Loading summary
Transcript117 lines
- [00:01]
A
I'm grateful for marketers like you. Not the ones waiting for their boss to tell them what to learn, but marketers who actively plan for their future. Because you listen to this podcast, you're already ahead. You're seeking to understand AI instead of waiting to see how this AI thing turns out. But here's what I've learned after a decade of running conferences. Interest doesn't create results, implementation does. That's why we created AI Business World 2026, where you'll master AI skills that make you indispensable, where you'll get your questions answered by experts, and where you'll connect with over a thousand marketers who are implementing AI right now. Years from now, you'll look back at this moment and remember this is when you got ahead. Head to AIbusinessWorld live and secure your competitive advantage.
- [01:00]
B
Welcome to the AI Explored podcast, helping you put AI to work.
- [01:06]
A
And now, here's your host, Michael Stelzner. Hello, hello, hello. Thank you so much for joining me for the AI Explored podcast brought to you by Social Media Examiner. I'm your host, Michael Stelzner, and this is the podcast for marketers, creators, and business owners who want to know how to put AI to work. People say that AI agents will change the way work is done. If it's true, when will it happen? In today's episode of the AI Explored podcast, we'll explore the future of AI and marketing. My special guest is an AI analyst and general partner of the AI Fund at Blitzscaling Ventures, a venture capital firm founded by LinkedIn's co founder, Reid Hoffman. He's also started the Llama Lounge AI Startup event in San Francisco. He was previously an analyst at Forrester Research and the co founder of the Altimeter Group. Jeremiah Oying. Welcome to this show for the first time.
- [02:00]
B
How you doing today, Michael man? Sure, I'm new to the show, but I've known you for like 15 plus years.
- [02:06]
A
I know, it's awesome.
- [02:07]
B
I am so delighted to be here with you and to all of your amazing community that is listening, thank you.
- [02:13]
A
Well, let's start with your journey into AI. How in the world did you actually get into AI? Because I'm sure there's a story there.
- [02:20]
B
Yeah, well, I was working in the Web3 space and that collapsed, and I was asked to exit, which was quite unfortunate and never a good feeling. And I said, well, what's next? And it was very clear that generative AI was picking up steam. And my wife, you know, nudged me. She goes, you should just. Just grab that and So I did. And Michael, I have a history of like when I see a new tech trend and I, I have kind of an advantage because I've been in Silicon Valley for 27 years. I can kind of see what's going to come because, you know, the events are happening right here in my backyard that I just went for it. And so I committed to attending three to five events in the generative AI space. And then I launched my own event series in April 2023. Now just to remind, OpenAI kind of went public around November, about three years ago in 2022. So just a few months after that. And my event series grew and grew and then I was offered a position at Blitzscaling Ventures to do investments in early stage startups. And I continued to get on the speaking track and just getting in front of this and I seen it grow from not that great of tools that you could talk to. Now they're starting to be actionable agents to integration even last week into the Google Apps. And it's just growing and it's an exciting place to physically be and to watch and see how businesses of all sizes, including small and medium sized businesses and moms and pops and individuals to actually use all of these tools. So that's kind of my story.
- [03:47]
A
Okay, so about you said April ish, 2023, you started doing these local events. What was the vibe? What was kind of the feeling up in the Bay Area then and how has it evolved?
- [03:59]
B
Oh man, it was scrappy. So there was just a handful of events when I started Llama Lounge and I chose the name Llama Lounge because many of the other events they call themselves AI this or Machine Learning Meetup that or AGI this. And I said, gosh, I need to use my marketing skills. By the way, I have a background in marketing at four Star Discovery Marketing and I have a marketing degree from San Francisco State University. And I said I need to come up with a better name. So I came up with a memorable alliteration, Llama Lounge. And boy did it stick. And the first event was in kind of a bad neighborhood. It was a ninth in market, which is kind of seedy. And at that time Uber had moved out and you know, we're just coming out of the pandemic and it was just a rough neighborhood and people still came into this rough neighborhood. And I noticed, Michael, that they would open their laptops and show off what they were working on in AI. And I said, ah, okay, so that's cool for the first event. For the second event, guys, we're gonna have a sign up form where you can sign up to demo and then it's grown and grown. So fast forward nearly three years. Llama Lounge is one of the premier AI startup event series in the world. So the average event has three to 400 attendees, mostly AI founders, typically 75 VCs, then 50 buyers from large companies and hundreds of founders apply to demo. And there's tons of social media for it. We have a group picture and it's being hosted at some of the world's biggest corporations now. It's been hosted at Zoom, SAP, Amazon Web Services, next is Snowflake, then to Google, Stanford again. And it's just gone bonkers in terms of demand. And these large companies host it at no cost to me. And then we feature startups and then we feature the host as well. So on the average we turn away decline 900 people per event because we just can't fit them. People fly in from all over the world. So it has become kind of this premier thing. And I was asked to speak about community and events even at TED AI, which happened in a few months ago in San Francisco. And so I talked about community and I talked about the culture in Silicon Valley where people are collaborating and building and it's resulted in some new cultures which we can talk about in a bit. So needless to say, and I saw this, and you saw this too in web2, it was like scrappy meetups turned into larger formed events and of course there's conferences. And I recently saw you at the Marketing AI conference in Cleveland. And the industry is booming now, it's formalizing more and more.
- [06:42]
A
Love it. Okay, so my next question has to do with misconceptions. So outside of Silicon Valley, and I know you probably are interacting with people from all over the world, what do you think are one of the biggest misconceptions right now when it comes to AI? Maybe in the business world or in the marketing world.
- [06:59]
B
I think the misconception right now is that anybody would be surprised that there's an AI bubble. Of course there's going to be a bubble. Every tech trend has a bubble. And so how does it start? So it starts as a small energy and then the world gets really excited and businesses lean in and investments, they lean in and it gets over indexed. And this happened in dot com web two sharing economy web three multiple times and now AI. So of course it's going to happen, but you should prepare for it to happen because the trillion dollar companies come out of that and that's okay. So right now there's 42,000 AI startups. Michael. By the way, that in itself is incredible because the cost of building these startups and these are small businesses is very low. Right, because you can use AI to build these new startups. So 90% of them will not sustain. There's just not enough room in the market. And that's okay. And so I think that the fact that people are surprised that there might be a bubble is surprising to me. So that's the major misconception. I think the other misconception about Silicon Valley or AI is that it's the only cause for layoffs. And it's not just yesterday. Did you see the news? One of the big agencies is, I think laying off like 4,000 of their workers. But the main reason is because there's a merger and acquisition and so they just have redundant roles. So in some cases AI is kind of like this scapegoat, say, oh yeah, of course, AI. But when the CEO says they're laying off people, they have to also appease shareholder demands and say, yes, with this smaller team we will be more efficient and of course we will apply AI when possible. And then everybody says, ah, of course it's all AI's fault. But no, it's not. There's often some cracks within the business model. So those are the two misconceptions that I just wanted to address.
- [08:56]
A
Do you think we're in the trough of disillusionment in the Gartner hype cycle? You know what I'm talking about? No, no.
- [09:01]
B
I think we're so in the Gartner hype cycle. And I was, I'm ex forester. So we used a different graph, so there's a different S curve. We're just coming off the peak of the first.
- [09:09]
A
Okay.
- [09:09]
B
The trough is coming. So the reckoning, if you will, the calling, hasn't happened yet. It's going to happen soon.
- [09:17]
A
Got it. Okay. Everybody's talking about agents. As you know, every major billion dollar company that's a tech company is screaming AI agents from every hilltop. And I know it's a hot topic in the Bay Area. What I'd love you to talk about is what are the benefits, what are the upsides? And to businesses of all sizes when it comes to agentic AI?
- [09:36]
B
Oh, this is great. Let's talk about small businesses. And I work at a small business, so I work at a very small VC fund, Blitzscaling Ventures. We have a very small investment size compared to the billion dollar Sand Hill firms. Like we're not that we're under 100 million, a smaller outfit. And so there's four partners and we have no staff. So like you, if you're a small business, we have to use the tools as best as possible. And we do. And so we just don't need a lot of junior staff. Now of course we outsource to consultants when needed, but we have to use all those tools too. Now the great thing about this AI and Agentix software is that an individual contributor can be like a small team or a small company. And a small company can be like a medium sized company and a medium sized company can be like a giant enterprise. There's one more wrinkle and an enterprise can finally act like a startup and be okay, all right, so there's like this flip over. You could, if they can figure out how to be this certain culture, which we'll talk about later. So what the heck is an AI agent? Well, let's first define it. And I interviewed actual humans. Shock, shock, shock. Because my background is market research and I interviewed the actual leaders in the agentic space and I say let's define AI agents. And I published this over a year ago. AI agents are software that can sense the world around it. And that could be inside of an app, by the way, or it could be physical cameras and then it makes decisions on things to do and then it actually completes actions. By the way, large language models don't actually complete actions. They just talk to you. And then it can learn and improve over time. In the advanced use cases, the AI agents can do orchestration with other AI agents. Maybe it'll partner with another AI agent or summon or hire or even build another AI agent. Like a fleet or a group or a a swarm of AI agents. Physical manifestation of an AI agent are self driving cars. It fits the definition. In fact, Waymo is rolling out throughout the Bay Area and also in many other cities including Phoenix, Arizona and beyond that, they are themselves AI agents that can sense the world, complete actual physical tasks and improve and be networked. So that's what an AI agent is. Now for a small business or an entrepreneur, it means you have access to all of these 24, seven staff workers that are virtual. They're AIs. Now let's be honest, right now they're not that great, you know, but as we saw from the recent MIT report, around 11% of roles can already be done by AI and AI agents. And we're just in year three, so imagine what happens in another three years. Now I know this sounds scary, but, but if you're a small business, this is also an opportunity for you to level up and use these tools for productivity. So that's what an AI agent is and the opportunities for businesses of smaller size. Michael.
- [12:32]
A
Okay, so a couple of quick questions before we get into some other questions here related on just what you said. What I heard you say was that AI agents effectively can sense the world around them, make decisions and complete actions. And I think a word you didn't use is autonomously, but that's effectively what we're saying, right? Is that.
- [12:53]
B
Yes, that's right.
- [12:54]
A
It's able to actually make a decision without human involvement.
- [12:58]
B
One caveat, the human is in the loop during this first phase, during experimentation. Just like there are human pilots in the self driving cars to make sure they're safe.
- [13:05]
A
Okay, good, that's important. So today humans can be in the loop and should be in the loop. What I'm curious about though is the way that you've described AI agents. Are they here yet? Meaning, you know, the way you describe them, where they can complete actions and can learn and improve, or is this something we are striving to achieve? Do you understand what I'm asking?
- [13:26]
B
They're here in an experimental set and they're being deployed, but they require some customizations. And we've done investments in companies that can do this. So we've invested in Please AI, which is for small business owners, could definitely use Please AI as it rolls out in public, or you can use it to plan your trips or your goals or meetings or events, and it will help complete those actions. Now what's interesting at the end, if it can't complete the action, like booking a restaurant or hotel or an airline ticket, it'll summon a virtual assistant, a human to actually complete that. So there is some breakages at the end, and that's part of the strategy that Please AI will do. We also invested in Crewai, which is for Fortune 500. It's big companies, but we'll also see versions for small medium business. HubSpot's got their version called Agent AI, which is a marketplace for any marketer to go and find agents on HubSpot's marketplace. And this is run by Dharmesh, the founder of HubSpot. And you can actually use these agents for creating content or to do competitive research or to complete CRM records. And of course we're already seeing AI agents being integrated into CRMs to do follow ups in an age like we're experimenting with that as well. So it's rolling Out. I wouldn't say we have any best practices yet, Michael. It's still early. I'd say we're in year one of a several year cycle on AI agents, but boy, did they learn fast. This is new to see this happening.
- [14:54]
A
Love it. Okay, so when you and I were at Macon, you were speaking and you had identified these different kinds of AI cultures that are inside of businesses. I would love you to kind of reveal each of them and maybe we can start at the bottom of your, your thing and work its way up.
- [15:12]
B
Yeah, okay. I identified through the research and you know, I'm living it. I've attended, I attend three events per week. I go sit in the startups office that I invest in. You know, I'm in the Bay Area. So I kept seeing this firsthand and everybody has a different perspective, but I'll just share mine. So the first culture I want to talk about is called the AI resistor. It's a culture or an individual that doesn't want AI and they may actually want to thwart it or stop it. And we've seen this before, we saw this in social media, traditional media, and journalists didn't want it and marketers didn't want it. Remember that? And calm didn't want it for a while. And before that, in the Industrial Revolution, there was a group that fought and broke their actual weaving looms, right, called Luddites. The second culture is called the AI followers. Well, as the name implies, they're not going to be the first movers. You know, maybe they work in banking or in government or in healthcare and they have regulations that prevent them from adopting this because they're worried about HIPAA or compliance. Or maybe their culture is like, we just don't need to be the first mover. We're, we're already successful. Like we'll let somebody else do it and find the best practices and I want to know there's proven roi. I don't need to experiment. Or maybe they just have a deep mistrust of maybe tech companies, which is understandable as well. The third culture is what is called AI Forward. And I could see this actual term emerge in their earnings reports from public CEOs. And this is a really delicate topic. So let's imagine your company, small business or medium size or enterprise, where your most valuable resource right now is humans. Your workers, your employees, the team, the team culture. Maybe you even use the term your work family, right? Which is, you know, I don't know if that's the best thing to use, but the people Come first. Now, how would it feel at your company's culture if you said, hey guys, we all got to use AI. Oh, there might be a revolt. So the term that the CEOs are using in their earnings reports is literally, we are going to be AI Forward. Everybody has to use AI and you have to integrate it and you got to collaborate with AI. But don't worry, it's not going to take your jobs, at least right now.
- [17:34]
A
Okay, so hold on, let's pause and let's talk about these first three. Okay, sure. So this is a pyramid. For those that haven't seen. Help us describe, like what percentage of people are in each of these camps. So far we said AI Resistor at the bottom, AI Followers, and then AI Forward. Maybe you could say all three of them represent a certain percent.
- [17:52]
B
I don't have percent, so I haven't shown this really as a, as a pyramid. I think of it as a rack stank, just a stack. I could go get data, but it'd be fluctuating a lot.
- [18:00]
A
What's your sense? You know, let's just go with intuition.
- [18:03]
B
Oh, yeah. So I do poll audiences, but again, I'm skewed because I'm speaking at AI conferences, but most people are raising their hand for forward or the fourth level, which is first. AI First. But that's so skewed. Right? That's, that's. The people who hate AI are not going to come to my conference to hear me speak.
- [18:19]
A
Let's talk about this AI Forward category here a little bit. How are these businesses? Well, let's get to all of them and then we can unravel a little bit. Let's go to your next one.
- [18:28]
B
You got it. So we talked about AI Resistor, AI Follower, AI Forward number four, AI First. Now this phrase, this self description of a culture is coming out of the traditional tech companies. You know, the ones that have been around for 25 years, you know, the box.com, aaron Levy or the salesforce Marc Benioff. Both of them. Both of them have said in earnings and in their social media, we want to be AI first. And there's actual formula for this. AI first means when you have a business goal or a problem, you first see if there's a AI solution that's already built off the shelf. And then you try to use it. If it doesn't exist, step two, you actually build it. And if you're not technical, that's fine. You can use the no code tools. And there's a plethora of those and then if it doesn't, you're not able to actually manufacture it, produce it, develop it. The third step is then you hire a human, in that order. Okay, so that's called the AI first method. It's actually a three step method and that's well known in Silicon Valley. So the people that say AI first are the traditional tech companies. Finally, number five, the fifth one is called AI native. And these tend to be young startups under 3, 4 years old, maybe 20, 25 employees. And they were built from the ground up at the birth of Gen AI. Their DNA is AI. It might be some young entrepreneurs that have a degree in AI or they worked at AI lab or they worked at big tech and they started this new startup. So everything they do is already AI and they use AI for everything and they're super efficient. In fact, at my TED AI talk, I showed some of the revenue per employee of the top AI native startups and they're out producing traditional SaaS, companies 10x in revenue per employee. On annual basis, they're generating 2 to 3 million per employee. Now imagine you're a small business. By the way, the average Small business makes 200k to 400k per revenue per employee. That's like agency, can be a small consulting firm, it could be a small tech company. The top 10 SaaS, companies like Salesforce, Box, Adobe, Atlassian, they make 500k to maybe more than that revenue per employee. But the top AI native level 5, these startups, it's not their culture, they have the highest revenue per employee. There's a leaderboard out there, you can go track my data on this. It's called the lean AI native leaderboard. They're making 2 to 3 million dollars revenue per employee. So the whole point of these five cultures isn't just to put people in the buckets, it's to show that the AI native startups that are producing the highest quality and they're using AI, they're generating far more revenue than companies that are not AI native.
- [21:09]
A
Interesting. There's a moment happening in companies right now. Leadership is discussing 2026 priorities. AI implementation is at the top of the list. And someone asks who on our team can actually lead this? That's when it becomes clear some marketers are prepared. Most are not. The ones who invest in real AI training. The hands on implementation focused kind suddenly become invaluable. Here's what happens when you attend AI Business World 2026. You walk in uncertain, you walk out confident, you arrive overwhelmed by AI tools, you leave knowing exactly which ones to use and how. You come feeling irreplaceable, you return indispensable, you multiply your output, you end guesswork on which AI tools actually work. You create content faster while keeping your authentic voice. And you automate hours of daily work and you teach others inside the company how to do it. When your company asks who can lead our AI implementation, you'll raise your hand with complete confidence. Head to AIbusinessWorld live and grab your tickets today so you can become your company's AI expert. Okay, so now that we have these upper levels here, we've got this AI Forward, AI first and AI Native. These are all companies that are putting a lot of value proposition on AI. The AI First. Obviously there's a lot of pushback on this, you know, because there's been a lot of companies that have said their AI first. They've gotten a lot of negative press on this Klarna shop. There's a bazillion of them. Right. But I like conceptually, whether we decide to call it AI first or not, the way that you're defining it is first, if there's a way we can solve with AI, ideally with off the shelf solutions, let's go there. If not, let's build it. And if not, then we can refer to humans. Now for small businesses, they might not have the capacity to build it. They might just try to use AI and if they can't, then they can go to a human. I would imagine because you're in the Bay Area where all these techie companies are and you talked about no code a little bit. But not everybody has even knowledge of no code solutions. Right?
- [23:27]
B
It's true, it's true. However, I think the off the shelf versions are, it's going to increase in size. Right. So we talked about Agent AI, a marketplace of AI agents from HubSpot. Right? So that's for small medium businesses. So imagine other CRM systems and other SaaS applications for small businesses do that. That Right. Shopify has made the promise that they want to offer more AI features as well. So you can imagine that once you start to see the patterns that those traditional software companies that small businesses already rely on will have the most popular AI features already embedded. So we're still at the early era where people are having to build features that don't exist. But it'll come. But there's also this next breed right of worker that's going to experiment with no code. So I wouldn't assume that every small business doesn't know how to code there's some that are quite adept at it.
- [24:19]
A
Okay, very much appreciate that. Let's talk about some examples of AI native businesses that maybe you are familiar with. Maybe they've been at your llama event. Maybe there are investments you guys have made just to help people understand what these AI native companies are doing. Because I think it's fascinating exploration.
- [24:37]
B
Sure. So if you can all open your browsers, go to lean AI leaderboard.com and this is put together by my friend Henry. There's a few misnomers, like I cleaned out the data when I did my report on this. Like I don't think Telegram belongs on there, but you can see like the company. So it lists out the companies, it lists out a description where they're based at their annual revenues, number of employees and their revenue per employee. And again, these are all AI startups. So Mid Journey is on there, they have 40 employees and they're making 12 million employees revenue per employee.
- [25:12]
A
That's crazy.
- [25:12]
B
Wow. So that's, they're ranked at the top. There's another one called Mercor which is kind of a marketplace for AI export. So topical experts. And those topical experts are hired to train AI platforms. By the way, some people call them defectors. Right. Because they're, they're training people, they're getting paid to replace their jobs. Right. That's literally what's happening. There's Lovable, which is a Stockholm based company, which helps with no code. So those are some examples.
- [25:40]
A
Yeah, there's Gamma, which makes presentations and.
- [25:43]
B
There'S 11 labs voice capability.
- [25:45]
A
Base 44 is like an agentic solution I think. But most of these are AI startups. Do you?
- [25:51]
B
These are all AI startups.
- [25:52]
A
What I mean by that is they're providing solutions, they're picks and axes. I almost feel like they're using AI to sell AI solutions. Do you know of any? How can I say this?
- [26:01]
B
Non tech.
- [26:02]
A
Yeah, exactly.
- [26:03]
B
So a couple of startups have pitched me and they're working with restaurants and they're trying to help them automate their whole front of house and the back of house.
- [26:12]
A
Oh, okay.
- [26:13]
B
And so this company is called AIO and he's a farmer, he is a restaurateur, his name's Alex and he built this device that sits on tables and it can sense when people are done with their meal or when their glass is nearly depleted and it triggers to the waiter, waitress, humans to come refill it. And of course billing is there and the menu's there and it wants to connect all the back office with your supply chain and your billing systems and then of course tracking employees and time cards, all of that would be there. It's very bodacious project to try to do all that in one place. AIO stands for all in one. We're not investors. I just think it's interesting way for small businesses to use that. We're looking at another startup. I can't tell you the name right now, but they're building this easy to use app for non technical small businesses and nonprofits and small governments to quickly communicate with their customers. And it'll be AI enabled and create automated content so the small organization wouldn't have to worry so much about communications. So we're seeing that. I just continue to see tech companies serving small businesses. It's, it's an important part of what we're. This whole market is just making everything efficient, no matter the size of your company.
- [27:23]
A
For aspirational or entrepreneurial audience that's listening, that wants to do something and try to be AI native. How are these founders thinking differently? Like, can you give us any insights that allow them to kind of look through a different lens when they're building?
- [27:39]
B
Well, they're definitely following the AI first method. Right. And so that is the main thing that they do that's different. They try to integrate all their data as well. That's the second thing.
- [27:50]
A
Okay, so what about like their vision? You know, like vision is obviously really important, Right. I would imagine they're not constrained by any kind of limiting beliefs, right?
- [27:59]
B
Yes. So one of the interesting things about some, not all, is that there is a belief that there's a premium on humanity. What does that mean? It means that if we can automate all the labor and the repetitive tasks, you know, the stuff we don't really want to do, then you can have more time with customer interactions. You can have more time greeting customers when they walk in the door. You can have more time tailoring the customization for the customers, therefore increasing your margins or more time for customer care. You have a premium on your humanity where you are offering greater services, human services, because the rest of the operations should be automated. That's the hope and promise. So I hear that from time to time, not all founders in Silicon Valley agree with that. But those that are very B2C focused or have a services component, I really.
- [28:51]
A
Like that a lot actually, because there's a lot of us who are a little older, you know, I'm in my 50s, and we've been through multiple waves. Right. Like there was the initial Internet wave that disrupted Everything. Then there was a mobile social wave and now and There was the web3 thing and now there's AI. Any kind of mental frameworks that you are using since you've also been around for a little while or that people are using to just kind of like keep up with the change because it can be a little bit overwhelming, right?
- [29:21]
B
Of course I use all the standard foundation models and I think you should too. So I use Google Gemini, which is really improving. I use ChatGPT from OpenAI and I use anthropic Claude and sometimes I run the query through all three and to see the different permutations and where things are aligned or different so I can get better knowledge. OpenAI has done a great job becoming kind of my co. Like my co partner. So it's trained off me. And by the way, this morning I said, hey, GPT should I create a custom filter called the GPT? And it says no, you should not because I already know so much about you and it went through all my biography of all the questions and it says you're a venture capitalist, do this, you like this, you like this type of fitness. And I said, great job. You really know what I like and want. That's fantastic. So that was good by the way. Obviously I'm comfortable with my data being used that way. And then Claude, I've been using it more and more. I wouldn't say I've really been been an expert, but here's the really interesting one, Michael Gemini is killing it. Google is really doing a good job and this is why Sam Altman of OpenAI is scared. And he called Red alert this weekend because he's concerned that Google is going to eat their lunch now, which was the opposite from two years ago. So for those small businesses out there, if you haven't tried Gemini in your Gmail, it works really, really well. By the way, I don't have Google shares. Like this is just my honest opinion as somebody who's jumped into this area. So you can ask Gemini, hey, summarize my inbox or summarize this super long email thread, it's so annoying. Or help me schedule available times to have that call. Or here's a prompt, I want you to reply to Michael in a nice way that I would love to get his latest book or his autograph and it will compose that email. You can review it and then you could adjust it and then it will send it. Wow. Okay, this is great. It's as if every entrepreneur now has their own personal assistant. I love it, right for how much? $5 a month.
- [31:23]
A
It's built in with the workspace pricing. You know, it's fascinating. I've been recently on a couple of different podcasts, which I don't normally do, and recently Gemini 3 dropped, which that very same day I was training my entire staff that we are moving exclusively to Google ecosystem. And I was training them on Gemini and I was training them on 2.5 pro and three dropped. An hour earlier, I got an email from Ethan Mollock. I'm like, whoop, update the slides. But what's really cool about Gemini is the keyword is integration.
- [31:54]
B
Yes.
- [31:54]
A
So when you open up gemini.google.com and you tag, you can call any app in the Google ecosystem. You can call sheets, docs, email, calendar, you can say calendar and you can say, look at my schedule for next week and find me one hour during my normal working hours that I can experiment with Gemini. It will not just do that, it will place it on your calendar. So you start to think about this. And then they just rolled out. I just found out today that memory is now coming to Gemini. So that's really important. And memory is the thing that Jeremiah was talking about with ChatGPT. It knows so much across all the different chats and threads that he's had with it that it's highly customized, which I've got to ask Jeremiah. This is a lock in effect, isn't it?
- [32:37]
B
100%.
- [32:38]
A
Talk about what that means to people, because I don't.
- [32:40]
B
Also, it integrates Google Photos. So I'm asking Gemini now about you. And it found a photo of us of your name in Google Photos.
- [32:48]
A
Really?
- [32:48]
B
We were in a clubhouse room and for some reason I took a screenshot, so I see your name in there.
- [32:52]
A
That's hilarious.
- [32:53]
B
So, yeah, it's a lock in. So we call this a network effect or a data effect. Yeah, sorry, Blitzscaling. Where I work as a gp, there's two things we're looking for. Viral effects and distribution, and two, network effects. Network effects, this is an important term. This is a business strategy. Came out of the Stanford class. And then Reid hoffman, founder of LinkedIn, and Chris Yeh, my business partner, wrote that book, Blitzscaling. So a network effect is when every new user that joins the value increases for everybody else. Google Apps workplace has that in spades. And the second thing we look for is viral effects and distribution. Well, Google has distribution right already. It's one of the top websites in the world and all the apps and plus they have viral effects because, hey, we're talking about it right now. And plus there's viral loops and emails naturally. Right. So yes, it is a lock in. So when you have those things, and this is unfortunate by the way, I want competition. I live in Silicon Valley. I don't want to see tech giants, I don't want to see hyperscalers, I don't want to see trillion dollar companies, I want to see startups. When I'm betting on, I'm literally betting on startups. That's my job. So I don't love the fact that Google is the only like big player in town. Like we should have other players. So Perplexity also has email integration. There's been rumors that OpenAI might have a social network and maybe email. Right. To combat this, obviously Microsoft is partnered with OpenAI. Amazon has all their versions. Someday Apple will get their stuff together.
- [34:22]
A
Well, and they're all coming out with browsers too, right? Yes, that Comet and Atlas, they're trying to messaging.
- [34:26]
B
Right. And then Meta wants to play this game too. So I guess we do have competition, but it's just not a lot of startups because you don't have distribution and network effects at the level that in these companies we just mentioned. So that. Yes, you're right, it is a lock in.
- [34:39]
A
Yeah, it is, it's fascinating. But the good news about Google is it does have an open system. I mean the Chrome browser has a massive plugin ecosystem where you can actually get your software to integrate within their system and hopefully they'll start to bring that. Like Asana was one of the first integrations that we use. Asana, which is project management tool and it's fully integrated into the Gemini ecosystem. If you tag, it's just right in there, you know, and it'll connect.
- [35:06]
B
How many people work at your firm?
- [35:07]
A
Just 19.
- [35:08]
B
Well that's 19 is a lot still. Right.
- [35:10]
A
We all use Asana and it was super awesome to see that it's integrated with an tag. You just have to authorize it. So okay, you're a marketer, you have a, well, you have a background in marketing. I want to talk about marketing a little bit. AI is going to change and it's already starting to change the marketing world. And you have some thoughts on how it's changing the funnel. So I would love to explore that with you a little bit.
- [35:33]
B
100%. So by the way, this is slowly changing because Google's catching up. So over two years ago I heard this quote from Bill Gates, Bill, and it helped to shape my investment thesis and he said, whoever has the personal AI agent, that's the big thing, because then you'll never have to visit a search engine again, you'll never have to visit a productivity site again, and you'll never have to visit Amazon again because your AI agent will go do the top of the funnel, all the way to the bottom. The AI agent will do the research, do the product comparison, do the product pricing, negotiate and even do the actual purchase. And this scared Google so much that they declared a code red and even Sergey Brent went back into the office literally and is helping to lead at Gemini even coding. Right. So that's why we're seeing them publish all these new features. So I do believe this will happen. So OpenAI wants this to happen and they want to offer product recommendations. And Perplexity said that they will make it available for marketers to have sponsored answers in the results section and that's pulling it off site where you're finding products off Google. And of course, Amazon just made a press release last week saying that a lot of the holiday shopping on Rufus has increased in demand, which is interesting because then you don't have to go to Google search to find a product. Rufus is their AI assistant embedded in Amazon.
- [37:02]
A
Real quick aside, I was looking yesterday on Black Cyber Monday, December 1, for toner replacement for my HP color LaserJet and I went to ChatGPT and fascinatingly enough, the very first recommendation was not Amazon, it was Walmart. And I wouldn't be surprised if they have some sort of partnership. You might even know more than I know, right?
- [37:23]
B
I don't know anything about that, but I won't be surprised.
- [37:25]
A
But here's what we know. We know Amazon is heavy invested inside of their competition, Claude. So I'm imagining that eventually there's going to be arrangements with some of these major merchants, right? And it's going to be interesting. I don't know.
- [37:41]
B
I mean, isn't this good for marketers though? I know it sounds scary, right? But don't we need some competition? It's not great for small businesses to be dependent upon one search engine. Yeah, let's have multiple. That's better for the world, that's better for consumers, that's better for small businesses. So I think we should have a variety of ways to get information. So AI agents will help with that. And we also bet on a startup called skyfire XYZ which is partnered with Consumer Reports and also larger companies as well, where it will actually you can verify an AI agent for a consumer. So the AI agent will be verified as Michael's AI agent that's shopping for him. And Michael, you would have to show Skyfire you have a government ID to prove that you're a real human. And then you would say, yes, that AI agent, I don't know, let's just call it Frank, represents me and will go do shopping for me and can complete transactions up to this amount. And so you've heard of the term kyc, Know your customer. So KYA is their main feature at Skyfire where it will actually know your agent. Know your agent. Exactly. And so that's coming.
- [38:47]
A
Are these blockchain based solutions?
- [38:49]
B
Partially, but they also have Visa ability and credit card ability. But yes, blockchain is part of it. And the token for certifying is like a token. Right. And they are backed by A16Z and Coinbase. So you're right. But again they also have credit card. But the overall trend is the AI agent will be doing the shopping for us. I mean, have you booked a flight on United States recently? It sucks.
- [39:10]
A
No, no. I mean I'm a Southwest guy.
- [39:12]
B
All right. Even that like the amount, I mean they do a good job to make it easy. But United, the amount of clicks and pop ups and do you want to buy insurance and do you want extended, you want extended leg room for 2 inches for 50 bucks? I don't know. I don't know. Just like too many things to think about. It's so many clicks and so imagining the AI agent doing that for you on your behalf to make it easy. Right, let's. We don't have to do the laborious work, we just get to our destination.
- [39:37]
A
Well, you know what's really funny, Jeremiah, is I remember when people started first doing E commerce on the web and people said, I will never put my credit card online. Okay. There was a time where people just said, I do not trust putting my credit card online. But I think those very same people are going to be like, and a lot of us are like this right now. I am not going to trust that an agent is going to make an intelligent purchase on my behalf. I think that's a fair and reasonable presupposition today. But do you anticipate that that's going to change pretty rapidly?
- [40:07]
B
Remember the cultures? So that's like the AI follower, right?
- [40:10]
A
Yeah.
- [40:10]
B
If you said I'll never do it, I want to fight it, then you're AI resistor. So I think that's AI follower and that's okay, right? You can wait for the best practices. You don't want to put yourself at risk. That's okay, right. We have an adoption curve, like a bell shaped curve. People play in different parts of it and that's fine. But over time, that whole bell shape is going to move towards progress.
- [40:30]
A
So how does that change marketing? Because honestly, if the human is out of the loop, that's a problem. Right?
- [40:35]
B
Okay. So yes, this has massive changes to marketing. So if at the Macon event where I was delighted to speak, it means that one of the top visitors to your website over the next few years, small businesses, is not a human, it's Michael's AI agent that's certified. Right. Or it's a chatgpt that's going to refer traffic. Right. And search and get that information. It also means in our theme, so the humans don't need to go to the websites, but the AI agent will grab the information and complete those actions on behalf of the human. And then it'll bring that information back to the human user, back to you, Michael, and it will recreate it in the way that you want. Do you like audio, visual or text? You can choose. Like, do you want a lot of information? A little. You get to choose. When do you want your information? You get to choose. You don't have to see that website, even if it's got a bad UI or great ui, it doesn't matter. You don't have to go to it. So for marketers, this means you have to create content that is catering to the AI agent. It also means that the AI agent might strip away ads. So instead of ads, you're likely going to be brokering and bartering with the AI agents where the ads become offers. Another thing it means is that your CRM system will have to have a new field in it called AI agent. And there will be AI agents that visit your systems or your E commerce or your website and you don't know who they're associated with. Is it Jeremiah's agent or Michael's or Franny's? We don't know. But eventually it would say, yeah, I belong to Michael, I'm ready to make a purchase. Then you would append that in the CRM. Right, so that's another change, right? Operational.
- [42:12]
A
So does that mean we could make special offers to the agents then? Is that kind of where we're going with this?
- [42:16]
B
Yes. So let's go with that thread. So it means that marketers will offer their own AI agent to communicate with human visitors and also an AI agent to talk to the buyer side AI agent as well. Right. So basically the big trend here is the presentation layer. Whether it's your website or app and the data, the content, they're decoupling, they're separating, they don't need to be together. Right? So that's the big change from marketers that they need to prepare for that. So now, one caveat. If you have an amazing brand offering, like it's something people want to lean in, like luxury or entertainment or sports or chat or news, people want to visit those websites then, or gaming, right? Or adult content, people are still going to visit that stuff, Right? So, but the all the other stuff, you know, filling out expense reports, I'm going to have my AI agent do that. Booking my flight, I don't want to do that. I'll let my AI agent do that. Shopping for my aunt's new sweater, I don't need to do that. I'll let my AI agent do that. So I think that will change depending on which industry you're at. What do you think?
- [43:22]
A
I don't know, man. It's like, how do we even prepare for something like this is the first question that comes into my brain. I mean, we're going to need to have insights that are going to need either come through cloudflare or Google Analytics or some other tool that's going to be intelligent enough to know when an agent is coming to the site and how to deliver the content to the agent in such a way that strips out all the unnecessary stuff.
- [43:46]
B
So I don't think you have to think about it that deeply. So any CMS system, like the data is already, like it just won't show the HTML and it could just be a text file. Right. So I don't think you have to overthink that.
- [43:56]
A
Well, how will you know though? How will you know as the.
- [43:59]
B
So Skyfire has partnerships with those bot blocker companies, the ones that compete with Cloudflare, actually.
- [44:05]
A
Okay.
- [44:06]
B
They already have two or three of the partnerships, one with Data Dome. So once you have the verified agent Kya, then when they visit your website, then you can say, ah, this AI agent represents Jeremiah and so that's verified. So that exists.
- [44:22]
A
And then also with the large language models like Google and ChatGPT, there's always the question of how do we even get our stuff in there to be discoverable in the first place? Right?
- [44:34]
B
That's a whole nother topic.
- [44:35]
A
Yeah, but sometimes those things, sometimes people start there right before they make a buying decision. Right. They don't even know what they want. They're going to do some research in these tools. Any thoughts on that or is that Something outside your purview?
- [44:48]
B
No, it is tied. Okay, so there's a new term called.
- [44:51]
A
There's a bazillion of these terms. Yeah, it's called like AI Optimized or something like that.
- [44:55]
B
Yeah, there's. There's a couple of them. So you know Pete Blackshaw, right? Former word of mouth woman guy. And then he was Nestle CMO. P&G. CMO.
- [45:03]
A
I feel like. Yes.
- [45:04]
B
So he's a friend of mine and he started a company, obviously. He's this marketer called BrandRank AI. Oh. And they've demoed at Llama Lounge. And so Brandrake AI is doing that for marketers. And by the way, they have competitors. I just don't remember the name. Some of them were at Mekon, by the way. And Brandrake AI. You're a marketer. Brand marketer uploads their company name and product name and then it tries to figure out what does GPT think about it, what does perplexity think about it? Anthropic, Mistral, Japanese Sakura, like all of them. And tries to understand where are those large language models getting that data from? And then from there you can understand, oh, they're scraping Consumer Reports. Oh, they're scraping Wikipedia. Oh, they're scraping Reddit threads. And then once you have that information, you send in your content teams and your marketing teams to go create that content or reproduce that content or influence that content, whatever you got to do. Right. So that's how you unravel that. And it's the same exact strategies that we saw with SEO right now. It's just applied to the large language models. So that's a whole little industry, cottage industry that's emerged for this.
- [46:13]
A
Do you feel like marketing is going to become less or more important in a world where AI is everywhere? What's your thoughts?
- [46:19]
B
Less important. And I'm a marketer and you can see, right? Because I mean, even with the layoffs and marketers are struggling just to hold together, like it's just catastrophic. And the agency market is just turning upside down right now, by the way, brand doesn't go away. Brand still becomes important because the AI is going to be doing that.
- [46:37]
A
Content marketing might be more important than ever. Right. Creating videos and writing and get in front of a human audience. If you still have a human audience that's making these purchase decisions, I would imagine would be important.
- [46:46]
B
I still go back to that. There's a premium on humanity. Now. Is that part of marketing or, you know, is it. That's customer experience. Right. User Experience, Right. Or brand promise? Is that marketing? Maybe, maybe not. Is it customer care? Could be. Is it brand fulfillment? Yes, could be. Is that part of the brand experience? Yes, of course it is. But I still think like we're going to see some, some new experiences emerge, but I think the traditional way, search engine to E commerce click, I think that's not going to be the same and I think we're going to see more direct routes. All those steps in that funnel, I think become collapsed in a different way. With AI, it becomes more immediate. So we can skip a lot of those steps. I was asked to speak at one of the big marketing agency conferences, you know, the one in France, and they are telling me like the budgets are being slashed, marketers are freaked and they're struggling to be relevant in this space right now. I think that's going to continue for the next few years.
- [47:42]
A
Jeremiah, you have a great pulse on what's happening in the future. You've always been an analyst that's very much on the edge of everything. And I first of all want to say thank you for bringing your insights to us today.
- [47:52]
B
My pleasure.
- [47:53]
A
People want to connect with you on the socials. Do you have a preferred platform? And if potentially they want to get involved with the Llama Lounge or any of that other kind of stuff you got going on, where do you want to send them?
- [48:03]
B
Oh, cool, thanks. Yeah, so I'm publishing on LinkedIn probably the most and then I have a blog and a newsletter. You can find them. There's links out there. The good thing about my super long name is there's only one Jeremiah Ow Yang out there. So I'm so easy to find in my emails and my DMs are open so you can hit me up at any time. Thank you, Michael.
- [48:21]
A
Jeremiah, thank you for properly pronouncing your last name, Jeremiah. I've been saying it wrong all these years.
- [48:27]
B
Oh, it's okay.
- [48:28]
A
Don't worry. Really appreciate your your wisdom today.
- [48:30]
B
Thank you for having me.
- [48:32]
A
Hey, if you missed anything, we took all the notes for you over@socialmediaexaminer.com A85. Be sure to follow this show on your favorite podcasting app. If you've been a listener for a while, we would love a review and you can share this with your friends and if you want, you can tag me on Facebook, LinkedIn and or X. Do check out our other shows, the Social Media Marketing Podcast and the Social Media Marketing talk show. This brings us to the end of the AI Explored podcast. I'm your host, Michael Stelzner. I'll be back with you next week. I hope you make the best out of your day and may AI help you become more successful. The AI Explored podcast is a production.
- [49:08]
B
Of Social Media Examiner.
- [49:12]
A
Get your tickets to AI Business World right now by visiting AI businessworld Live.