Nvidia CEO Warns: Every Company Needs an AI Agent Strategy
Eric Siu and Neil Patel discuss autonomous AI agents and how companies should build strategies around them. They demo tools like Instantly and Karrot for automating marketing, sales, and recruiting, share results from autonomous outreach campaigns, and warn that agents still need human oversight to avoid security risks and low-quality leads.
Chapters
Jensen saying that every software company in the world needs to have a claw strategy. Okay. And I do think a lot of companies are converging into becoming software companies. Not— I mean, physical companies, maybe not so much, not like LVMH. But I think that part is interesting to me because I just, I see the evolution in the company right now where people are just, they're talking to these claws like they're actual partners and getting things done. Again, data pool strategies, actual execution. And so I think that the fact that he said that and he spent 17 minutes of his talk time on it, I think is really interesting. And I don't know, like, what are you feeling about this right now? Have you played with it or no?
I have not played with it personally. I've seen it from my team and them sharing the results. There was a marketing tool that went viral on X, I think in the last week or so, had 12 million views when I saw it. I was traveling, but I saw it. I know something that are like first AI CMO, it'll do all your SEO for you and all this stuff.
Oh yeah, yeah, yeah, yeah.
It broke a lot, so people weren't happy. You can play with it and judge on your own. But I was talking to my team and I've been talking to my team about this narrative since middle of last year, and I keep telling 'em, you know, and OpenCLAW does it 'cause OpenCLAW actually takes action for you. Like people talk about agents, but they don't understand how agents work or practical use cases. OpenClaw can go and do everything for you as an agent, right? It just goes and executes and then it brings you back the end result. People, when they're paying a company or paying a software solution, I believe they're tired of dashboards, they're tired of reports, they're tired of analyzing things on their own and figuring out what to do even if you give them a list. They just want more of a chat-style interface being like, and listing out, Here's what's wrong, we're going to fix this in this priority, here's the estimated results, is this approved for us to fix? Yes. Oh, we can't fix it because we're not connected with your WordPress, click here to connect with your WordPress, boom. Oh, we want to track your results, integrate with your Google Analytics, click here, we'll do the rest for you, and it just goes and does everything for you. I think that's really the future of a lot of the internet and businesses and even marketing. I think like, You can pay softwares to do outreach for you. Like for the Writesonic guy, founder was showing me how he's using OpenClaw to do outreach for influencers. So he hit up 400 and something influencers. It would bid on his behalf. He gave it a max budget. It would have to go to him for approvals all the time. So it was going back and forth in Telegram. He got 5 approved out of it. The campaigns went live. I don't have the results from the campaigns. He didn't share that, but he shared how he did all of this with OpenClaw and it was pretty quick. And you're going to start seeing more things like that. Now that result is a little bit skewed because there's already software companies that do that for you before OpenClaw that you can just pay $20 for per month. And he may have spent more than $20 spinning this up with servers and stuff like that. So I think at the same time, even though we're looking to automate a lot of marketing, people need to also not get carried away and reinvent the wheel through through AI when there's already solutions that do it for you that are cheaper. And I would recommend doing that than running an OpenCL instance and having to deal with bugs or issues or anything like that.
Yeah, that's actually a key thing. So we actually, even though I talk about this stuff, it's like, hey, if you can find a solution that's $50 or $100 and it does it, like, don't try to engineer this stuff. And so, but for the things that you don't want to do, so I'll share my screen real quick. So I was playing around with this last week, I think. So what I have on the screen over here, this is Instantly, which allows you to create email inboxes, allows you to send emails, things like that. Just think of it as email infrastructure. So I was like, well, I don't want to make all these email domains. I don't want to figure out what's available, what isn't. I also don't want to come up with the names as well. So this actually made— it's like it came up, it figured out the naming convention. So we have the company name and also like a modifier at the end. It was like clickfollowgrowth.com, carrotreach.com, And it went, it bought all these domains, okay? It started warming them up within instantly. And not only that, Neil, it started to write the sequences too. And so it's like, okay, hi, first name, here's what we'll do over here. We'll analyze your presence at no cost, whatever it is exactly, want me to run the analysis? And so it's just testing this. We're starting to test it as of Monday on this one, but it's already working for us on the recruiting side. And so my point is, We're talking end-to-end solutions. Neil just mentioned it. The key thing here with any of these CLAWs is that it can do the work for you end-to-end. It's not only going to analyze for you, it's not only going to do a data pull, but if it can execute for you end-to-end where you just have to review the work, that's why Jensen is saying this is so important because that changes how we work across the board. So Nemo Claw, Open Claw, whatever it is that you use exactly, that's why he's like, it is the most important piece of open source software ever. And that's from his mouth. Not mine. And guys, test at your own risk because you don't want to burn everything. So don't just connect it to everything and like give it all your financials and your credit cards and all that. That's not going to be good for you.
So instead of just taking his word, it's not just the most important open source project out there. I believe the last time I checked, it's the most popular one. It's used more than anything else.
Yeah. So there's a chart where it shows like the number of GitHub stars, right? And it shows like Like 2, I forgot what the 2 major GitHub things are, but OpenClaw is just like a line that comes out of nowhere and like surpasses them. So yeah, yeah. OpenClaw GitHub star chart versus others. Let me see if I can pull it up for you.
Dude, it's so popular. Same with OpenClaw. Uh, not too many people talk about it outside of the United States from what I've seen.
Dude, you know who, you know which country is crazy about it right now?
Brazil.
China. They literally, I saw, I saw like it was like a park and everyone had their laptops out and they're on OpenClaw. Like, isn't that crazy?
That is crazy. I didn't realize that. I know quite a few people who use it in Brazil, but we have a team out there, so I'm biased with Brazil.
Check this out. So, um, here, here, here's what I wanted to show you. Okay, so this is OpenClass spikes to 250K+ stars, Linux and React as context. So you have Linux, which is in the yellow line on this chart. You can see it growing steadily over time, right? It goes, goes, it's a hockey stick.
Okay.
It gets up to like 200K stars or so. Then you have React, which is from Facebook, and it pushes up here, and then it's like 250K or so. OpenCLaw. It's not even a hockey stick. It's just like, it's just an eye. It's not even a hockey stick.
It's just straight up a line going up for anyone who can't see. All the other ones are hockey sticks, so it's going up and to the right, like a triangle, side of a triangle. This is just a straight line up. Straight up, like, you know, 90 degrees. Yeah, straight up.
And this is what Jensen is saying exactly. You can do more with more. And so again, if you're someone who is very curious about the world, you're about to become way more curious. Um, anyway, that's enough about NVIDIA, I think, for today. I just thought it was very exciting that, uh, that they pushed this out so quickly. And it's— he's so smart to have partnered up with OpenCLAW on it, um, because I thought they're gonna make their own, but the fact that it's partnered up, I'm like, that's, that's smart.
Dude, I thought the same thing too. I thought he was going to create his own. On a side note, did you see the Amazon and Microsoft fiasco?
No, what happened?
You know, OpenAI is doing a big partnership with Amazon and they're not allowing them to use APIs. Yeah, APIs and Microsoft deal. I'm like, uh, someone's gonna get sued for this one. And Microsoft doesn't mind suing. They have so much money in the world, they'll sue and they'll make a killing.
It's crazy to me. And like, by the way, like Elon tweeted out, he's like, hey, whoever, like if I win this OpenAI lawsuit, like I'm just going to give the money to charity or whatever. But he just wants to do what's right. And so I don't know, there's a lot of weird stuff going on, I feel like, with OpenAI. So yeah.
Yeah. On a side note with marketing, and I know we're talking a lot about AI, if one thing you really should take from Eric and I, is as you become more efficient with your marketing and AI, do more with that extra time. If you don't, you're going to probably eventually end up getting replaced or not do as well or not progress in your career as much, and we don't want to see anyone fail, of course.
Yeah, yeah, so I guess you're right, we should try to tie this in with marketing more. I will say like all the things I'm doing right now within the dot charts, like I have this Chart Factory thing, you know, it just, it ingests all my content, And then it just makes those dot charts that you see on my Instagram. And some of them do pretty well. Like some of them, it just goes and publishes it to LinkedIn. I see people fighting in the comments on like, oh, it's not about IQ, it's about this. And I'm not even engaging.
So, okay. So do you want to show people your dot charts and how you do it?
Yeah, sure.
And how you use AI to create this content for people on marketing?
Yeah, this is funny. Let me see if I can dig the one up where everyone's like fighting each other. So here, let me show this one first. So, all right, sharing screen.
Do you see this?
Not yet. Do you see this?
Now I do.
Okay, so this is a dot chart that was created completely from AI. And all I did, Neil, with my claw was I said, I really like these dot charts on Twitter. How can you apply this to us for my content? That's all I said. And I was like, you already have a content ingester where you pull from my content every day. What would be relevant? So it actually pulled a study from McKinsey on 1,300 companies, and it shows 88% of these 1,300 companies use AI. Only 6% get real value from it. Less than 1% are running autonomous systems. So this publishes, okay, and it gets 56,000 views on it, which is not bad for an Instagram post.
That's great.
Because this is the type of stuff, by the way, it's not like, oh, it's like a dating post or something like that. We are talking companies using AI. And we're getting more and more companies on my side, at least on the AI implementation piece. So that's one. Now, I want to show another one. Okay, this one's great. Let me pull this up for you. Let me move to this screen. Do you see this?
Yeah, I can see it. Your LinkedIn.
Okay, so this is a LinkedIn post. So this caption, by the way, let me just read it. I'm not gonna read the whole thing, but AI went from an IQ of 83 to 128 in 3 years. Yours didn't. Let me put that into perspective. November 2022, GPT-3.5 launched, IQ around 83, below average, below human average. People called it a fun toy. March 2023, GPT-4 hit IQ 105. That's dead center human average. People started paying attention. 2025, okay, it just goes down this right now. We're looking at it like IQ 145 plus. 3 years from toy to genius. So the question isn't whether a AI is smarter than you. It already is in measurable ways. The question is whether you're using it or competing against it. One of the paths has a future, the other doesn't. So then the graphic that you see here, Neil, AI's IQ is growing exponentially, yours isn't. So it kind of shows a chart over here. And then all these people in the comments, 113 comments over here, okay? 78 likes, it did fine. But all these people start arguing about IQ is not the right thing to measure, IQ cannot solve this, blah, blah, blah.
And I'm just like, I'm not even going to engage with this bullshit. No, there's no point. Even if you just use stats and data, people will tell you how you're wrong. There's just no point. But let's go into how you created that. So let's go, let's talk about workflow. Let's say someone wants to end up creating social content here. This chart here, I'm assuming you had AI create it. You didn't pull someone else's chart.
Yeah, and it's doing this, uh, this one. So what it does, Neil, it will look around on the internet and it'll cite— I have it cite the sources at the bottom. Okay, so, you know, by the way, uh, let me, let me give you a little bit of the workflow. So imagine you and I are talking about stuff because you and I share a lot of news and we'll share like what we're testing. So it It will take, it'll ingest all of our content and based on a claim that you or I make, it'll try to do more research on it to back it up and then it'll make it into a chart.
And you're using what for that? OpenClaw?
Yep, I'm using Claw. Yeah, OpenClaw, one of the Claws.
All right, so you guys can try that out. My process is a little bit different. We end up taking, so instead of taking recordings like that, We take what we get asked frequently from potential customers and we have a repository of that. It could be audio, it could be, you know, people writing it down. Audio's easy for AI to just scrape from there. And then from there we end up doing research and then you can create charts on it. I've mentioned this in the past, we've tried using AI to help with the charts. It's not always 100% accurate, so we tend to have a lot of humans in the loop. AI does help with a little bit of the design, but— We have humans in the loop because we just don't want the 1 in 5 or whatever errors, because then people just chew me apart anytime anything is a little off.
So, Neil, as you say that, I literally typed that into my Open Claw. I'm just like, go do that. You see what I mean? And you know how I like— I'll see what people post on Twitter now, or I see like a YouTube video. All I do is I just post that YouTube video in. I'll say, hey, how does this relate to our goals on a scale of 1 to 100? How relevant is it? Oh, it's very relevant. Okay, go do it. Most of the time, I would say 9 times out of 10, it's like, this is not relevant right now. I'm like, okay, good. Good to put me in my place. But sometimes I'll be like, oh, this is like a 95 out of 100. I'm like, okay, let's have a conversation around this. And literally all I'm doing, I talk to this more than I talk to anybody. And when I'm pointing at my phone right now, I mean my claw. All I talk to is Alfred. That's all I do all day. Coming back from the gym, talk to Alfred. Sitting, waiting for the restroom or whatever, I talk to Alfred. That's all I do.
So Eric is overstating it here. He does have a personal life and he does other things than talk to his phone and OpenCLaw. Yeah. But yeah, it's super efficient.
I just, I can't wait for you to activate it because then that's all you're going to talk about with me. You're going to call me all the time. Maybe you're not going to call me anymore because you're just going to want to talk to yours. I don't know what you're going to name yours. Maybe Beverly.
I know you don't believe it, but I don't think I'm going to end up ever activating it.
I don't believe you.
You don't think so?
This one I'll bet you my life on too. I've been willing to bet my life on very little things. This one I'll bet my life on.
If I look at my life each year, it's changed to progressing to just like focus on like extremely really large deals and brand building. So I agree with you, I would be a really high user of OpenClaw, if it was 6 years ago, 7 years ago. It's just what I work on today has shifted and it's getting more extreme to that way where it's harder for me to use tools and technology to help automate what I'm doing because it's very heavy relationship business. Like it's getting to a point where I don't even do outreach anymore and I'm just getting set up and just doing the calls myself and just building relationships with people and just trying to work big deals and that's it.
I completely understand. And yet you still use it. So we'll see. We'll see in a year.
What way do you think I'll use it? Give me an example.
I don't know. I just know you're going to use it because here's the thing, right? You just saw the chart. You saw the data. You're a very data-driven person, right? And you hear some of the richest people in the world saying like, this is like the biggest thing. And like, you know, every company needs to have a strategy. I think eventually you'll come around to it and you use it in your own way. Um, like for example, like, uh, I can only talk about me. I can't talk about you, right? But the things that it does for me, Such as how it negotiates deals for me on the— because everyone tries to do these. By the way, I think I'm getting 8 to 10 inquiries a day because I think these other influencer managers are using AI as well. And so I think I'm having the AI negotiate with them and that is real revenue, right? But even when I have it look through my LinkedIn or say, hey, here's who you should be talking to automatically, or look at my calendar. Hey, you're traveling over here right now. Here's who you should be talking to based on your goals over here. Maybe you should set up a dinner for this or that. And then maybe your team can take that over for you. But I think you're going to be able— like what your team's doing right now for you setting them up, I think you can do even more of it. That's my point.
No, but my— so I think you nailed it. I'm not saying my organization won't use it. We already are using it. We do it for outreach for influencers. We do it for outreach for me in meetings. It's just more so I personally don't have to download it and install it. My team is already executing on it and they have been for a while now. If they didn't, I would be pissed off. But it's just, I don't think it'll come to a place where I need to install it on my phone or my computer. It's more so my team's just going and executing it all on my behalf.
So I think you're going to be— the reason I say you're going to be using it for sure, maybe you're not using OpenCLaw, but you're going to be using a version of it because everything's kind of becoming like an OpenCLaw. So like Perplexity Computer or like Mantis's like new computer use or whatever. So you're going to be using a version of it because it's going to be built into whatever you're using and you're just going to use that. That's what I mean. That's true.
From Gemini and all those solutions and all that kind of stuff. And the real reason why I just don't touch OpenCLAW, I don't want to deal with any security issues.
Yeah.
That's the real reason, or else I would already had it on my phone.
Yeah.
But the, the, um, NVIDIA version, I think it calls it NemoCLAW or something like that.
Yeah.
I will end up using that because it'll be secure and for corporations. I just don't want to use OpenCLAW myself. Until they figure out security issues. I think you nailed it. It won't be called OpenCLAW, but it'll be a quote-unquote version of that in whatever LLM I'm using because it'll already be integrated.
Yeah, and you will love it, and we will be talking about it all the time. You know, I just came from a podcast interview with Andrew Warner, so I was just sharing all the AI stuff for like an hour, and he was telling me, uh, Heaton Shaw, your brother-in-law, texts him all the time on how he can't get off of it, right? But Heaton's more similar to me in that respect, so.
Yes. On that side, from a personality standpoint, I don't think I'll ever be a power user like you or Hiten. I just don't see it.
That's your personality and that's fine.
Yeah. It's like for some reason in business, like my thing that I love more than anything is the deal. Like I just, I'll kill myself for the deal.
Have you read that, The Art of the Deal?
I have not.
You know, I actually have it. I actually, because it's all about the deal with that book, obviously. So I'm actually going to read it because it's definitely The fact that you brought it up, I'm like, I've never read it, but I have it in my bookshelf.
Dude, so I was talking to Ethan about marketing a few months ago and he was telling me about all the stuff he's doing with his MacBooks and how he spends so much per computer and he orders a ton of them and all this kind of stuff. And he was just like, business is going to change, marketing is going to change. And he's like pushing me and here he knows my personality. I'm like, yeah, this is fine. And he's like, are you going to do it or not? Personally, not so much, but I'm like, I'm definitely hiring a lot of people who are going to do it for me.
And how did he react to that?
He's just like, he's like, that sounds about right. That is more my personality.
Yeah. And that's fine. We're all different, right? So Neil and I, we're not saying like there's any right or wrong way to do it. I think we've figured out by now, like, you know, approaching— well, I'm approaching my 40s, but in our 40s, I should say that we are how we are. And, you know, I don't think there's any right or wrong way to do something.
So yeah.
Yeah. By the way, go ahead, dude.
I think someone should end up doing in marketing. Um, and I was talking to, uh, the WriteSonic founder Sam about this. I was like, someone should take everything in marketing, like every single thing that someone does, uh, create tons of SOPs. I know you're already on the process of SOPs. automate as much as possible, adapt it to that organization, and package it up into one interface that just makes it easy for people to do every little thing.
Yeah.
Because the hard part in marketing is, and the reason I didn't say package it up for everyone is you can't. The way you do marketing in an agency that's an SMB agency is different than the way you do marketing for an enterprise agency, even though they're in the same space. The way you do marketing for a D2C e-com business that's doing $1 million in revenue is very different than a D2C business that's doing $100 million in revenue, right? So even if you're selling the same product or same service, the way you do marketing is a little bit different. So ideally what you need to do is optimize and automate as many systems and processes for that one company, for the type of products and services they specifically sell. I don't think— we were talking about how effective it would be to just generically, you know, create tools that do marketing and it works for all organizations, I think that'll fail and it just won't do a great job.
Yep.
But if you optimize it per vertical, per company size, per the way they do businesses or business, it'll work well.
Just a quick break. If you want to run LinkedIn personalized ads and landing pages, you can't do it on LinkedIn ads right now. But you can do it with Karrot. It's www.karrot.ai. Again, karrot.ai, and we'll see you on the other side. So I'll show you something over here. We have this thing that we're working on right now. So we have a dashboard here. So those of you that can't see it, it's called Single Brain. And within Single Brain, this is focused as like a B2B dashboard. Okay. So we could have it focused on sales agents, recruiting agents, SEO, and all that. So This all lives within the command center here. Now, within this, I have this experimentation tool called Auto Growth. What Auto Growth does, Neil, is for, let's say you're sending a ton of cold emails out there, okay? And you want it to automatically experiment on different angles, different hooks, different body offers or body copy or different offers in general, right? So this is what it does. Like, this is actually set— this is working off the emails that I'm sending right now. And it's compounding. And like once it learns, it's going to try new tests. And once a new variation works, it's going to move over to that one. So imagine that whatever you're spending money on, as long as you have enough data that you're collecting, like you're spending more money or you're sending a lot of emails, this auto growth engine, autonomous outreach optimization, that's what it does. And it was built off of what Andrej Karpathy did with his auto research, which helps make quick improvements from a code standpoint. You can run continuous experimentations from a coding standpoint. All that to say is we have a dashboard, we're focused on B2B, and then we're trying to have that all in one where people can just kind of check their metrics in there and then the thing just kind of grows the business for them. That's the dream. We'll see what happens.
So we do a variation of that. So if you just take the email example for growth and sending out cold emails, Carlos and Caio on our end ended up creating that a while ago. The key though is you have a human, at least what we found to get better results, in charge, looking at it every single day and helping fine-tune it. Because if you just let it run wild, we found that the results aren't that great versus having a human in the loop and giving it feedback on how to adjust.
Yeah.
When we combine both those two, we found it to work well. But here's the caveat on this. At first, when we experimented and had someone in charge, we just put a random person in our organization like, hey, you know, like Marina, can you check this out and just monitor this and make sure it does a good job? Well, if Marina doesn't specialize in email outreach and hasn't done that for many years before, the results aren't going to be as great compared to like Carlos, who's sending emails every single day and has been doing that for more than 10 years. He's figured out what works, what doesn't, gives it direction. We found that it just learns that much quicker and produces better results so much faster.
Yeah. So you definitely need a human in the loop, but I think what to call out from this auto research from Karpathy, which just came out in March, by the way, is He just open-sourced how Frontier AI Labs do research. And so you couldn't do this before, but now all these people are building something on top of this auto research. But now it's like research is now entirely the domain of autonomous swarms of AI agents running across compute clusters, megastructures in the skies. The agents claim that we are now in the 10,000th generation of the codebase. So he iterates on the codebase. In any case, no one should be able to tell. No one could tell if that's right or wrong, as the code is now a self-modifying binary that has grown beyond human comprehension. The repo is the story of how it all began. So this is what he built. I would encourage you all to take a look at it and then you can build your own versions. But our version is just called Auto Growth. And then we're still— it's still early for us. We're testing right now and I'll report back on the results. So go ahead, Neil.
Is this tied to the Amazon gift cards that you're giving away?
On Autogrow?
Was it Amazon or was it AirPods or you were giving away something for calls?
Yeah, we do, we do multiple. So one is Amazon gift card and then one's just like a straight up call. And then the other one could be like a 30-day trial of like one of the softwares that we have. Yeah.
Uh, what's the call version?
The call version is like just what, oh, we, we, we did an SEO audit for you or an AEO audit for you using ClickFlow. Uh, cause we have like all the tracking in there and stuff. So yeah.
Oh, you mean like you just give them a call, like you're not giving them anything?
No, no. So yeah, yeah, exactly. It's like we will hop on a call with you 15 minutes to show you the audit, to talk it through if you want. Yeah.
Got it. And then I'm assuming you're using AI to do this at mass scale. What have been the results so far? Because you've been doing this for months now.
Well, the calls themselves work. I would say that maybe let me do the math on this. I will say on I think on 80,000 emails that we sent, okay, this was using the cold email provider before. We closed, was it 5 things from it? So that was, that's what we closed from it.
So 5 deals. What was the average revenue per deal?
Let's call it average revenue. Let's say each one was around $25,000 a month or something like that.
So that's not bad.
It's not. And so now what we have running as of, you know, very recently, is the things that I just showed you built out with Instantly and this auto growth stuff. And if before, if you were to ask someone to do it on the team or like you'd have to hire a whole company to do it, now I just have the claw set it up and then we get going on it. But obviously there's a lot of infrastructure stuff that needs to be set up correctly.
Yeah, the big thing that we found is when we have to automate a lot of emails in mass quantity, and then before you get to the closes, you have people jumping on a call and there's all these processes and trying to close a deal. What we found is we're, even though it's getting a lot of appointments, it's getting lower quality appointments. And we're finding that our sales team is spending more time trying to close people. I haven't run the math on that, but that's part of it that we include in the calculation because if something gets you a lot of calls and then you get closed deals from it, you just have to calculate your costs. But what we found out with a lot of the AI emails yet, and we just need to fine-tune more on our end, is instead of just cold blasting out tons of emails, it's a similar model. You're giving criteria, but we're finding that a lot of the calls that it's scheduling with us, it's doing a poor job. And I'll give you a very specific example. So we hit up a D2C company through cold email, and I ended up joining one of those calls as it got closer to pitch. AI had specific job titles to go after— VP of marketing, CMO, director of marketing of a specific channel, whatever it may be. A lot of the people that we got our calls on did not have that title. And the big issue here was they may have had a similar title for a past role, But when we tell it to focus on their current role, it would mess up on what their current title is for their current role versus what some of their past titles were. So it was scheduling calls with the wrong people within the organization.
Yeah.
So you actually—
so I had a follow-up point to this one, but I forgot what I was going to bring up with this one. But I think We talked about this actually from lunch last week. It's like, I think we're going to see a lot because of ElevenLabs and how good these voice models are. I think we're going to see a big rise in robocall spam, texting as well. We're going to see a lot more of that too. But I would just say that with the outreach stuff that we have going on right now, the whole premise of auto growth is that You'll have a lot more experimentations running for you. So the math that you can do on this is like, if it's a good growth team, maybe like human growth team, maybe you're running 50 experiments a year as a growth team, right? Obviously, each individual is probably running more experiments. But if you have this thing that's constantly running experiments and you're driving enough volume, if you can run 100 experiments a day or something like that, I'm just making numbers up. If you can truly run 36,000 experiments or micro experiments, How much faster is your company going to grow? And that's— it remains to be seen. But with how Karpathy is talking about how quickly they're doing AI research, my hope is that can carry over into marketing, which is why we're doing it that way. So we'll see what happens.
Yeah, I'm just hoping they can fix the little issues with these LLMs. Like, you know, when you want to get cited, a lot of companies are paying for articles and paying for placements on other websites, and they'll talk about how they're the best solution for X or the best product for X, and it's getting them cited by more by ChatGPT. The problem right now that you're seeing with ChatGPT and Gemini and a few of the LLMs, they can't decipher if the content that they pulled was paid or not. Shockingly, I don't know why they can't.
Yeah.
And it's causing a lot of people to pay and get results in organic when they shouldn't be able to get those results. There's a lot of nuances like that, similar to the example I gave you when I'm asking for specific job titles. It is pulling job titles, but that person could have had that job title in the past, even though we said, tell us so many times, no, it needs to be its current job at the current company they work for, title, it just messes up on little things like that. Ooh.
Yeah. Okay. You gave me 2 good ones. So there's, I remember what I was going to say, and then I'm going to pull up the, you know, how this guy's like hacking ChatGPT or Gemini rankings. So we can talk about that next. So the sales workflows that we have right now, so Aero is our sales agent. So it will surface cold emails to target like cold companies, like based on our ICP. That's one that's not necessarily new to anybody, but the second one is Deal Reviver or Deal Resoruptor. So obviously, you know, you're going to lose deals, right? But oftentimes, if you come— if you're coming in number 2, you just have to find people at the right time. Maybe, you know, in 3 months to 6 months, again, you can resurface these. So we have a Deal Resoruptor, but we also have one, Neil, for any inbound contact that comes in, whether they come into the email list, maybe they watch one of the webinars. What happens is it will try to see if they're an ICP fit and then reach out with a custom message As it relates to their role and something that, that like some type of custom, like personalization in the beginning, right? So the whole idea here is now, now that you have these CLAWs, you can build these different sales workflows and then that enables your team to get more done with less. So anyway, all that to say, guys, that is it for today and we'll see you tomorrow.
