#16: AI Will Turn Every Business Into an AI Business (No Exceptions)
28:21 Share

And my fundamental belief is AI is going to team up with humans to change the way we work, live and play. Welcome to Divot, a community for people trying to make their mark on the world, where each week I'm interviewing the best minds in business, tech, sports and entertainment to see how they made their mark. Go to Divot.org to see all the episodes or you can watch them on YouTube, Apple Podcasts or Spotify. This episode is brought to you by Salesforce.

Tell us why you're so excited about collaborative intelligence. What is it and where is the future going with it? Absolutely, right? So I think AI is yet another technology. It's yet another tool.

and I look at it as the horse. The jockeys are sitting not only here but everywhere in the audience. So it's up to us on how we use this technology to elevate our skills to superhuman skills. And my fundamental belief is AI is going to team up with humans to change the way we work, live and play.

And as I mentioned in the clip, it's 100x force. First, machines finally understand our language. And I know we'll talk about vibe coding. Till now, in the IT industry, the last 50 years, we have been learning machine language. Learning this kind of coding, that kind of coding. We're slaves.

Now finally we can just sit back and we can talk and machine does the work and secondly with all the innovation that has happened in the information technology industry, machines can finally

not only do research they can plan, but they can take action. So if you take a 10x disruption on the front end user interface becoming conversational and 10x on the back end, it's a 100x disruption and now it's up to humanity on how to use this technology to essentially collaborate with us to get to new heights.

And that's why myself and Mayfield is so bullish on it that rather than using this form of enterprise software to improve my productivity, now you and I can be on a beach and tell the machine to do her stuff.

So that's the era we are entering where technology and machine should be. So it's a pretty exciting time. And there are many, many use cases. Yeah, in the short run, there'll be pain. Jobs might get displaced. But humans win. We're not dinosaurs. So it's a very, very exciting time. It seems like it frees up time for us to be more creative. It frees up time for us to do more interesting things than-- if a machine can do it and it can be automated,

We're more capable than that. We should be doing more interesting things. It frees up time to simply be human. So let's take an example, right, like of coding. They're like 30 million developers. Maybe they're 40 million developers. People are worried developers are going to lose jobs, right? Like a lot of those people are here. I look at them as builders and creators of the future.

My belief is now using conversation based coding which some people refer to as vibe coding, everybody will be a developer, everybody will be a builder. We'll have 8 billion builders in the world. So essentially if you understand your natural language, you can just sit down and creatively code and create things.

So first the developers are going to be 100x or the creators. So what are these people going to do? Not only can they code, build an app, but they will go out and design new things. We know those tools which are available. They'll sit. If they're in product management, they'll be able to do their job better using wipe coding, if you will. Same thing will happen in sales, marketing, HR. And finally we'll get

personal productivity improvement. So I really think this vibe coding, I want to use a new word, whenever you and I get together. You create new words. I just thought about it, right? I think this is an era, like what I have said before, CAS, which is Collaborative Intelligence as a Service, it just clicked to me. This is like three Cs. The way coding will happen now, A, it'll be conversational.

Second, it will be collaborative. There was a time where two or three best coders used to sit together and do tandem coding. And finally this will be cognitive. I have ideas, the other person is giving me the stuff. So I believe that's the era we are moving into. It's going from vibe development to vibe building to essentially vibe operations. And this is 100x.

30 million developers going to 8 billion builders is more than 100x. And we're going to see this again and again and again. And when people are worried about entrepreneurs, only developer and product people can become entrepreneurs, many people think that way. I don't. I think anybody who has an idea in the room or elsewhere,

should have a chance to be an entrepreneur. This is the great equalizer. Everybody here should lean forward and say, you know what? This is my time that I don't need to have computer science and engineering skills to be able to imagine and do new things which weren't possible before. So that's why this technology can be so interesting. So vibe coding, from what it sounds like you're saying, is way more than just prototyping. It's

fundamentally changing the way software and things get built. Everything, right? Like it's not just for developers. It's for every function, right? Like basically if I'm a product manager, I'm a designer, I don't need to learn these low-level things. I'll just type. I don't need to be the slave of if I'm in sales. They're like maps of 100 sales text tools, right? Now I'll just sit and talk to a conversational interface, what is called an agent,

And they will go work with these hundreds of apps. It's the death of software interfaces. That's what this vibe coding phenomenon is about. It's not just restricted to writing software code. It's broad. Do you think it's going to permeate through all shapes and sizes of companies? Or is it going to be limited to high growth tech companies? I think to me, the software development, wherever there are software people, they need to essentially

use these tools for automating tasks, accelerating productivity, but I'm talking of market expansion. The number of developers is going like what happened with Airbnb, what happened with Uber, Lyft,

the market used to be much smaller of developers. The TAM is now going to be massive. Massive people who could be developing. And then when that happens, with every wave we have seen, just creativity starts, new kinds of companies. We need to just have an open mind. So I don't even know what's possible now, right? Basically you're democratizing entrepreneurism. I'm wondering when does venture capital get democratized?

It will be fun, right? Like this is the great equalizer. So I want to energize people to be trying using these tools. And like you could do things that a handful of people could do in the past.

Well, it's clear that it's fun to you because, and this is one of the reasons I love learning from you because your energy is really infectious. I think some of the other people we've had on the stage, I think they're worried about, more worried about maybe AI disrupting them in venture capital versus them being part of that disruption. It's okay. When the, basically when it gets tough, the tough get going. I love challenges. Right.

Right? Like, why should entrepreneurs only feel challenges? Everybody should be in the hot seat, and maybe it's our time now. I like it. I don't know if people think that should be the case or not. Right? Like, everybody needs to be nimble. Dinosaurs never survive. Entrepreneurs, they love hearing that, you know, the VCs are going to feel a little pain too, you know? They should, always. I've been a serial entrepreneur three times, right? Like, why? We need to be all on the same side. I like being on the bright side. Right now, the entrepreneurs are on the other side.

Let's talk about what seems like a renaissance in hardware in the AI space. What's happening and what do you think the future of hardware looks like? Yeah. So I've been thinking about, right, like where we live is called Silicon Valley. And I don't know when it became a software valley. I actually know. And then it became a media valley in the social arena.

I think with what we have seen now with the advent of AI and cloud and even with mobile, the biggest companies are hardware companies. So if you look at what Satya and other people say, if you look at a cloud, what is it? It's a mainframe. It's a supercomputer that is connected to

through the internet and it integrates all the semiconductor, all the software, all the services that you can just go approach it. The only difference is in the 1970s when you use mainframe you used to have dumb terms even I wasn't born right but now the intelligence is split on the cloud back end and some on the edge but most of the stuff we do is browser based

So essentially what is happening is if you look at the world move from mainframe, then it went to many computers where you had these Dell servers, Sun Microsystems, it went to the desktop. But with the revolution we are in, it's gone back to centralization. That's where when I use GPUs, they're not sitting here, right? They're somewhere in ether. So the reason I'm so bullish on hardware is it's reflecting on

the market cap of the biggest companies, whether it's Nvidia's of the world, AMD's of the world, Tesla's of the world. But then if you look at hyperscalers, AWS, GCP and Azure, they're nothing. They're called IaaS, infrastructure as a service provider. What are they doing? They're bundling hardware, software, network,

connectivity, storage and offering it as a service. They're hardware companies, just the different business model. It was the same what happened to enterprise software. They got delivered at a service. So to me, this is the time and I don't know if they're entrepreneurs thinking here, the largest market cap companies are Apple and Nvidia. And after that is Microsoft.

So I believe there's a renaissance of silicon. There was a saying that, hey, software is going to eat the world. It already did. It already did. Yeah. Right? So it's back to the golden era of hardware and semiconductors. So the number of problems that need to get solved at the hardware level is a lot. Clearly, GPUs and AI accelerators, the big companies are going to own it. But there are fundamental things

that need to get solved. So for example, power and cooling, neglected areas are one of the biggest costs when you run an AI data center. 40 to 50% of your costs go, these things are heat sinks, these GPUs and the other things you put. You need to do air cooling, you need to do liquid cooling, you need to do rack level cooling, what's called immersion cooling. Then,

These things are so quick, you have to connect them. You can't connect them over wires. You need to put optics. You need to put intelligence in those. Some people are saying, "Oh, I'll figure out how on copper to send one terabit." You go, "Really?" Really. So if you look at now for software to come back and even live, hardware is where the innovation is happening.

And that's where I believe it's the renaissance of the hardware era. Not that software won't come back again, but till these things get fixed and the cost has to go down,

and intelligence has to keep going up with these components that are being provided, that's when again those things can wave and I think it's a contrarian thing. We started investing again in the silicon hardware ecosystem 8-10 years back and those companies are growing phenomenally. Phenomenally which most of the people may not know, it's okay.

Yeah. It's okay. It's part of the hardware infrastructure to power these systems and data centers. Everything. Yeah. You have to build an agent. It just runs, doesn't run on software. That software has to run on some hardware. Yeah. Do you think with what AI is doing that

electronics and sort of any hardware application that solves problems in our lives is up for disruption? Do we think that the appliances in the home or our vehicles or other things, do you think all of that's gonna go through

sort of AI wave with new companies coming in and disrupting them or will the incumbents similar in software we've seen some of the bigger incumbents be able to kind of latch on to AI and and have the kind of infrastructure advantage that they have. Yes, I would say the magnificent seven right like the big incumbents have their act together and

So if you go after, like, I can't see many people. I hope people don't have tomatoes and stones. Are there any people from, like, legacy industries or are people here primarily from the tech industry? No, we kicked those people out a long time ago. You kicked them out? Yeah. That's good. So I think here is the thing, right? Like, AI, right now what we are seeing, the prevalence is with...

the digital economy, right? This chat, that chat, this agent, this software development accelerator, it's going to come to the physical world.

And it was tried 10 years back with autopilot. A lot of people use self-driving. I don't know how long it will take. But basic things around us, jobs that humans were doing, many of them in the physical world are going to get augmented and done better by AI-based machines. And that's the concept of physical AI, physical AGI.

And what I mean by that is, it won't be... First, it may not happen in consumer, but there will be applications of it. So for example, manufacturing. The last time there was innovation in robotics and manufacturing, God knows when it was. But it's an aging population. People don't want to do certain jobs. With all the geopolitical stress and the dislocation of the supply chain, there has to be nearshoring. So everybody is going to build.

manufacturing capabilities, it's a crown jewel. But now if you combine intelligence with physical hardware, what we have seen with digital agents, you can do wonderful things. So I think one by one physical LLMs are being created. There's four or five companies. I won't mention them. We are in one of them. But they're going to go find applications

for applying this stuff, what people call humanoids. They may not be running on the streets, but car manufacturing companies need more and more of this. Tesla is already building it, where you use physical intelligence

to have robots do things that current generation of companies aren't doing. Same is going to happen in construction. Same is going to happen in difficult things like fire. You have to climb up on buildings that's too risky, right? Like healthcare for old age parents, old age homes. So we are going to see these machines entering us

Right? And helping us in places where they just can do a better job. A nurse is supposed to help me with clinical stuff, not to remind me to take medication or to come take my temperature if I'm there. So that's what we have to imagine, the industrial applications, the B2B applications, warehouse, logistics.

It'll be B2B first. And yes, for home cleaning, cleaning the dishwasher, like making bread and other things, there'll be machines which will keep coming up. I'm still waiting for automatic chai makers. They're like only coffee makers. I wanted the Indian way. So there will be, even in home stuff,

the next Roomba, the next things, it'll happen. So that's what, it will come to hardware. Today, the hardware I was talking about was the supercomputer in the cloud. Yes. But it's going to come to the edge. Just powering the software and then... It's going to come to the edge. Yeah. Right? Like, there will be

100 billion connected IoT devices which will all have intelligence, right? So security cameras. It's very hard to figure out that you have a physical breach. You analyze the data, you sit on screens, it's going to detect it at the edge. So this physical thing, AI, what we are seeing with this chat, that chat, this sales agent is going to be in the physical world.

and it's going to be prevalent.

And there, there's no fear of job loss. It's only protecting us. It's work we don't want to do. Right? Like, it'll be amazing, man. Like dishwasher, physical thing comes, cleans it, brings it out. That's further out. But we have to imagine. What are my kids going to do if I have a dishwasher? They'll figure out something. They'll come spend more time with you. Yeah. Right? Like at the end, we need to essentially do things like humans should be doing. Yeah. Right?

So it's okay. They'll do other things, trust me. They'll create more work for you. What are the barriers that exist now that is keeping the physical AGI from happening? Yeah, so I think the first thing is there is a limitation with the current generation of models. They need to all...

get augmented and expanded. There's nothing wrong with transformer based architectures. You take that, they have been trained on the open internet. Let's take a robotic application in a manufacturing car plant. How do I train this model? Then you need to understand, you need proprietary data. Each car manufacturer is different. So training there's an issue.

Data is an issue. Then when these robots work in the physical world, they have context. They have to mimic human behavior. So you have to go solve all these problems. Need great entrepreneurs.

and VCs hopefully can make some money along the way. So these are like tough problems to be solved and all of us are funding companies but I'm not so existing models need to be augmented and expanded. They have been trained on certain other things not on the physical world so there's a flurry of physical AGI companies which are spawning and that's why some of their valuations are much higher.

at the start than many of the public companies you and me know in my favorite space, software as a service. So many industries are being reimagined and it seems like even the core business models are being reimagined. Are there particular industries where, you know, I mean we see some that are already kind of falling into place and others that are coming, but do you think AI has an application

Almost every industry, do you think it has limitations and say, hey, okay, we've started software. Hey, there's some hardware things you can go to. Or do you think it just...

10, 20, 30 years from now, is the future just full of AI inside of basically every application and problem that we need solved? I think it's hard to say whether it'll be 100%, but we should imagine everything we are doing working in a company, the same thing that happened with when the internet came, there was this notion of e-commerce, e-business, right?

I think the same thing is going to happen. Every business is going to become AI first. They have to figure out if they were starting today, let's look at a law firm. Let's look at a consulting company. Let's look at an IT services company. What would they look like? So I've been studying and we launched a company today called Groove. I don't know if the people are here in the audience or not. It is reimagining IT services businesses.

companies which do implementation of IT, they do managed services, they do BPO. It's a 1.5 trillion dollar market globally. Most of it... This is like outsourced services that you have. Yeah, Accenture, TCS. Most of it, you don't do it in-house, you have outsourced it. Unfortunately, a lot of it has been offshored. Yes. But now, what is the difference with the past to this thing?

is the following, right? IT services companies charge per hour, they charge per project, or they charge on a monthly basis. If I am an AI native IT services companies, some people in the front will be humans, but most of the work, which is repetitive, it's not new software development, can be done using AI. And you'll be left with some people in the back end because humans will always be in the loop.

So you could build a new kind of an IT services company, let's pick security, a managed security services company, which doesn't charge per hour, which doesn't charge per month, doesn't charge per contractor. It only charges you when you use them. So if you see a security event, that's when you charge them because most so it becomes like utility computing.

So an IT services company could be charging for outcomes or on a consumption basis the same way

a cloud computing company does. And this is the same way the SaaS industry happened, which is you take enterprise software, pay me three years upfront, five years upfront, and SaaS companies came and said, "No, pay me on a monthly basis." Now it's going to change. And this is the problem SaaS companies are going to have. Their business model is being disrupted. People aren't going to pay per month.

They're going to pay per use. And the model is moving to consumption. So it's SaaS companies are under attack. But to me, going after services businesses, IT services is 1.5 trillion. Any people from investment banking here?

Thank God. - No, we threw them out there. - Okay, you don't have any sponsors, right? Or any people from management consulting, McKinsey? - No, no, no, no. - Okay, thank God, man. Now I'm feeling relaxed. So think, yeah, man, like what e-commerce, you never know when the stones are gonna come. Or tomatoes, I'm just kidding. Right, like think about if you're building a new wealth management company, I really think you can have a billion dollar AUM, charge very, very little,

and have one or two people. Just do the client interface part when you're managing somebody's money to give them trust. Applications AI can do, stocks AI can buy based on research, reports AI can produce. They're mad at you, AI can answer.

So I'm just saying if you look at it, you can imagine the way Tesla built a company. Imagine you could be a two-people billion-dollar AUM company. I'm trying to figure out what is such a thing in venture capital. So you have to reimagine the next McKinsey, the next Goldman Sachs. There's so much money. The next law firm, the next VC firm. Basically, there's just...

It's just a lot of, right, like everywhere there's a lot of opportunity. Disrupt using business models and delivery models. So it's the great equalizer. I love it. Right? Like everybody needs to have some heat. And by the way, IT services is 1.5 trillion. Professional services firms, which are like knowledge workers, is $20 trillion. Wow. Information technology is only a $3 trillion industry.

2 trillion is telecom and devices. Enterprise software is only 600 billion. We're talking about 1.5 trillion, 20 trillion. So that's why with AI, shouldn't we be like, shouldn't this- I can see why you're optimistic about it then. No, no, no. You and I will be sitting like startup grind five years from now. Won't be here. We'll be in Hawaii.

With all the great wealth these guys are going to create and all the sponsorships and tickets. No, I'm not kidding. Let's dream, man. Why should? Why should?

Right? Like the money and the wealth be concentrated in less than 1% of the people, right? This is a great equalizer. Knowledge is getting democratized, right? Expertise is getting democratized, so why can't everybody be an entrepreneur? This is the great equalizer and that's what this country is about. It's about opportunity, it's about entrepreneurism and sky's the limit. And this technology, I'm telling you, really, really, really

allows a normal person, a human to become superhuman because you don't need that much capital. I mean it, right? We're going to Hawaii in five years. Naveen Chhara, thank you so much. Absolutely.