What Founders Can Learn From Students Cheating With AI

0:00Eric Siu:When a buyer asks AI for a solution like yours, does your business come up? Most companies have no idea. And by the time they find out, they've already lost the deal to someone who did. HubSpot AEO helps you show up in those moments with the right answers buyers are looking for. Before the first click, before the first form fill. That's the moment HubSpot AEO is built for. Check out HubSpot.com, the agentic customer platform for growing businesses. These seven prompts cut our ad spend by 52 % in 60 days. I'm like, oh, that's totally AI generated. I'm like, what are the prompts? I don't know. I don't remember any of these.

0:35Eric Siu:So here's the thing. Let's talk about prompts for a second. I think having to remember prompts and like saving all these prompts is theater. I think it's unnecessary to have this stuff because I don't remember the prompts. You don't remember the prompts. And the agents themselves are getting much better at prompting. So really, it's just about studying, doing more than anything, and just constantly experimenting.

0:56Neil Patel:But I agree with Eric. But back in the day when you started seeing ChatGPT first come out, prompts were really important. It's not the same anymore. And the big mistake we're seeing people make now isn't really related to prompts. It's actually what data are you feeding them before you do anything else. And I'll give you an example of this with a homework because I see a lot of kids these days cheating. And I was talking to some students. I was speaking at a college. And when I say cheating, they were using AI to do their homework. And they were telling me how AI does a terrible job. And I was breaking up.

1:34Neil Patel:I'm like, you're thinking about it the wrong way. And I was like, why does it do a terrible job? They're like, if I ask it 10 questions, you know, one or two of them will be off. And I was like, before you actually have AI help you, are you actually telling it what to pull from, what to read, what's your test based on? And I'm like, if you tell them where to look for the answer and only look there, instead of eight out of 10 or nine out of 10 being right, you'll probably get 10 out of 10 right because you're giving them the exact source to pull from and you're telling them to only pull from that source and nowhere else.

2:07Neil Patel:And I'm like, then you'll start getting all the answers right.

2:10Eric Siu:And they're like, oh, never thought about that. You know, I just realized something as you're talking, Neil. AI is not just, we're not really talking about intelligence amplifier because kids, there's a lot of intelligence kids. I actually think AI, IA, intelligence amplifier is now JA, JA. It's judgment amplifier because children, they don't have good judgment yet. They don't have good experience yet, right? But they might be very intelligent, but they don't have good experience. And so the people, let's say you have people on your team, they have poor judgment and they use AI. Well, if they have poor, what was that?

2:42Eric Siu:BJM, BJ, right? Bad Judgment Amplifier. Yes. Bah, right? Whatever you want to call it. So, yeah. Anyway, I think that was interesting. Okay, so let's move on to this one now, okay?

2:58Eric Siu:Here, you pick one of the hooks. I was going to go with this hook. I wasn't sure if I wanted to use it.

3:04Neil Patel:Three roles on your marketing team that won't exist in 18 months. If you're in one of them, listen up. Okay. the first role that won't exist in 2027 is a data analyst someone who is actually just looking at spreadsheets analyzing data and telling you what to do with that data and there's a few reasons and eric and i will name off a few of them but one is dude neither eric or i or any company wants their data lag we want it in real time why do you need a data analyst helping you look at the data that's not in real time. I'm not saying you won't have people in your data and analytics team helping out, but you don't need someone reporting to you on the data.

3:47Neil Patel:AI should have it right then and there, notify you when something's wrong. And then when anyone on your team, whether they're in data or in marketing or any other role, they should just be able to ask AI a question related to your data and it pulls up the answer for them.

4:01Eric Siu:Yep. Number two, I would say is the entry level content writer, the entry level SEO writer. I think that is something that's going to you need someone with good judgment and good experience overall. And so someone that can just, you know, generate basic copy like eHow. Remember demand? What was it called? Was it called eHow? What was eventually changed into?

4:25Neil Patel:I don't know, but I know you're talking about it was like that big company in L.A. Back in the day, demand media. Demand media, there you go.

4:33Eric Siu:So they used to generate, people used to generate a bunch of articles that were like 500 words or something like that. They would hire people in India, for example, right? And just crank out a bunch of these for the game SEO. Nowadays, you already have generative AI tools like ChatGP, Cloud that can generate basic content for you. So I think entry-level content creators or writers, I think what's going to happen is you're going to have, you want a single human editor instead, those are going to be hired more. Those are going to be more demand than just the entry-level content writers.

5:02Neil Patel:Yes. The last one I would say is marketers, especially on the entry level, who are jack-of-all-trades, will not be as needed. Specialists will do way better in today's role than jack-of-all-trades because when you specialize, you have really amazing domain expertise, or you should, and you should be better at things like strategy or the nuances where sometimes AI will miss or get wrong. We look at an organization as a pyramid flipped upside down. You have more people on the higher levels and mid-levels and you have less people at the bottom. And a lot of the people are in the mid and higher levels.

5:43Neil Patel:Sure, some of them will be jack of all trades, but the majority of them will be like extreme specialists who are A players at one specific thing.

5:51Eric Siu:Dude, so we should talk about the levels here because you're talking about that. So there are really four levels when it comes to AI marketing. 90 % of teams are stuck at level one. Here's how you jump to level three. So the four levels here, number one, level one, unacceptable, okay?

6:05Neil Patel:90 % are unacceptable, sorry. Number two - But describe each level. So what is unacceptable when you're saying unacceptable? Well, let me go through the first and then I'll come back. Perfect. So level one is unacceptable.

6:16Eric Siu:Level two is capable. Level three is adoptive. And level four is transformative, okay? I'm gonna go through each one, okay? And then we're gonna talk about how you can jump to level three, which is adoptive. So then this is stolen from Zapier's AI fluency table. Okay, I just want to give him credit. So unacceptable is obviously unacceptable. Okay, so unacceptable means that in 2026, if you say, hey, how are you using AI? Oh, I love using ChatGPT. Even these executives we talked to from the Holdcoast, oh, I'm not really into AI right now. If you want AI, I'm not your guy. Actually, this guy was a president for one of these companies, right?

6:47Eric Siu:He's like, yeah, I'm not really into it. How are you using it right now? Oh, I'm just using it to search for things. I'm using it for recipes. It's been tremendous for that. Okay, unacceptable. Level two, you have people that are capable. Capable now means that you are using like Claude Co-Work, for example, you're using to maybe schedule things for yourself, you're using it to analyze your spreadsheets quickly, okay? You're using for actual real work, not just at, you know, base level, okay? So that's, by the way, this chart changes. We're talking like every month or so this changes, okay? So you have unacceptable, you have capable.

7:18Eric Siu:Then you have adoptive, which is where we want most people to go, okay? Okay. Adoptive is where you are using something. Okay. And let's say you're using it to, you're inputting something into AI and you're having to do the keyword research for you, or you're having to create like 200 ads for you at the same time. Okay. You're basically have a human review, the AI work, the human asks for something, the AI does the work, the human reviews. That is adoptive. I'll give you an example. Someone, someone on our team's done this. They created something end to end. It plugs in with our agent and they'll say, hey, with this agent, hey, can you spin up 200 ads based on the data over here?

7:56Eric Siu:Literally, it can do that. That's weeks or days of work that's condensed into just this workflow. That's where we want most people to go. The only difference with transformative and adoptive is that workflow still requires a human in the loop there to constantly check it more often. Transformative means you have a closed loop where the system is improving on itself recursively. And you want a lot of these workflows in your company. transformative you're probably only gonna have like maybe one percent two percent three percent max maybe five percent of people that are transformative right now as i record this that's why you want most people to get to adoptive because neil today's capable is tomorrow's

8:32Neil Patel:unacceptable that's true this is very true dude it's it's moving at such a fast pace one of the biggest questions i tend to get asked by marketers right now is how do we keep up with all of the change. And you know what I tell people? You can't keep up with all of the change. And don't worry about being the first adopter to every single new platform. Wait to see what settles in a few months after launch and focus on those things that are gaining a lot of traction and look like they're going to stick around.

9:03Eric Siu:Yeah. I would say, one, if you try to keep up, you can't. I feel like I'm behind every single day. I'm sure, you know, the people you talk to, the friends I talked to that think they're out that we think everyone's on top of it. Nobody's on top of it. Like you just figure out what you can based on the time that you're willing to, to allocate to it. Um, and I would say, look, if you feel like you're very AI pilled right now, the only concern I have for people that are very, very AI pilled is the risk of burnout. Um, you're definitely going to feel momentum, but the burnout can be really real now.

9:30Eric Siu:Cause you don't stop working. Cause there's people on my team that work till 2am. I'm like, nobody told you to work till 2am. Right. He's like, no, I'm like, then why are you doing it? He's like, because I'm having so much fun right now. Right. The other person's like 1am and he's like, I found out the chat TPT new images model. I'm like, that's fantastic. But please, I need you guys to, you almost need to have a mental health policy. Like I used to kind of like, we used to be like, okay, yeah, mental health policy right now. It's like, no, no, no. You really do need a mental health policy now.

9:56Neil Patel:Real quick, if acquiring customers has been a struggle for you

9:59Eric Siu:and you are trying to figure out how AEO works, answer engine optimization, or the new version of SEO, how paid works, how all this AI stuff is going to play into your customer acquisition strategy, then check out my ad agency, Single Grain. It's singlegrain.com. And if it looks like it's a good fit, we'll help you with a free marketing plan. So again, go to singlegrain.com and we'll see you on the other side.

10:18Neil Patel:You have a lot of people that you know that are entrepreneurs and I am in a similar camp as you. It's funny when I talk to my entrepreneur friends, one of the biggest problems we see in many departments, especially marketing, because you're adapting so much with AI is actually mental health. And we're seeing become a bigger and bigger problem because people are overwhelmed, they're working too many hours and they're not taking care of the health. And people have this misconception that you just get more done by working more. If you don't get enough sleep, you'll actually be less productive. And sleeping, like I sleep eight hours a day, sleeping a solid eight hours a day, and again, this isn't health advice, but what I found for me and most people, you actually are more productive and you get more done in the hours you are working compared if you have a lack of sleep because you have less brain fog.

11:08Eric Siu:I think this period right now, maybe the next five, 10 years or so, most people will have, most companies will have some type of mental health policy where probably your tokens get shut off or something like that. But I just think, I think so. You just turn the intelligence off.

11:21Neil Patel:That's so good. Yeah. I never thought about that. Like your tokens are shut off. You're working too much. No, right?

11:27Eric Siu:But like, I think people are going to realize that, hey, like maybe this thing is intended to help us make us work less overall in the long run, right? I think right now everyone's just so excited by this new technology that I think we're going to over-index, over-work right now. I think longer term it might decrease, right? But I also think the motivating people are always going to be the same. Like, that's never changed in history.

11:46Neil Patel:Yeah, dude. It's ridiculous. But, yeah. And I feel bad for people, too, because it's like the people that don't work hard, I have to push to work harder. The people who are A players, they work too hard and I actually have to push them to, like, work less, spend time with family. and I'm like, dude, this is just a recipe for disaster for your personal relationship with your kids or your spouse, just any of it. And I'm like, you need to work less. And you just can never push those people. Like you have to push them, but you can never convince them to work less. Like sometimes they'll take like a vacation.

12:22Neil Patel:I had someone tell me like, oh yeah, I'm gonna now have to travel to Mexico more because of family, but don't worry. They have better internet than Cuba. I'm like, dude, I don't worry about you. You work seven days a week and you need to work less. Internet, if it's shitty, is actually a good thing. So that way you actually work less for once and you spend more time with your family.

12:42Eric Siu:Dude, speaking of vacations, I'm going on vacation tomorrow. So that's our recruiting day. But it's because there's like, part of the reason is like, I've never been to Mexico City. So that's where I'm going. But it's just like...

12:53Neil Patel:Oh, this is the trip we were talking about during lunch.

12:56Eric Siu:The work is nonstop.

12:57Neil Patel:And are you going to stop using AI during this trip? No. That's not a vacation. Do you actually go on vacation? I'm going on a Disney cruise. For the first time.

13:06Eric Siu:But usually you worked on your vacations. Well, I still get to work on the Disney cruise. They have internet. Well, we work because it's fun. That's part of my vacation. But you know what? I don't have to get on all these calls. I don't have to be in all these meetings. That's the fun part. I'll do the same stuff. So you'll still be working. You're obviously going to spend time with your family and all that. But let's just be honest. We enjoy working. And this is my time to learn more, to experiment more. That's all I'm doing. Yeah. yeah okay we get to pick one more here so let's pick

13:39Eric Siu:okay I'm gonna go with this one this is a this is this one's a bit of a like a what do you want to say headline e clickbaity one let's go with this one okay here ready Neil yeah go for it so we replace four people on our content team with five AI tools output went up 3x here's the stack so I'm like number one over here. Okay. So we using Higgs field, Higgs field has been great in terms of cranking out way more ad creative or way more videos, way more UGC content. It just wasn't previously possible before. And by the way, when we say replaced, we mean augmented RT.

14:19Neil Patel:Yeah. Number two, and we're not saying we actually need 3x output. This is that headline that AI came out for us, but we've done tests with Answer the Public, that is my own tool, usually can cut down time spent on content creation by a third to a fourth. So you're saving quite a bit of time because not only does it help you find the topics, it helps you write, you still should keep a human in the loop, it'll help you post, it'll help you come up with the images, it pretty much does a lot of it soup to nuts, and it can do it in a lot of different languages. But still, even with that, you still need a human in the loop.

14:57Neil Patel:And I really advise, no matter how many tools Eric and I give you, we are not advising you to take out humans from the loop for content creation. You may not need as many humans, but if you replace humans fully, your results aren't going to be as good.