, like is,
- 12:52is everyone really is the answer like you know, is bringing AI
- 12:55into their into their enterprises.
- 12:57Subscribe to our newsletter Join our community Go to cxotalk.com
- 13:02Subscribe to the newsletter. Check it out.
- 13:06So it's following a traditional enterprise software, the
- 13:11purchase process in a way because as you just described,
- 13:15you've got IT and the CIO and at the same time you have a
- 13:20functional line of business leaders, HR, whatever,
- 13:24marketing, whatever it might be. Both of these groups are looking
- 13:28at AI products. And I'm assuming this comes
- 13:31right back to what you were saying earlier, which is the
- 13:35core issue. It's not actually the
- 13:37technology, it's the business problem that's being solved.
- 13:41We have a very interesting question from LinkedIn and keep
- 13:47your questions coming in. We have some questions on
- 13:51Twitter as well, and we're going to get to all of these.
- 13:54Risha Varshney asks, how do your offerings address the AI ethics
- 14:01jailbreak, prompt leakage and AI transparency slash
- 14:07interpretability risks? And I'll just mention that Risha
- 14:11is Senior Director of Risk Systems Development at Freddie
- 14:16Mac. And of course, Financial
- 14:18Services is keenly interested in these topics.
- 14:21So Lean is an enterprise AI solution.
- 14:25We work with the largest enterprises out there in
- 14:29different industry verticals. We work, we work a lot with
- 14:32financial services and these risks are all very real.
- 14:35Like, you know, I think with AI, the, you know, it's a very
- 14:38powerful technology And, and it's also like a, it's, it's,
- 14:43it's, you know, it's grounds for, you know, innovation and
- 14:46like folks who are trying to actually create security
- 14:48problems and, and you have to be very, very careful in terms of
- 14:52rolling the technologies in the right way, in a secure way.
- 14:56And also like, you know, in a way which basically ensures
- 15:00that, you know, the technology is unbiased and, and, and it's,
- 15:05you know, it's doing stuff that you think is, you know, safe,
- 15:08you know, for your, you know, in your context, in your
- 15:10enterprise. So some, I'll give you some
- 15:12examples of like, you know, things, you know, problems that
- 15:14you have to solve on the security front.
- 15:18One of the key, you know, things that AI is that if you're going
- 15:21to make it useful for your enterprise, your, for your
- 15:25organization, you know, you have to take these language models
- 15:29that are built using the public web data.
- 15:32They don't really know much about your enterprise and your,
- 15:35your, you know, you know, your information, your data.
- 15:40And so somehow you have to figure out how you're going to
- 15:42actually connect that enterprise context with the power of these
- 15:46language models and, and do it in a safe and secure way.
- 15:50For example, if you train the model and you connected all of
- 15:53your enterprise data and knowledge, you know, with that
- 15:57model and now anybody can go inside your company and ask
- 16:01questions, well, it's going to leak a lot of sensitive data to
- 16:04people who should not have actually, you know, have access
- 16:07to that information. Because enterprise data and
- 16:09knowledge is very, you know, it's it's private in nature,
- 16:14like, you know, like a given document, like, you know, only a
- 16:16few people may have access to it inside the company.
- 16:19So you can't take like, you know, our content wholesale,
- 16:22like you know, in your enterprise and just train or
- 16:24build any I system with it. Any AI system that you build, it
- 16:28has to understand who's using, you know, that that particular
- 16:32system or software. And how do you make sure that
- 16:35you know, AI only uses information that you know, this
- 16:39person, you know, is entitled to is, you know, allowed to use
- 16:42inside the company and create those safe and secure AI
- 16:45experiences. So that's one of the problems
- 16:46that we solve. Like in Glean, any AI usage, you
- 16:50know, any agent that you build on Glean, you have to sort of
- 16:55use that as an employee and you have to be signed in.
- 16:59And then we will actually ensure that whatever we do for you is
- 17:02done with knowledge and information that you could
- 17:05access to. I would like you to join the
- 17:08Cxotalk community, so go to cxotalk.com and sign up for our
- 17:14mailing list so we can notify you about upcoming shows because
- 17:18we have amazing discussions like this.
- 17:21All right, let's go to Twitter and to Arsalan Khan who says, do
- 17:27you think AI will become a simple plug and will become
- 17:33simply plug and play even for non-technical employees in the
- 17:39enterprise, so the the broader ease of use for the rest of us?
- 17:45Well, absolutely. I'm both hopeful and confident,
- 17:49you know, that that's what's going to happen.
- 17:50I think the true power of AI will be realized when it's, you
- 17:54know, so easy to use. You know, you can actually do
- 17:58things with AI as as a business user, you know, you should not
- 18:04be required to understand, like you know, how to code, how to
- 18:07build systems. You know, be an engineer, you
- 18:10know, you like, you know, as you know, for example, like let's
- 18:12say you are, you are an, you know, you are an, an employee in
- 18:16the legal department and you review contracts everyday.
- 18:19And you know, you know, it's a, it's a process, you know, that
- 18:22takes you a lot of time, you know, and you should be able to
- 18:25just, you know, ask AI. You can, you know, like, you
- 18:27know, and say that like, look, this is, this is, this is how I
- 18:30review a contract. This is the process that I
- 18:32follow. And you should be able to just
- 18:34say that, you know, to an AI system that actually then goes
- 18:37and identifies, you know, that particular business process for
- 18:40you, automates it for you. So AI has to truly become
- 18:43accessible like that with our Asian PIC platform.
- 18:46That's what we do. You know, we are actually really
- 18:49thinking about how to make this technology super accessible
- 18:53like, you know, make, you know, make sure that it doesn't matter
- 18:56who you are. You could be in HR and, you
- 18:58know, in finance and legal and you, you, you know, you don't
- 19:02need to know anything about AI, how it works.
- 19:04Like, you know what, what are language models?
- 19:06Like that's, that's not like, you know, that's not what you
- 19:08need to worry about. You need to just like work with
- 19:10a smart. You're like working with AI
- 19:12should be like, you know, working with a really smart
- 19:15person who you could actually go and get some work to, you know,
- 19:18you just tell them, you know, like how you know, like how this
- 19:21work needs to happen and then AI just makes it happen for you.
- 19:23That's, that's the, that's the model that you're going to see,
- 19:27you know, with any like the agent, you know, AI agent
- 19:29platforms that are going to succeed in the market are going
- 19:32to be of that nature. Like, you know, you have to
- 19:34elevate, you know, like the, you know, the capabilities of these
- 19:37systems, you have to make them more accessible to non-technical
- 19:40users. And the other thing that that
- 19:42that also add to it like to go one more step, you know, beyond
- 19:45that people are not seeking, people are not seeking help, you
- 19:49know, from AI as much as you would expect.
- 19:53Like, you know, we all have habits, you know, we do things
- 19:55the way we do. Like, you know, like there's a
- 19:57lot of inertia in terms of like, you know, thinking about like,
- 20:00well, should I do this task differently?
- 20:02Like people don't think that way.
- 20:03Like, you know, you, you don't have time often.
- 20:06Like, you know, most of the times you yeah, you don't even
- 20:07have time to think about, well, should I change?
- 20:09Like, you know, the way I work and so.
- 20:12It's not only that AI has to be easy and you need to be able to
- 20:17summon it and, you know, get it to do work for you.
- 20:19You know, AI also has to follow you.
- 20:21And it has to come to you and say that, look, yeah, I'm, I'm
- 20:23observing that you're doing this work every day and spending, you
- 20:27know, two hours trying to get this work done.
- 20:29And I can help you with it. So, you know, it has to come to
- 20:32you. And and that is when, like, you
- 20:34know, you really, you know, see, people will embrace the
- 20:37technology at a large scale. I'm looking forward to that day
- 20:40happening because I can tell you I use large language models
- 20:44every day and I use multiple language models and different
- 20:48modes, research, non research and so forth.
- 20:52And so much depends on the model you're using.
- 20:56It seems the time of day, the phases of the moon, how you
- 21:00construct your prompt, and then the whole thing's just a pain in
- 21:03the butt. Yeah, it is hard.
- 21:04Like it's not easy today. And then by the way, you are
- 21:07using like what I would say probably one of the most, you
- 21:11know, one of the most accessible tools, like, you know, you're
- 21:13just talking to a system like in natural language.
- 21:16Yes, you have a few knobs to select hidden there.
- 21:20So it's, but it's going to get easier like, you know, and
- 21:22that's what like our job is, you know, at clean, like one of the
- 21:24things that we do is so we're not an LLM company, We're not,
- 21:28you know, building and training these foundation models.
- 21:31But we are like, our goal is to see that you as a business user,
- 21:35how do we make sure that all this innovation that's happening
- 21:37in the industry, how do we make it more accessible to you and,
- 21:42and make things easy, you know, as seamless as like, again,
- 21:45like, you know, the working model for me with AI is, well,
- 21:49AI is like a smart human that's been in your company from day
- 21:52one. They know everything about your
- 21:54company. They know all the people,
- 21:55they've read all the documents, all the, all they've been part
- 21:59of every single meeting and they're ready to now help you
- 22:0124/7. Just ask them what you want and
- 22:04they do it for you. Like that's, that's the, that's
- 22:05the right model for when you know AI is going to deliver true
- 22:08value in enterprise. Greg Walter says glean looks to
- 22:12be the glue or connector between disparate databases.
- 22:18Do you see a future where this function is no longer needed?
- 22:22Enterprise environments are so complex today.
- 22:26You know, any large enterprise, you'll get like, you know,
- 22:29thousands of systems, you know, the data depositories,
- 22:33databases, you know, unstructured, you know, data
- 22:39depositories, you know, documents and the like.
- 22:43The true magic of AI happens when you have access to all of
- 22:47that information across all of these different systems.
- 22:51Maybe like, I think things are actually going to get simpler in
- 22:54the sense that, you know, AI is going to get smarter and
- 22:57smarter, like in terms of, you know, being able to connect with
- 23:00all of those different systems over time.
- 23:02And, and a lot of things that we do today at lean, like, you
- 23:06know, which we have to actually do a lot of hard work for to
- 23:08actually connect with these different enterprise systems.
- 23:11Like we do expect, you know, it to get easier over time.
- 23:14So yes, like, you know, some like, I think that's the, you
- 23:18know, if we're, if we, if the answer is that like, you know,
- 23:20the complexities are not going to go away, you know, then I
- 23:23think he is not doing his job. Dave Brace on LinkedIn asks, is
- 23:29it likely that non deterministic agentic systems can honestly
- 23:34prove that they are truly trustworthy in the enterprise
- 23:39and that business leaders can trust products like Glean to
- 23:43make thousands of business decisions every day?
- 23:48AI is non deterministic. It can also be wrong.
- 23:53It can hallucinate and to some degree, like, you know, humans
- 23:59are also like that, like, you know, if you, you know, if you
- 24:01have questions and you go and ask somebody, like sometimes,
- 24:04you know, they don't have the full context, they're going to
- 24:06give you an answer and it may be wrong or it may be incomplete.
- 24:10And, and so I think the, it's, that's, that's, that's the
- 24:13fundamental thing to remember that, you know, these AI systems
- 24:17are actually more like humans, you know, and, and less like
- 24:21machines, you know, that they used to and, and now you have to
- 24:25figure out like, well, how do I use this technology?
- 24:28Like this is not perfect, is going to make mistakes.
- 24:31So how do I trust it, you know, with, you know, my mission
- 24:34critical processes where precision is actually, you know,
- 24:37absolutely required. And so, so there are a few
- 24:40different strategies that I would say that like, you know,
- 24:43you know, as an enterprise, you can think about there's a lot of
- 24:45work where you don't need precision, you know, you know,
- 24:48work where you need creativity. And that's where like this
- 24:51technology is already really, really well suited to do.
- 24:54But then when you think about like your task with that require
- 24:58precision, AI can be actually used in a few different ways. 1
- 25:02is that, let's say that there's a business process and now
- 25:05you're going to agentify, agentify that business process,
- 25:08you're going to automate it. You're going to ask AI to, to
- 25:12understand that business process and come up with a plan with the
- 25:15workflow to, to actually execute that business process from now
- 25:20on. And so when you, when you use AI
- 25:22in this fashion, like, you know, you will go and you know, ask AI
- 25:25to actually build that agent for you and like, maybe to make
- 25:28mistakes and but you know, you should, you should be there.
- 25:32Go and supervise it like, you know, go and like ask it to
- 25:34tweak, you know, it's work, you know, go manually fix, edit it.
- 25:38And ultimately in that collaboration with AI, you
- 25:42actually encode and build that agent.
- 25:45You know that workflow. But now this workflow is
- 25:47deterministic. Like you, you, you put some
- 25:49investment in it. AI helped you like, you know,
- 25:52build this workload very quickly, but you were totally in
- 25:54control, you were monitoring it. You've gotten into a coordinate
- 25:57into a place where now, now as I said, it's deterministic.
- 26:01And now this business process can actually run and it is OK.
- 26:04Like you know, you don't, you don't need full automation.
- 26:06Like you can actually work with AI, you know, and spend like you
- 26:09know that, you know, initial time like an hour or two, but
- 26:13now you're going to have like, you know, this automation for
- 26:15years because and it's no longer non deterministic.
- 26:17So, so you don't think about it like the technology allows you
- 26:20to to know great things and you don't have to rely on AI to
- 26:25actually make complex decisions. You know, behind the scenes you
- 26:29can actually work with it that initially and and build
- 26:32deterministic systems with it as well.
- 26:34So you're describing essentially the role of that phrase.
- 26:39We often hear the human in the loop and what's the appropriate
- 26:43relationship between the the the person and the AI system that at
- 26:50this stage of development is a tool rather than I'm looking at
- 26:55LinkedIn and Greg, Greg Walters says the true magic of AI
- 27:00replacing old standard applications.
- 27:02Let's take some examples. You know, we initially were
- 27:05talking about a legal person that reviews contracts.
- 27:09Now, if you get 100 page, you know, you know, agreement,
- 27:14customer agreement, and it's going to take you like, you
- 27:17know, a week or two weeks to actually go and review this and
- 27:21make sure that you know, all the terms and conditions, you know,
- 27:24meet, you know, your enterprises requirements and you're to
- 27:27redline this document. And and you didn't say that
- 27:31like, you know, well, I don't trust AI to do this for me.
- 27:33Like this is pretty sensitive stuff, but well, you know, get
- 27:36AI to do the first version of the redlining and you know, it's
- 27:39going to do a great job. Like, you know, if you just tell
- 27:41it like how you do it, you tell that to AI, it's going to get
- 27:4490% of the way there. And if it gets 90% of the way
- 27:47there and now you can actually, you know, like fine tune that
- 27:51and finish that work, you know, with the context and know how
- 27:55well, like, you know, that two week task, now it's actually,
- 27:57you know, a one day task for you.
- 27:59And that's, that's, that's big. Like, you know, you don't have
- 28:01to, you don't have to actually like aim for 100% automation and
- 28:06like, you know, remove yourself from the task completely.
- 28:09Like, you know, there's, you know, we can, you know, I'm very
- 28:11happy if I get 90%. You know, it's a big impact to
- 28:14the business. This is from Gersheron S on
- 28:17LinkedIn who says which agents are driving the most value in
- 28:23large enterprise functions. Even better if you have examples
- 28:28in an industrial AI context. And I should mention that
- 28:33Gershwan is an AI product manager in metals mining.
- 28:38So which agents are driving the most value in enterprise
- 28:41functions? Top use cases for AI today in
- 28:46the enterprise are the following three.
- 28:50Number one is general knowledge access and assistance.
- 28:56So you know, you are a knowledge worker, you, you may be in, you
- 29:00know, healthcare or financial services or you know, industrial
- 29:03sector and you have questions, you know, you have questions
- 29:07that you need answers to. You have information that you
- 29:09need to do your tasks and, and you use AI to actually, you
- 29:14know, help you with that. So like tools like ChatGPT or
- 29:18tools like clean, like inside your company that are just
- 29:21general purpose, they're not meant for a specific use case.
- 29:24They're basically knowledge tools like, you know, they help
- 29:26you, you know, they make knowledge accessible, you know,
- 29:29from the world, from your enterprise to you so that you
- 29:33can move faster with, you know, with your tasks.
- 29:35That's the number one use case for AI today.
- 29:37And it's not surprising because like this whole revolution was
- 29:40catalyzed by chat GPD. And so that's, that's the,
- 29:43that's the application that comes to everybody's mind when
- 29:45they think about AI. The the second use case is, I
- 29:49would say for, for software development, which is around
- 29:52code generation, like developers tasks, like how you build
- 29:55systems, you know, right, you know, develop technology that's
- 29:58fundamentally changing with AI. So that's, that's a very
- 30:00powerful use case. Lot of correction there.
- 30:02And then the third one I would say is, you know, is around
- 30:04service. When I say service, I mean like,
- 30:08you know, I know folks who are actually like taking questions,
- 30:13taking, you know, complaints tasks from their customers or
- 30:17their internal employees. So these are like customer
- 30:21service teams, IT, you know, internal IT help desk, you know,
- 30:24HR help desk, folks like who are servicing other people's
- 30:28requests and demands, like, you know, you know, taking, taking
- 30:31your company's data and knowledge and automating a lot
- 30:34of that interaction with AI. That's the, that's the big use
- 30:37case. So functionally those are the,
- 30:38those are the three top applications and, and, and I
- 30:43think like across industry verticals like those, that's
- 30:46what we're seeing. So like I'll give you some
- 30:48examples on customer service. So we have like these large
- 30:52telcos, you know, that have like, you know, 50,000 or even
- 30:54100,000 customer care agents day in day out.
- 30:57Like, you know, they are getting questions from their customers
- 31:01that they need to quickly answer.
- 31:03And you know, if you, if you make those transactions twice as
- 31:06fast, like, you know, that's, those are hundreds of millions
- 31:08of dollars of savings, you know, for those teams.
- 31:10So like that's a very powerful use case for AI today for for
- 31:13glean, you know, we already talked about software
- 31:17development, like, you know, like, you know, we have large
- 31:20enterprises in industrial sector, in retail, you know,
- 31:24where you are fundamentally changing how, how you use AI to
- 31:30build systems faster, to test them to, to review, like, you
- 31:34know, code to troubleshoot. These are some of the, you know,
- 31:37key use cases as well, like, you know, for, you know, across the
- 31:40industry. This is, again, from Arsalan
- 31:43Khan on Twitter. He says AI might be able to
- 31:47automate standard operating procedures, but what about the
- 31:52institutional knowledge that resides in people's heads?
- 31:57And he says, isn't this a significant job security issue?
- 32:03Today AI like even with Glean, we are able to actually tap into
- 32:07a lot of their institutional knowledge to then create these
- 32:12powerful like, you know, GPT like experiences where you can
- 32:15come and ask questions and you can answer that using, you know,
- 32:18that institutional knowledge. Then think about institutional
- 32:21knowledge like it's actually present in a few different form
- 32:25factors. You know, you have documents
- 32:28inside your company, you know, where people like, you know,
- 32:31write stuff, you have ticketing systems, you have like
- 32:36databases, you know, CRM systems.
- 32:38So there's a lot of wealth of knowledge inside each one of
- 32:41these different systems. But then there's also a lot of
- 32:44knowledge and communication tools like, you know, e-mail or,
- 32:47or Slack. And there's an, an increasingly
- 32:52like, you know, enterprises are changing behavior so that they
- 32:54can capture more and more of that institutional knowledge.
- 32:58One example of that is that, well, like, you know, if, if two
- 33:01people are going to actually talk and have a meeting, you
- 33:03know, record that meeting or at least, you know, like capture
- 33:06the summary of, you know, that what happened in that meeting
- 33:09and make it available to AI so that, you know, it can be, you
- 33:12know, tapped into in the future. So, so I, so I think like so, so
- 33:17that's, you know, these are all the things that you can do
- 33:19today. Like, you know, we, we capture
- 33:21all these forms of institutional knowledge to then actually make
- 33:24that knowledge work for you as an individual and help you.
- 33:28And, and like, you know, is driving that.
- 33:30Like now people are also more motivated to, to actually
- 33:33capture this information more in our company, for example, like,
- 33:37you know, we now record like every non, you know, one-on-one
- 33:42meeting. Like, you know, I could say, you
- 33:43know, confidential meeting with between the employee and a
- 33:46manager. We won't.
- 33:47But like, you know, if it's about, if it's about, you know,
- 33:50a technical discussion, it's about like getting some work
- 33:52done, you know, we'll typically record those meetings so that
- 33:55like all of that data is available to AI in the future to
- 33:58help us. So that's but but now coming
- 34:01back to your second question on, well, does it actually, you
- 34:04know, create an issue with job security?
- 34:07Like, you know, we, you know, I believe that, you know, the like
- 34:10as an individual, the strategy for you to make sure that you
- 34:14stay relevant is you don't go and learn AI like you don't like
- 34:18this. These tools are amazing and
- 34:20don't think of AI as something that's going to take a job away
- 34:23for you from you. I don't think, you know, AI is
- 34:26that powerful. I don't think it's going to do
- 34:27it for most people, but you'll certainly like, you know, lose
- 34:31edge against somebody else who knows how to use these AI tools
- 34:35and work faster and better than you.
- 34:36So, so I think that the, the, I, I think that what, what as
- 34:41individuals, what we need to do is like, you know, like, like,
- 34:44like, like, like this have been always like that.
- 34:45Like you know, when new technologies come, folks who
- 34:47embrace them, like you know who quickly try to learn them, are
- 34:51the ones who come out ahead. I think this is a very important
- 34:54point and I certainly give people that same advice that
- 34:58you, you must be learning how to use AI to make yourself faster,
- 35:03better, more efficient. And there are so many jobs that
- 35:07will get displaced. So if you're listening and
- 35:12you're not doing that, you know, I'm assuming if you're listening
- 35:15to CXO talk, you are doing that. But so tell, tell the folks you
- 35:19work with who maybe are not so far on the cutting edge, give
- 35:22them that, that advice. It's good advice.
- 35:26So let's talk, come back to startups and we have a question
- 35:29from Twitter directly on startups.
- 35:31And this is from Gus Bechdash, who says, Arvind, what advice do
- 35:38you have for people forming AI ventures?
- 35:42Many make the mistake of choosing problems that the
- 35:45platforms will take over, making their companies irrelevant.
- 35:51I actually don't believe, first of all, that, you know, if you
- 35:55start a company, if you start to solve the problem that you're
- 35:59going to fail because of somebody else, because of like,
- 36:02for example, an incumbent or a large company actually solving
- 36:06that problem and solving them before you.
- 36:09You're going to have a firm belief like as an entrepreneur,
- 36:12that you can solve the problem faster than a large company.
- 36:15A large company always has lots and lots of things to actually
- 36:17worry about. Do you know, they're
- 36:19structurally not designed in a way that they can move faster
- 36:23than you like as a really, you know, as a nimble early startup?
- 36:26So so I don't actually agree with that premise of like people
- 36:30making that mistake. Most of them startups fail
- 36:32because people give up like, you know, because you lose
- 36:34confidence in your own idea. You don't have enough conviction
- 36:39because like, you know, if there's a real problem, well,
- 36:41you have the right to go and solve it and and you will be and
- 36:45you know, competition is not something that's going to
- 36:46matter. So, so with that, with that
- 36:49said, I will add to what's the right strategy for you to choose
- 36:55as a founder? Well, like, you know, pick, I
- 36:58always like to pick problems which are first of all, they're
- 37:02obvious. Like, you know, like you have to
- 37:04go and talk to five people. They will not argue with you
- 37:07that, hey, is this a problem or not?
- 37:09I think is it all like, if you talk to the first five people
- 37:12and there and you you don't have that clarity, but you know, from
- 37:15them, then there's something wrong and you got to like work a
- 37:17little bit more on it idea. So, so get to that, get to that
- 37:20level where like, you know, whoever you talk to actually
- 37:23agree with you that yes, you know, you're solving a real
- 37:25like, you know, important problem.
- 37:28And and also like, you know, like I like to work on problems
- 37:30which have broader impact. So big problems, you know, that
- 37:33a lot of people are going to have because you know, that's
- 37:35going to actually create more opportunities for you.
- 37:38Like even if there are a few other companies that also solve
- 37:41the same problem where there's a large market that you can tap
- 37:44into and you'll have your own success, you know, along with
- 37:46them. And then over time, you know,
- 37:48like, you know, if you do the best that you're going to win
- 37:50like over everybody else. And then, and then the, the, the
- 37:53last thing I would add is don't like come up with an idea where
- 38:01you feel like, well, like all I need to do is use AI and it's
- 38:05gonna actually solve this particular task for me.
- 38:09Like, think about, you know, if it is easy to build, if it is
- 38:13super easy to build, then everybody else can also build
- 38:15it. And then you're not really
- 38:17adding value. So, you know, like AI should be
- 38:21no more than one of the tools in your tool kit to solve that
- 38:25problem. But like, you know, make sure
- 38:27that you know there's something substantial that you're
- 38:28building. There are so many companies that
- 38:32are building wrappers around the models.
- 38:37What about consolidation among these types of companies as the
- 38:41models advanced their capabilities?
- 38:45If you are a startup and you are a thin dropper over the core
- 38:49capabilities of an LLN, well, you'll be irrelevant quite soon.
- 38:57Yeah, you have to like in the mindset, like I'll tell you what
- 38:59we do again, you know, you know, we, we know that like, you know,
- 39:02so the way our product works is that, you know, we actually work
- 39:05with all the LMS, you know, that are out there in the market.
- 39:08You know, whether it's, you know, LMS from Open AI or in
- 39:10Tropic or Google or Meta or, you know, like and all like, you
- 39:15know, so many LMS from open source, we work with all of
- 39:18them. And our model for like what we
- 39:21do is, well, we're going to use all the capabilities that these
- 39:27LLMS, you know, provide to us and they will actually bring
- 39:30those capabilities to our customers.
- 39:32But we'll also be, you know, ensuring that we're building a
- 39:35very deep technology stack on top of that platform.
- 39:39And as the LLMS advance, as they actually, you know, add some of
- 39:43those capabilities that they've built ourselves that actually
- 39:45throw our, we have to throw what we built like, you know, that
- 39:49the LLM providers can do already for us and actually keep going
- 39:53up the value chain. Like, you know, and, and so
- 39:55that's, that's the model that we choose.
- 39:56Like, you know, like to get out your space and maximally use
- 40:01like the innovation that's happening in the industry, but
- 40:04then you don't build build a significant layer on top of of
- 40:07that so that you can make that technology accessible to your to
- 40:09your customers. Given the importance of the
- 40:14foundation models to your business, how do you manage the
- 40:18fact that AI is evolving at such a rapid pace?
- 40:24And how do you balance the stability of your business and
- 40:29product direction while these underlying capabilities are just
- 40:33shifting all the time? We have no choice.
- 40:36You have to like take advantage of the rapid innovation that is
- 40:40happening in the industry, otherwise you're going to be
- 40:43left behind and you have to fundamentally change like, you
- 40:46know, your execution model. Like, you know, we have to like,
- 40:49you know, like we, when we started, you know, when we
- 40:53started glean the we sort of very understood that very, you
- 40:57know, at a fundamental level, like, you know, what
- 41:01technologies, you know, we had available from open source and
- 41:04cloud. And, and then you get to sort of
- 41:06build your road map and you build a, you build a one year
- 41:08road map. And and that's, that's not no
- 41:11longer how things work today. Right now.
- 41:13Technology changes on a monthly basis.
- 41:15And, and so you have to fundamentally change your
- 41:17architecture. Like, you know, like one of the
- 41:19things we change, as an example, is that we don't have the annual
- 41:22like, you know, or a quarterly planning process.
- 41:24We switch to a monthly planning process in terms of like, like
- 41:28how we're going to actually, you know, build our technology
- 41:31because every month, like, you know, there's new things to look
- 41:33at and you have to quickly adapt.
- 41:35The other thing that we, you know, also added is this concept
- 41:39of the, well, figure out what you're going to throw like every
- 41:43month, you know, in technology that you've built.
- 41:46Like this is a new fundamental way of, you know, building tech
- 41:49startups is that, you know, you will become obsolete if you
- 41:53actually hold on to technology that you build for multiple
- 41:57years because all of that technology that you built two
- 41:59years back is most likely obsolete at this moment.
- 42:03And so you have to constantly like, think, you know, of this
- 42:06execution model where not only are you thinking about like new
- 42:10things to build, you're also very actively thinking about
- 42:13like, you know, things that you need to actually throw away and
- 42:15actually leverage like, you know, the innovation from the
- 42:17industry to replace that. We have another question from
- 42:21Twitter from Arsalan Khan again who says since most AI and data
- 42:27is created in English or Chinese, do you think start-ups
- 42:31should also focus on non-english or Chinese?
- 42:36And does this create a digital divide?
- 42:391st, the content is created in many, many different languages.
- 42:44Yes, you know, I, you know, English is dominant in some
- 42:48ways, but there is, there is plenty of like, you know,
- 42:50content, plenty of systems, you know, in different languages.
- 42:54And, and I think what AI makes it possible today is it actually
- 42:58helps you build products that work like, you know, that, you
- 43:02know, that are that, that are global in nature.
- 43:05Like it's much easier today to build a product, you know, that
- 43:08you can actually bring to customers in, you know, all, all
- 43:12parts of the world. You can localize your products
- 43:15much easier with AI. You can make it work in Japanese
- 43:18and in Korean and like Hindi and all the different languages at
- 43:23at a much faster pace. Like, you know, so you know,
- 43:26this is, this is one of the, you know, a, you know, a core
- 43:31capability of AI. But then in terms of biases and
- 43:35the digital divide, it is true that the, and it's been true
- 43:40like, you know, forever, like, you know, even on the Internet,
- 43:42even Pai, when you go on Google and search for information,
- 43:45there's always that bias that creeps in because like, you
- 43:49know, English dominates as, as the source of knowledge in the
- 43:53world, right? And, and so the, that, that's
- 43:57it, that's, that's a good one. Like there's a problem I don't
- 43:59have a good answer on, like how, how to sort of resolve that as
- 44:04as as you know, AI becomes more and more capable.
- 44:06How do you make sure that it's taking everybody's point of view
- 44:09and taking, you know, all the knowledge that's out there?
- 44:12I think I think like one thing maybe I should add is that we
- 44:16have more capability today with, you know, with AI to process
- 44:20like even non digitized, you know, content in a much, much
- 44:24easier way. So hopefully, like, you know,
- 44:26there's, there's some silver lining on, on on that side that,
- 44:29you know, you can actually make use of, you know, data
- 44:35knowledge, you know, content and like different languages more
- 44:38than ever before with AI. Let's talk about start-ups now
- 44:42from the perspective of enterprise buyers, which is a a
- 44:48very important part of course of any start-ups life cycle.
- 44:52If you're, if you're an enterprise start up, do you have
- 44:55advice or a a framework that enterprise buyers can use for
- 45:01evaluating AI startups? Especially given the fact that,
- 45:07as you pointed out, every startup these days is using AI
- 45:11and the hype is so intense, very often it's, it's hard to sort
- 45:18through the claims to find out what's real and what's not.
- 45:21And I'll just mention one one thing here that I remember
- 45:26traditional software companies, ERP vendors and other enterprise
- 45:29software companies. And I have to say this
- 45:32situation, you know, 20 years ago was no different from that
- 45:36standpoint. Now.
- 45:37I mean, software companies make outlandish claims.
- 45:39And So what should IT buyers and line of business buyers do about
- 45:43it? This is one of the toughest
- 45:45problems like, you know, for, for an enterprise buyer today.
- 45:50I think the AI industry has done a bit of disservice making bold
- 45:55claims, but then not not being able to follow through.
- 45:59You know, like it's very easy to create really, really amazing
- 46:02demos and visuals, you know, for what, what AI can do for you.
- 46:07And enterprise buyers have like, you know, and they've realized
- 46:11that like, you know, as they try these products out that like,
- 46:13you know, the claims often, you know, like, you know, are are
- 46:16much larger than like, you know, the reality.
- 46:19So I think like I would say like, you know, maybe maybe let
- 46:25me take a step backwards and then first talk about like,
- 46:27well, what should we even do? Like I think like there should
- 46:29be a plan for like how you're going to roll out AI inside your
- 46:35enterprise. And I feel like, you know,
- 46:38centralizing that, you know, you know, and having a core, you
- 46:44know, AI strategy for enterprise, first of all, is the
- 46:47right way to start. Like now think about all the
- 46:50things you'd like to do this year with AI.
- 46:52What are the different areas that you would like to actually
- 46:55see AI make an impact? You can pick a few departments.
- 46:58You can say that, well, you know, for our engineering teams,
- 47:00for our customer service team, you know, like these are the top
- 47:04two or three priorities, you know, where we, you know, want
- 47:06AI to actually make an impact. So first build that road map and
- 47:10build that together in your enterprise, you know, the CIO or
- 47:13if you're just, you know, if you've chosen to have a, a head
- 47:17of AI, chief AI officer type of, you know, role in your
- 47:20enterprise, like, you know, let them, you know, give them that
- 47:22charter of like making sure they're working with all the
- 47:25different, you know, functional teams.
- 47:27And you come up with a, a desired road map for AI for you
- 47:31for this year. So you start, so you start
- 47:33there. Now in terms of like the number
- 47:38of vendors is not just like, you know, Michael, as you said, like
- 47:40it's not just the startups, it's actually every existing software
- 47:44company is also an AI company. They all have, you know, AI
- 47:47things to sell to you. And I think you to make some
- 47:50decisions there, like what's the right strategy for you?
- 47:54You know, we feel that like, like, no, like, you know, you
- 47:57have to be, you have to control like how many AI products you're
- 48:01going to actually bring in. It's very, very hard to
- 48:03evaluate. It's not easy, by the way, like,
- 48:05you know, it's just like, you know, the, the time that you
- 48:08have to spend time to evaluate every single AI product is
- 48:12enormous. Like, you know, you like, you
- 48:14know, like setting those systems up, getting them up and running
- 48:17in your environment and testing them.
- 48:19And like, you know, you sort of like companies have gone through
- 48:21this exercise where they spend 6 months, you know, and just to
- 48:24find out that, you know, that this thing didn't work.
- 48:27So like, what are the right strategies?
- 48:28So, so like 2, So, so, so my suggestions, like, you know, I
- 48:32have two pieces of advice. Like #1 like, basically work
- 48:36with fewer, you know, number. Like don't have too many POC's.
- 48:40Like, you know, start with a few so that you can do justice to
- 48:44them. And, and 2nd, instead of like
- 48:48relying on demos and, you know, getting excited by a
- 48:52presentation. Well, you can have a safer
- 48:55strategy, which is like, we'll go and what every vendor that
- 48:58you work with, like see if they have proof points like go and,
- 49:01you know, make them, you know, connect, you know, connect you
- 49:04with, you know, customers that they were able to create success
- 49:07for. So I think that is like, you
- 49:09know, it's a lot more helpful like, instead of evaluating
- 49:12like, you know, like actually had those conversations with
- 49:15your peers in the industry and see like, you know, where they
- 49:18are achieving success. Like if you tried 4 things,
- 49:21you'll see one thing that's successful, share that story
- 49:23with others and, and, and they will share with you.
- 49:25And that's probably the way to sort of scale and, and maybe
- 49:29maybe this is a plug for Glean, but the, the way we think about
- 49:33AI is that, well, like fundamentally all AI in your
- 49:37enterprise is about making, you know, working with some data
- 49:42that's in your enterprise using the reasoning and intelligence
- 49:46powers of the language models. And then after that doing some
- 49:49work, which you're again going to save, you know, in your
- 49:53enterprise systems, you know, that work is going to get
- 49:55recorded and saved in your enterprise systems.
- 49:57So, so fundamentally, when you think about all AI use cases,
- 50:01they're about working with your enterprise information and, and,
- 50:07and then you're making and applying AI on it to do some,
- 50:10you know, to, to actually make some magic happen.
- 50:12And so we chose a different strategy with Glean, which is
- 50:14that, well, why don't we actually build a system that's
- 50:17connected to all of the enterprise data, right?
- 50:19That's what Glean is. And, and now we're giving you
- 50:22this platform where you can go and build like many, many of
- 50:26these agents, many of these applications yourself.
- 50:30And, and, and that way, like, you know, with this horizontal
- 50:32strategy, you can, you know, you can do a much better job at
- 50:35security governance, you know, also like not having to buy
- 50:38many, many different products. Like it's more cost effective
- 50:42and, and and and and allows you to sort of like, you know, get
- 50:45more value through like, you know, through 1, like, you know,
- 50:4711 product. All right, we're almost out of
- 50:49time. So Arvind, I'm gonna ask you a
- 50:51bunch of questions, some from me, some from the audience
- 50:54listening and I'll ask you to respond very quickly from an
- 50:59enterprise perspective, the build versus buy decision, how
- 51:04should enterprises make that choice?
- 51:06Very quickly please. It's a build plus buy.
- 51:09You have to make sure that you get as much turnkey technology
- 51:14as you can, but you have to also remember that it's not going to
- 51:17be enough like, you know, to add true value.
- 51:19You will have to, you know, you know, build on top of those
- 51:21systems. From Twitter, a big hurdle for
- 51:26AI and LLM platforms is that they are not integrated with
- 51:30work flows. How do you see the market
- 51:32evolving? That's exactly right.
- 51:34And I think the you you have to, you know, build that layer.
- 51:39One of the key techniques for that has been rag like so you
- 51:43take all of enterprise data and knowledge systems for flows and
- 51:46you actually build the, you know this, you know, a retrieval
- 51:50system and an action system on top of all of your enterprise
- 51:53data and systems. And then you connect with AI
- 51:56like you know, that's exactly what we, you know, the problem
- 51:57that we solve with you. Another one from Twitter who the
- 52:02CTOCFOCOO chief digital officer? Who creates AI strategy?
- 52:08And should the chief AI officer report directly to the CICEO?
- 52:13That's a good idea. Like so if you, if you actually
- 52:15have a chief AI officer, having them report to the the CIO or
- 52:22the CEO could be a good strategy because this is a company wide
- 52:26effort. So like the, if you don't
- 52:27associate it with a function that's actually better in my
- 52:29opinion. And but otherwise, like, you
- 52:32know, I think like, like, you know, it's not don't be rigid.
- 52:35Like, you know, like whoever is motivated, who's excited and
- 52:40like, you know, like and has the capability, let them drive their
- 52:43efforts initially And like if you can always figure out how to
- 52:45reorganize later, like, you know, in a more scalable way.
- 52:48Here we have another question from Twitter, and this is about
- 52:51pain. If AI startups are getting more
- 52:54lean, are older AI startups at a disadvantage?
- 52:59And as these older startups become more lean because more
- 53:04efficient using AI, what is the implication for employment and
- 53:08their workforce? So on, on the 1st question,
- 53:11there is indeed a like a first mover disadvantage in AI because
- 53:15when technology changes so fast and you, you have some of that
- 53:19legacy code base in your systems, like it makes it like,
- 53:22makes you a little bit less agile.
- 53:23So well, that's, that's part of like, you know how it always is.
- 53:26Like, you know, as a new startup, you always have the
- 53:28agility advantage. And as a start, that's a little
- 53:31bit older. Well, like, you know, work hard,
- 53:32like work hard on modernizing, you know, your systems, your
- 53:35stacks so that you can, so you don't fall behind for those
- 53:38reasons. And then in terms of employment,
- 53:41I think the like, I'll, I'll just say one thing.
- 53:45I've not seen people losing employment because of AI.
- 53:49Yeah, like, you know, we work with so many large enterprises.
- 53:52They're bringing so much automation and efficiency for
- 53:55them. But like, you know, every, like,
- 53:57you know, every enterprise that we work with, you know, they
- 53:59actually are concerned with both bottom line and top line.
- 54:02And so when they get some bins, nobody's, you know, giving up
- 54:05their team members. Like, you know, they're not
- 54:07actually, you know, they're just actually thinking about like,
- 54:10well, I can do more. And, and so companies are
- 54:13actually building products at a faster pace.
- 54:15They're getting more things done.
- 54:17So, so I'm not, I'm not super worried like, you know, I think
- 54:19I guess like, like I will come back to the point that well,
- 54:22like just stay relevant. You know, that's that's the
- 54:24thing that matters for you as an individual.
- 54:26If you if you learn how to use AI you know you will have no
- 54:29problems you know in the future. You're advocating maintaining
- 54:34intellectual curiosity. I'm not trying to put words in
- 54:37your mouth because as AI drives efficiency gains, if you're
- 54:42intellectually current on what's going on, you can adapt and work
- 54:48around that the changes that are taking place.
- 54:51Yeah, and companies are really hungry.
- 54:53They're really hungry for AI experts, for AI talents or like,
- 54:57so become one. I think it's going to be good
- 54:59for you. What advice do you have for
- 55:00entrepreneurs considering starting an AI startup?
- 55:04Number one conviction like you know, we know.
- 55:05Stay, stick with your idea, don't give up and, and #2 well,
- 55:10like, you know, if you're, if you want to succeed, like, you
- 55:12know, be ready to, to work very, very hard.
- 55:16And, and, and then, then the last thing I would say is that,
- 55:20you know, a company is all about people that you, you know, build
- 55:23it with. So like, you know, focus on like
- 55:26having, you know, having a great following team and, and, and
- 55:30spend a lot of time, you know, trying to get the right people
- 55:33that employees, you know, in your organization.
- 55:35And like you know, when you get the right people, you know they
- 55:38will do the you know they will build great products.
- 55:40They will make you succeed. Final thoughts or advice for
- 55:43CIOs who are making AI investment decisions very
- 55:46quickly, please. Focus on security, focus on the
- 55:53like, you know, centralizing, you know, like your AI software
- 55:57stack and, and, and, and I think the, the, the last thing is, you
- 56:03know, that I would say is that like we don't try to sort of get
- 56:08small wins. You know, this is my advice.
- 56:10Like, you know, don't like, you know, don't like, you know, like
- 56:12create projects where you know, if it doesn't deliver 50%
- 56:16success, it's a failure. Like, you know, like, you know,
- 56:18force every, you know, team, every function in your team,
- 56:21like in your organization to, to pick, you know, one or two wins
- 56:24for AI, like, you know, every quarter and see how that goes.
- 56:28All right. And with that, we are out of
- 56:30time. A huge thank you to Arvind Jane
- 56:34from Glean. Arvind, thank you so much for
- 56:37taking your time to be with us today.
- 56:39Thank you so much, it was a lot of fun.
- 56:41And thanks to everybody who watched and especially you folks
- 56:44who asked amazing questions. You guys are truly awesome.
- 56:48Before you go, subscribe to our newsletter.
- 56:51Join our community. Go to cxotalk.com.
- 56:54Subscribe to the newsletter, check it out and we'll see you
- 56:59again next time everybody. We have awesome shows coming.
- 57:02U talking about AI? See you later.