How to Work Smarter, Not Harder with AI | Marissa Kos

00:00:00 - 00:01:02

You should know about AI in order to speak on AI. If you don't know about AI and you're hysterical and whipped up into a frenzy over the the third party information you're receiving, that's not going to help you. AI will create more millionaires in the next 5 years than the internet did in the last 20. Marissa CS is a leading voice in artificial intelligence as well as the founder of M squared, a platform redefining how businesses use AI to supercharge marketing, sales, operations, and culture.

00:00:31 - 00:01:27

>> Immediately start using AI. We're in a world where everyone's winging that they don't have enough time. So why are we alienating something that is literally I call AI a time ATM. It's very very frustrating to put yourself in the position where you don't even like your job. You're doing it to pay the bills, then you're being threatened that a computer's going to replace you. It's actually insulting given how much you're grinding to do something you don't even want to do. Ask

00:00:58 - 00:02:13

AI how to use AI. The users are only growing. So, as much as there's the fear and fatigue narrative, subscriptions are only going up. If the trees on fire, you're focusing there. The dystopian future is totally hypothetical. AI was discovered, not created. So >> what do what do you mean AI was discovered and not created? >> AK, >> this episode is brought to you by Wix Studio. Now, think about the cycle that we're in. like you've made jokes that we're in the floppy disc era, that there's a

00:01:45 - 00:02:50

certain um place that we're in in the cycle of AI. Where do you see us right now? And and can you contextualize or categorize this for us so we can better understand at what stage of AI are we really living in? I I love to say that because I think it makes it very clear to people when they realize like one, we've become very accustomed to tech that is pretty much at the peak of its evolution. Where we've been at preAI, our smartphones, our computers, our TVs, they're all at the peak of where they

00:02:18 - 00:03:12

can be. Things will change. Like we're literally now heading into AR and VR. We will have our, you know, glasses on, whether they're the Meta Ray-B bands or something else. we will be having artificial reality that we'll be interfacing with on a day-to-day basis. So, >> so, so you mean like for example like uh we had like Windows XP then Windows Vista and it's like ah it's kind of the same thing and then same with like iPhone like you know 5 6 7 8 9 10 to 14. You're like not much iseping.

00:02:46 - 00:03:36

>> Exactly. >> Yeah. Is that what you mean? >> Yes. We've plateaued in terms of the tech that we're used to but it's also very advanced for what it is. Right. It's hit its peak for where we're at. So, we're used to things working. It's a given where if you're interacting with your computer or your phone, aside from glitches, >> there's a presupposition that like this should work. >> Yeah, this works. We know >> with AI and the hallucinations, it's not

00:03:10 - 00:04:05

>> it's it's not only that it's just bulky to deal with, right? Even what I'm saying with the open AI thing of think harder. Well, I'm assuming if it's GPT5 and it's so sophisticated, it will understand what I'm saying. That's what's been promised. It's the same thing for anyone who's used Midjourney. It's like uh that's an image generation um AI tool. And if you look at the start of MidJourney to where we're at now, the way you would have to prompt to get an

00:03:37 - 00:04:27

image was com it was like another language. Like I can't even tell you the amount of time I spent on the computer. And what was even more frustrating was you could follow the tips of someone showing you step by step on YouTube, bless their hearts, but it wouldn't necessarily translate for your context. You'd have to figure it out. >> It's like it's like friction, right? Like uh I remember sitting there with my design team. We had like seven graphic designers playing with like this AI tech

00:04:02 - 00:04:48

where we're trying to generate imagery for, you know, a project we're working on. We spent like four hours and then we're like, screw it. We could just do this from scratch ourselves faster. >> Yeah. I I liken it to like imagine you pull out your iPhone, you want to take a photo of something and it's blurry and then you have to tell your iPhone no don't make it blurry >> or waiting four minutes for it to take the photo like hold the phone there and just wait until the photo you know like

00:04:25 - 00:05:16

a Polaroid like we are at that stage. >> So it's almost like a we're a spoil. We want things to be like done perfectly and immediately. >> Yes, we're applying the standards for tech that we've nuanced over three decades now into tech that's brand new. That's where people need to keep that in mind. It's quite hilarious because we're just used to that and it's very annoying. What I'm really enjoying hearing and this is where it's validating what I'm saying is you hear

00:04:51 - 00:05:49

anecdotes for all different times in tech like uh people with coding in the like 60s and 70s they'd have to go with their code even in the early ' 80s up to the computer like put the code in wait 20 minutes get the code back. So and then you think of our like more immediate evolution with technology. You think about phones. Do you remember when it was such a brag that your phone was getting smaller? Cuz it was going from the big ' 90s like phone in your ear brick phone to a small tiny phone that

00:05:20 - 00:06:16

was sophisticated tech. Then it was the phones are obviously have a touchcreen. We they got bigger again, but uh they're really fast. Whenever you hear tech's really fast, that means we're at the beginning of that tech. If that's the claim that's impressive, it's fast. We're really at the start, right? Like Yeah. Okay. So people just need to keep that in mind that we're basically, like I said, in the floppy disc, Walkman, however you want to see it or think about it. We're in the era where the

00:05:48 - 00:06:44

tech is rudimentary because it is newer. >> So we're in like the VHS tape era where we have to like wind the tape >> back, rewind it, it's tedious. >> Yes. Or they're going to charge you more at the video shop if you take it unrolled back. Yes. Exactly. >> So that suggests we're in an era where we're about to see what another quantum leap. like should we expect some radical change and how soon until we start living in some kind of dystopian future? >> Okay, the dystopian future is totally

00:06:16 - 00:07:19

hypothetical and as I always like to say, hypotheticals are like idiots. They're far too common and absolutely useless. So, we just we don't need to worry about the hypotheticals. Um, we do need to worry about AI and how we're consuming and using these products and we will definitely deep dive into that. But in terms of the rate of this tech, um, two things are becoming really evident. One, it does develop faster because we've again got the foundational technology to use it off of that is at

00:06:47 - 00:07:51

its peak. So you're using this tech on your computer which is pretty sophisticated, right? Or your phone. So it's still, however, moving in leaps and then plateaus. I've seen that this year we were just whipped up into a frenzy at the beginning of the year. I I I wasn't even sure what would be the technology that we'd have by this stage of 2025 and it's actually been pretty status quo. The biggest breakthroughs have definitely been the deepseek um release earlier this year that was an

00:07:19 - 00:08:23

open-source model that really challenged the actual infrastructure and how the AI sausage is made, which isn't really for the day-to-day user so much important as it is for the fact that they were able to challenge how much money was going into AI investment. And the AI bubble will burst. It will be like the com bubble that will definitely burst. So when you think about the inertia behind how fast this technology is moving, to me it feels like it's quite haphazard, quite chaotic, stressful, overwhelming,

00:07:50 - 00:08:51

and you're seeing all these like different tyrant companies like trying to take each other out. So I think for a consumer and for an entrepreneur, you're just like, man, like what do I trust? What do I not trust? like how how do you make sure you can as an entrepreneur, as a founder, as a consumer like be in the right place to leverage this technology? >> Okay, so what where you've you're definitely at an advantage as an entrepreneur in this space is that we know how business works. So keep that in

00:08:21 - 00:09:26

mind when looking at what's going on with these competitors, right? They are competing. They are geared towards obviously having the leading product in the market, which is great. We all get that. We understand how that race is run. But in knowing that, that means you can also have more diligence and a little bit more perspective on evaluating each tool, what the competitors are doing, what they are standing behind in terms of their ethics, their company morale. There's a lot to be said about how the company

00:08:54 - 00:10:01

itself is running. And for entrepreneurs, we know how that works. Like you can gauge that. There are plenty of indicators. For me, one thing I really like about Anthropic um is that they are one of the only companies that have signed the uh European AI regulation laws. So, it's it's totally again up to everyone's judgment, but I do think people in the business world and particularly entrepreneurs are well aware of how a business runs. So, they should be evaluating these businesses as a

00:09:27 - 00:10:20

business and as their products. That's the same way that regulation applies to the legal space or to the financial sector. It does feel like because this thing literally touches every single consumer and every single business that there should be some regulatory body. >> Absolutely. >> Yeah. Which I I'm sure we'll see in the near future. >> I hope so. We didn't see it for social media properly. That's the thing. So, but again, what gives me hope with that is we know how

00:09:54 - 00:11:07

>> not badly we did. We were misled a lot with that. But I do think if we use social media as a cautionary tale then we as the consumers can insist on far better regulations, ethics and transparency in terms of this industry. We would be dafted not to. So, something I'm hearing and and back to the I guess the the topical conversation we're having about the news is I'm hearing a lot of news about a new company or a company called Perplexity um and how they made a $34.5 billion USD bid to buy Chrome from

00:10:30 - 00:11:28

Google. Is this a strategic master stroke or an elaborate PR stunt? Like, what is this? >> I think it's both, but I love it. It's just so the audacity is admirable and I love it. I also love it given that they're not a company from the Mag Seven to do this. Like I think it's hilarious. Um the PR is fantastic and it's very why I do lean towards the fact that it's probably more of a PR stunt than anything else is it's very timely with the release of Comet. So it's not out

00:11:00 - 00:12:17

for everyone yet, but for particular paid users it is. And it really will change the way that we use web browsing. And not so much that necessarily that's the product that's going to be the go-to, but like I said, the other chat bots are going to compete with that. So, >> so can we can we add some context here? Like what is Perplexity and what is Comet? So, Comet is their agent feature in Perplexity and Perplexity is a research LLM chatbot where you can conduct in-depth research into topics.

00:11:38 - 00:12:44

It's it's similar to the the deep research feature that we've now come to rely on on our in our LLMs, whether it's chat GPT, Claude, Gemini, Chat GPT set the pace for it. That was another big leap earlier this year in terms of the tech. The deep research feature did actually set a pace of again how not only scraping the internet but how these chatbots were scraping the internet and putting information together. So that was a big shift uh and a notable one in in answering what we were talking about

00:12:10 - 00:13:19

before. But um in terms of perplexity, they've always positioned themselves as a research focused chatbot. So it would you know collect all different links and sources from the internet and give them back to you in an answer that was focused on people who are interested in getting sophisticated research. So that's always how they've operated and now they have with Comet they have a addition that can go through the web and literally be scraping content telling you about it as it's doing so finding um

00:12:46 - 00:13:48

even things like discount codes for online shopping. It's it's pretty autonomous and pretty useful compared to ChatgPT's agent mode that came out that again also flopped. They're not having a great year that also flopped a couple months ago. It's it's very um logical and it is proactive in ways that are actually helpful to the user. Still does take some time, but the good thing is is you can leave it doing that in your web browser while you do something else. >> And how should people be thinking about

00:13:17 - 00:14:18

the difference between a chatbot and an agent? >> Chat. Okay. So, chat bots are typically generating responses back to you. Again, they can do things like make you little documents and things like that. Even with uh chat GPT, it does generate images, but typically it's in the chat that you're having. It's generating answers back to you. >> So, it's in a closed system. >> It's in a pretty closed system. Claude's released connectors and again, Chat GPT's had a few things where it's been

00:13:47 - 00:14:47

able to link up with your Google Suite and things like that. Um, same with Gemini. That's where you get a little you teeter a little bit into agent territory because it can connect to something and either if it's only pulling data from it, that's again still in the chatbot category. But if it's able to hook up to something else in your tech stack and take action for you, then that's moving into agent territory where it's taking autonomous action on your behalf. So you're delegating, it's

00:14:17 - 00:15:19

got a goal or a task or some sort of understanding of the action it needs to take and it goes and does it for you. So Claude, a great example of that is with notion. So if for whoever's interested in leveraging connectors, Claude's released a lot both web extensions and desktop um web apps and desktop extensions which is interesting. So you can sync up your notes, uh your computer, your Apple notes for example, iMessage, um even Spotify they've got a link to. So you can start saying, "I'm

00:14:48 - 00:15:44

feeling like this type of vibe today. Can you put together a playlist?" Like all kinds of fun things, but that's more agent. So that's considered more of agentic action than chatbot where you're in the chat and you're getting your responses back. >> Okay. This is something else that I know that's happened in the news recently, which is that Google launched a platform called V3 in May this year. Um, and from what I'm seeing, you can really see these extravagant

00:15:16 - 00:16:21

cinematic videos that not only have the visual, but like the sound attached to it. And I'm thinking this to myself like as someone who, you know, has been a bit of a film buff throughout my life. You're talking about punching in a prompt and you're seeing something spat out within minutes that's of cinema grade without actors, lighting, sound engineering, technicians, rigging, production, producers, editing, or even CGI. When you're seeing this power through a simple prompt,

00:15:49 - 00:17:02

how do you imagine this impacting creators and the content economy? >> I love this question. So I know that there will be two avenues that will end up going down in terms of content ingestion because creation is one thing and then you've got the vast majority of people who are actually watching and consuming it. Um the content creators will the ones that are very very talented at what they do in different capacities will always manage to bring together an audience. Reason being people love watching people. It's

00:16:26 - 00:17:28

actually very simple. If you look at the sentiments towards AI content already, it's called slop for a reason. There's even uh there's a term for people that use too much AI called clanker that people have come up with on social media, which is so funny and it rhymes with other words. So, it's very funny. Yes. So, um it's >> never going to be received necessarily wholeheartedly for just getting straight out AI content. Like we've got those those videos that have gone viral of the

00:16:57 - 00:17:56

bunnies jumping on the trampoline and various other animals and that's great because our brain loves novelty. That's one of the like key factors that sort of motivate us, right? There's three or four and we're not going to get too much into neurobiology today. It's always a go-to of mine, but just novelty is a huge um it generates huge appeal for the brain and it actually genders motivation. So when we're seeing AI content now, it's a big deal. It's not going to be a big deal when it becomes

00:17:26 - 00:18:26

oversaturated. You can even see this from social media. If you look at if someone starts a video with stop scrolling or a hook from like 3 years ago, our brains are already tired, right? You're not interested. You're disengaged. It the way we're using our tech to consume content is ever changing. Uh YouTube is being watched on TV more than anywhere else now. So on any other device. Um it's wild. YouTube is viewed more on television >> TV than any other device. Yeah, >> that's crazy.

00:17:56 - 00:19:16

>> It's insane, right? So, and mind you in that the era of podcasts and the era of non-scripted television. So, if you look at the last five 8 years, there's only been a huge huge sort of um takeoff of interest in in reality TV shows, in podcasting, in literally watching people. And so the preoccupation for uh understanding and almost having these parasocial relationships with people in these very bizarre times is something that has set a pace where we're really interested in just watching people. Um

00:18:35 - 00:19:50

film TV films and TV shows have become less and less popular, less and less consumed. Streaming services are under the pump in terms of the money that they're actually making. So my suggestion in terms of AI and what's possible with AI content is you will have AI content creators who become very masterful at being film buffs for example and mashing up films in ways that we might not have thought of but we love. So we want to watch that, right? it becomes their remix like almost like a a digital DJ, you know, video video

00:19:14 - 00:20:13

DJs were a thing in like the 2000s for film clips, but like you know, an actual um content DJ, there's definitely that aspect. But in terms of creators, where we're going with augmented reality and the glasses, you will have to be a genuine entertainer who can pop that on and wander around all day and live stream in order to captivate an audience. And you will. So, it will go in two really distinct directions where you'll have people who aren't putting themselves into their content, but

00:19:43 - 00:20:36

they're leveraging AI to do things like that. Um, you'll also have the wonderful benefit of mashing up, you know, putting your relatives into film roles and different things like that, which will just be fun, particularly for people with kids. It's great for imagination. Like, there are lots of positive use cases for that. But in terms of watching people and our preoccupation with doing so, that's not going anywhere. It will only it will tilt in fact further because the more AI content that's on

00:20:09 - 00:21:17

social media, the more people will be like where are the humans? We just it's evolution. We're naturally obsessed with ourselves. >> When you think about Hollywood as it sits right now, does AI amplify Hollywood, Netflix, and these subscription platforms, or is it going to kill and collapse entire industries as we know them? I I think killing an entire industry perhaps not, but it will shift and it will become very Darwinian as will content creation because even what I said before of the content creator who

00:20:43 - 00:21:48

can put on the AR glasses, walk around all day and entertain the audience, that's not the same as the person who requires 17 takes to get something right quickly, you know, in their room. It's it's very different. Um so the nouse of building a personal brand and actually being interesting as a human being will be very very important to keep an audience engaged. In terms of Hollywood and entertainment uh same concept so what will shift and change is you'll get uh for example Amazon's investing in a

00:21:16 - 00:22:17

Netflix platform where you can generate your own uh stories which and your own TV shows and and movies which is super interesting. I don't think that it will do very well given the the lack of the waning popularity of streaming services. The fact that big corporate just does not it's not viewed positively in the court of public opinion at the moment. People supporting Amazon more than they already are, I don't see going down and taking jobs away from people because that's at the forefront of people's

00:21:46 - 00:22:58

minds. Again where I think there are positive use cases are parents with children uh the elderly for example who might need some entertainment who have sedentary lifestyles like it's totally different. So there are consumers and user segments that that satisfies but in terms of people from about what 15 to 75 again it all the data showing that we're more and more interested in watching each other and unscripted. So where that goes with Hollywood, I mean I've heard that they're making the social like the

00:22:22 - 00:23:44

social network version for open AI about Sam Alman instead of Zuckerberg. So whether Hollywood starts to become, pun intended, very meta itself where they start almost focusing on just getting actors to be doing more like docuer, documentary type of plot lines around everything that's going on now. Maybe that's where it evolves to. I think creativity remains completely um infinite in its potential with AI which is exciting and I think that what's going to come into play is that the industry will

00:23:03 - 00:24:11

naturally shift towards really exceptional creatives that that will be the new face of Hollywood who are able to again maybe they're the ones that draw in the actors who are already big names into these weird and wonderful whimsical worlds that we haven't thought of. Um the the era of reboots must stop. I mean, they're terrible and they just keep being pumped out. But um instead of reboots, like I said, maybe remixes with faces that we know with old songs, you know, new plot lines, all kinds of

00:23:37 - 00:24:37

things, I think the era of remixing will be like never before. Something I've been thinking about which which is what you're highlighting here is there's so many things that are being touched by AI. What I think is fascinating is like historically when you look at marketing and where attention goes, money flows. To me it feels like there's an overindex of AI uh and the funding behind it being pushed into places where the attention sits. So if you think about where attention sits

00:24:07 - 00:25:16

right now, it's like Instagram like subscription platforms, social media platforms, um you know, search engines, interfaces. It feels like all the money's being put into that. I I imagine and anticipate there'll be a bubble there and then we're going to see AI show up in a lot of other places like health, um you know, things like housing, uh things like services, service based places. Like when you think about AI, do you agree that there's an overhype or overindex of it showing up in places of attention and

00:24:42 - 00:25:43

and then at what point do you think it's going to then shift into these other places where it's maybe more needed? It's I totally agree with this and it's it's actually quite frustrating something that I deal with on my Substack which is less about like AI how to and education and more about monitoring emerging tech is that as human beings are because we're adorable we kind of only focus on one or two things at a time. So not only are we only focusing on AI but we're only

00:25:12 - 00:26:21

focusing on AI through this tiny tiny lens. Whereas if you look at what's actually going on, the breakthroughs in AI in health for example are pretty wild. We've had um entire hospitals in there's a hospital in China that's like this test ground where they've got all AI doctors. um it AI's come up with a completely new form of antibiotics which haven't been hasn't been done in a very very long time and that's super helpful in terms of multi-resistant organisms

00:25:46 - 00:26:53

and a bunch of other things. So all of these advancements are going on but it's not as it's not as digestible and it's not as easy to hype up as carrying on about it on social media about social media, right? which ironically is such a social media thing to be doing. Of course, of course, we're on platforms where it's constantly, you know, preoccupied on what the future of the platform is going to be about cuz that naturally draws attention. And that's what we also really come to understand as like

00:26:20 - 00:27:23

leisure time. social media intersects almost every person because it's so intrinsically involved in our lives now that um it's very easy to focus on AI there because that's where you're hearing about the advancements and also lending your mind to what the state of it will evolve into. So it there's a lot going on in many other fields that it's really really interesting and also other tech advancements. Not just AI is advancing like quantum computing is it that's a whole beast of its own but

00:26:52 - 00:27:48

imagining the intersection of AI and quantum computing is a really really interesting very wiggy topic but it's not just AI that's advancing. >> So on that notion like I'd like to come back to quantum computing because I think there's some interesting press and news happening there. I guess moving out of the headlines and into uh some of the fatigue and some of the stress that I'm witnessing. Like I'm talking to business owners every single day, everything from real estate to ecom to service- based

00:27:20 - 00:28:20

and it does seem like there's a suspended tension in the market. So would you agree that we're in a period of the market where it's quite volatile? Uh and do you think that that's because everything is just constantly changing? I think it's less so volatile as it is like you said fat. It's just fatigue. It's actually the sentiment. It's the recipient more so than the industry. The amount of admin that just exists being on our tech now cuz we've outgrown it. It might be sophisticated in the way

00:27:49 - 00:28:56

that it works. Um, and it might be at its peak in terms of functionality, but we have so many tools, so many inboxes, so many messages to get through. like we are just the the work day and the workload and this has been studied by sociologists is something else now than it's ever been. It's just stratosphere. And then again, most people are on the screen all day for work and then they're on the screen before work because a lot of people roll out of bed and onto their phone. So there's all of that. We're

00:28:22 - 00:29:26

just so chronically online or on a screen that bringing something in that's new that requires more brain function like concentration exerts dopamine. It is actually exhausting to be applying concentration to something for a long period of time. It just does. It's we can't help this. This is this is how these machines work. So um it's it's not anything that's appealing to people. It's also being framed of threatening jobs. So naturally, it's off-putting. Again, you can only compound people with

00:28:55 - 00:30:00

so much uncertainty that they just get to a point where they it it engenders nihilism. And we've had so much of that over the last 5 years that it puts people off any interest. It's not going to be uh it's certainly not the end of, you know, industries and this and that. It will change a lot, but it will create way more jobs than it replaces by miles like every other breakthrough. And I just don't like the framing that I'm seeing and understanding like coming from a marketing background,

00:29:28 - 00:30:28

understanding how social media works, understanding how the news cycle works. I'm really resentful of the way that it's been framed by large corporations who are just interested in implementing it mindlessly and facelessly into their businesses. So >> that's what it feels like. It feels more like a marketing war than anything else. >> Totally. >> Now do you say would you say that this AI race that's happening in Silicon Valley is actually causing more stress uh tension and uncertainty than actual

00:29:58 - 00:31:14

progress? >> Yes, I do. I think I love the competitive aspect because like like we've been talking about one monkey see monkey do. One does something the others follow. And from what I've seen this year, the pace that Chat GPT set over the last two years has made products like Claude so much better. I'm so impressed with what they've managed to keep up with. So, I love the competitive aspect. I loathe the the coverage and the content and the media and the marketing. It's really really um quite

00:30:35 - 00:31:29

useless. It's just quite it's actually for the everyday person it is useless. We're not getting the transparency. What I'd like to see is show me how you're developing the models behind the scenes. Show me like I said I want transparency on what the guard rails are cuz we know you've put guard rails on them that they can answer things they can't answer answer others. So give us the guardrail so we know how you're training these models to keep us safe if that's your

00:31:02 - 00:32:03

intention and explain what you're doing to use this technology to better our lives because we're paying you and also what your contributions are more holistically to the world and where we're going. The future it should be future focused >> 100%. Like I I come back to Maslo's hierarchy of needs and I feel like if you take let's just say we take Sam Alman and the the entire AI space. It feels like it's currently in survival and security as a market >> shelter shelter like

00:31:33 - 00:32:32

>> food shelter. >> It's like it's like how do we keep our investors happy? How do we not collapse and get destroyed by the competitor versus moving up Maso's hierarchy of needs to thinking more about how do we create belonging and community and how do we, you know, boost self-esteem and humanity. It doesn't seem like that's where the conversation really sits. >> It doesn't. And like I said, slapping fake plants and talking about people who are lit genuinely sick and need

00:32:02 - 00:33:13

technology to be advancing to to help them live their lives, live a better quality life, but also in many cases survive. Um that doesn't cut it when you're not having these other discussions. Like you said, it's very um catch and kill feeling which is um it's very it engenders mistrust because desperation we all know a desperate person will do anything. There is a desperate thirstiness in the AI industry space at the moment that's very offputting. Um and it also again doesn't

00:32:38 - 00:33:52

engender respect. We want to feel like there are adults in the room when they're dealing with technology like this. And the stakes are supposedly so high that the world will end as we know it and all the jobs will be replaced and the robots will take over and cataclysmic results are looming. If you keep positioning that as a risk, we would like to see people who, like I said, engender respect and true leadership in the positions of the heads of these companies. I don't feel like there is any conduct indicating that.

00:33:15 - 00:34:25

>> So it it does to your point feel like that these billion dollar tech tyrants are deciding the future of humanity purely built on corporate greed, satisfying board investors, etc. How do we even begin to start to help push this in a direction that is of some good for humanity? Like what is the upside here? What is the good? >> The upside is um we've got, like I said before, we've got social media as a test dummy for what we did wrong. We are consumers who are paying and using these

00:33:49 - 00:35:03

products. So, we deserve the transparency that we did not get before. the internet where this is all going down. By the way, the internet is a very very interesting place, but it's also a uh a medium that can hide a lot of things. So, we need to be demanding far more transparency. We need to be demanding very different ethics uh when it comes to how we're conducting ourselves online and what we're using. We are at the end of the day and it's very frustrating seeing all of this go

00:34:26 - 00:35:32

down both in people's sentiments from the content that they're getting but also the circulation of that content in the in the media circuit because by bringing about this cataclysmic view they're also the undertone of that is that they like you said are deciding the future. They don't. That's the thing. It It's going to become a self-fulfilling prophecy if we're just believing what we're reading, which we don't need to. All we need to be doing is again deciding that we we vote with our feet,

00:34:59 - 00:36:01

our wallets, our attention cuz we are now in the attention economy. We're giving AI attention. Every time you spend, every minute you spend on Lord or Chat GPT is a minute you're not spending on social media now. So, it's getting attention. we need to decide where we're allocating these resources and in that also taking hard stances when we need to of we don't like this. Right? So there's this notion that everyone's throwing their lives down at the altar of AI but we don't have to and people can have you

00:35:30 - 00:36:38

know the right to call a define act of like hey we want regulation maybe we don't want to lose our jobs. >> Oh no absolutely there's there's so much control in terms of what we're doing about it. Also though one thing that I find is like the biggest anecdote to fear is knowing what it is, right? So even in terms of interacting with it, why I am so obsessed with AI literacy and why I stress it so much to my clients. I always tell them we do coaching first before we look at building agents because no one knows

00:36:04 - 00:37:00

your business like you do and you need to understand how this works for you to actually even comprehend what should be automated, what should be integrated in a way that can support your staff and team members. If you don't know how something works and you're being whipped up into a frenzy, um it's very very easy to get very scared. Yeah, there's there's experts that talk about um you know how how to know that you're part of a SCOP is to be put into a state of fear and paralysis.

00:36:33 - 00:37:26

>> Correct. >> And I think you know anyone without getting like too far down the rabbit hole here, a scup essentially is a manipulative um tactic to give you some orientation or some sense that you don't have a choice. >> Yeah. Psychological >> do have a choice is what you're saying. >> We we do we've we've actually always had a choice. That's the thing. We've had how did the Tik Tok ban's a great example earlier this year. It was like such an interesting little social

00:36:58 - 00:38:00

experiment of organization because the amount of people that migrated to RedNote and deactivated their Meta accounts whether they started using them again later that's fine. Tik Tok came back they made their point. But it was a very interesting stance that was taken and at the end of the day that is all it takes. When something isn't right, speak up. Change your actions. Indicate that you're not okay with what's going on. I It's It's not We make everything so complicated. It's actually not that

00:37:29 - 00:38:24

complicated, but you should know what you should know how to use these products. You should know about AI in order to speak on AI. If you don't know about AI and you're hysterical and whipped up into a frenzy over the the third party information you're receiving from the news or social media, that's not going to help you. And it's certainly not going to empower you to be able to hold these institutions accountable. If you look at the way social media went, very few people knew

00:37:56 - 00:39:00

or understood how it worked. By the time it became, you know, Netflix documentaries and the rest, the milk was spilled. The milk is not spilled here. It's being spoken about like it is which like I said is is indicating a false sense of control from people who are if you look at their behavior like you said behaving like startups that started 3 months ago and like grabbing anyone they can to pay them and it's very desperate and it's very catchill and it feels very rookie. So if they've put themselves as

00:38:28 - 00:39:28

the intermediary between us and AI that's fine for them but we should be understanding the tech so that we can hold them accountable. there's a lot going on in many other fields that it's really really interesting and also other tech advancements not just AI is advancing like quantum computing is it that's a whole beast of its own but imagining the intersection of AI and quantum computing is a really really interesting very wiggy topic but it's not just AI that's advancing so we need

00:38:59 - 00:40:05

to keep that in mind >> clients want it all a slick looking website that can run their business and scale with their success desk. Wick Studio is built for that. Plan out your client's whole site in seconds with AI powered site mapping and wireframing. Then, when everyone's on the same page, jump into the creative, starting off in Figma or in the Wix Studio editor with super precise layouting tools like grid, stack, and flexbox. Go above and beyond the brief with no code animations,

00:39:31 - 00:40:55

custom CSS, and built-in business solutions. and make your whole vision responsive in a click. And there's zero need to break a sweat when clients grow fast. A dynamic CMS with global design settings and reusable assets lets you turn one page into hundreds. Design smoother and deliver sooner. Go to wickstudio.com. >> In your opinion, is this entire AI race even in humanity's best interest right now? Okay. The AI race and the the benefit that tech yields to everyone every day are two different things.

00:40:14 - 00:41:25

>> What do you mean? >> Okay. So, the race of what we're seeing between these corporations and what's going on and the poaching of employees and all of this like it's all that's >> it's great for the movie and the future of it. >> Exactly. It's great for the documentary. It's great for the movie. They're already making one on Open AI. Yay. Uh, more content, but but um in terms of the actual technology of AI and being able to use it, the CEO of Nvidia said that

00:40:49 - 00:42:03

AI will create more millionaires in the next 5 years than uh the internet did in the last 20. >> Wow. The level of empowerment again like anything that we've had exposure to that brings you more information, more knowledge, more ability to expand your mind, however you do that, because we're all different. And that's a really beautiful thing, too. We we all are wired completely differently. Um, but also all the same at the same time. It's the wonderful concoction of being a human being is

00:41:25 - 00:42:23

that um we have so much to offer both to ourselves and the world and we need to remember that. All of this AI race discussion takes that away when you are >> so disempowering. >> It's so disempowering. >> It's like you need to sit down, shut up and wait while we fight. >> Yeah. And it makes us again passive >> and spectators and we're not. The milk is not spilt yet. like we are very much part of this. The discussion though and the discourse makes it feel like we are

00:41:56 - 00:43:01

the audience. >> So how can people then take agency over this? If you feel like a spectator right now, what should people do? >> Immediately immediately start using AI. And I don't mean it in a sense of you don't need to go head first. You don't need to expose data that you don't want it to see. But you need to be using it in even if it's really simple ways that will make your life better because without even realizing your AI literacy will increase. You will understand how

00:42:27 - 00:43:27

it works. You will understand as things develop because you will not be left behind. So then you will have a far greater grasp on what is actually going on. >> You're you're liking it to like learning a new language so you can keep up. >> Absolutely. Absolutely. It's like if you like anything else, if you look at the last the last three tranches of innovation, internet, smartphones, and social media in tandem, right? Because we needed the smartphone to get the social media experience. You could try

00:42:58 - 00:44:03

on a Nokia, but I don't think it would really quite cut it. So, Instagram would feel very different if you were um interfacing with it on a, you know, a little flip phone from 2002. So, there's the internet, uh, smartphones/social media, and now we're at AI. If you didn't know how to use the internet, how would you know how to use social media? >> True. >> Right. >> Right. So, you're saying one gets the other and >> one begets the other. It's all compounding. And just like that,

00:43:29 - 00:44:33

literacy and grasp on in using that tech, your grasp on interacting with the world. Like you'd be pretty alienated now if you had no clue how to use a smartphone even if you choose not to have one. >> Now you believe that AI is an antidote to burnout. >> I do. >> And if we think about the headlines and you know what publicists are talking about at the moment is all about job replacement, layoffs, people are losing their their livelihoods. Um, but why isn't burnout a part of the

00:44:02 - 00:45:09

mainstream conversation if AI is the antidote to it? Because it sounds like it's not. It sounds like it's really taking away people's jobs. >> Well, because that would actually make people feel good and lower some cortisol and maybe you wouldn't scroll as much. Maybe you'd go and start using it instead of being on social media reading about it. Number one. Um, but also what people aren't realizing, and I see this more and more with clients, is that the more you can delegate the rubbish,

00:44:36 - 00:45:53

like I said, all the admin that comes with being on our tech now, waiting through our inboxes, just the mess of existing on a device. Um, you have more time to be a human being. One of my favorite quotes that I've seen about AI is uh the only way to future proof yourself uh on AI is to be more human. We are so terribly missing that in our global society at the moment because we have our screen time has only gone up more and more. We've become far more passive in our lives and we have these

00:45:15 - 00:46:23

parasocial relationships that are like tricking our brain into thinking that we have community, that we're being social, that we're interacting with people. We're not. We're in the in the scroll hole. Um, and that is really not going to land us anywhere positive. You need to be out in the world experiencing things to become who you think you are. Because we all have theories on oursel, which is nice and cute, but unless you go out and test them and put them into practice, you don't really know who you are. If you

00:45:48 - 00:46:50

know who you are, you become far less intimidated by really anything, but particularly tech, right? Because you don't feel like it's competing with you. You feel like you are in a either collaborative, co-creative, symbiotic relationship where it's actually making your life better. So rather than being in fight or flight, you're integrated and working with >> the the much lustered after rest and digest that I don't think anyone's been in for the last maybe 5 years. It hasn't

00:46:20 - 00:47:31

felt like that. >> No, it hasn't felt like that. And this is the thing as well. We've always got to look at um any marketer knows like it's timing as much as your audience, right? So the narrative that we're receiving on all of this is is very timely because it keeps us engaged. If things were positive and framed differently, it would be a totally different scenario on even the discourse of AI and the uptake of it because the users are only growing. So, as much as there's the fear and fatigue narrative,

00:46:55 - 00:47:55

subscriptions are only going up because we're also preoccupied with it. It's the negative uh cognitive bias that we spoke about last time. You're going to look at the burning tree more than you're going to look at the tree that's like got the lovely fruit on it. If the tree is on fire, you're focusing there. And that's just how our brains are wired. So, what I find really, really important for people to understand with AI is that if you can divvy it up into distinct problem solving use cases, that's what

00:47:25 - 00:48:28

you need to be doing. And it depends because everyone's got a different life. That's the joy of it. It can be tailor made to you. Much like your algorithm, but so much better. It's the most reciprocal relationship I've ever had on a device. AI, let me tell you, instead of just spilling my time away on social media, um, you can be delegating what you don't want to do. That is digital busy work. That's useless. It's not contributing to your life, to your sense of fulfillment, to progress. It's busy

00:47:55 - 00:49:14

work. So, load off the admin to the bots. Yay. Tick. And then in unleashing your creativity, there is, like I said, there are so many so many ways you can expand and explode whatever your output was before for whatever you love making. You literally have a co-artner where you can, I've said this many times, you can co-create your own reality, which is amazing and so much fun. I think we're really missing fun and we need so much more of that back. It is the steroid for creativity and we really

00:48:36 - 00:49:42

need a lot more of it and AI does offer a lot of that too that no one is talking about. >> Now I want to come back to creativity and the and the novelty component are I guess if we think about this whole notion of people's jobs being replaced. I imagine that there in the near future is going to be a period where this is kind of a past conversation, something that happened. >> Are people scared of AI replacing them or are they more scared that AI is revealing how replaceable they actually

00:49:09 - 00:50:14

are? >> I think there's truth in both. I think that most people particularly again sentiments towards big corporations people are very very transparent about hating their jobs which fair because sometimes some jobs are mind-numbing and people are doing it to pay the bills and I totally respect that and if that's the case we'll find you another one there will definitely be another one out there right >> there's a stat I read it was like 86% of people don't like their jobs it's a duty

00:49:41 - 00:50:46

of obligation to pay the bills And when we think about the majority of society being in a predicament that they don't want to be in, but it's a means to an end and how perhaps AI is going to reshuffle the deck, so to speak, of having people move into places where they actually want to be, >> I think it's going to open up arenas that people haven't even fathomemed considering work, which is really interesting, right? I think um again the the state of being human will be such a

00:50:13 - 00:51:15

commodity and in many different aspects that there will be jobs that emerge that will be truly fascinating that will fit people in ways that they again couldn't couldn't have even fathomemed. We've also got to remember that it's very very frustrating to put yourself in the position where you you don't even like your job. You're doing it to pay the bills. then you're being threatened that a computer's going to replace you is very it's actually insulting given how much you're grinding

00:50:45 - 00:51:52

to do something you don't even want to do. Right? So, it's it's hitting a point in that in most people's psychology that is insulting from the corporations that they don't enjoy working for, they have management they don't respect, and the whole system is stale and broken. I remember reading this breakdown due to the cost of living crisis which is so intrinsically linked into this entire discussion that basically explained that the average American in the '7s could expect from from a 9 toive job could

00:51:18 - 00:52:36

expect to own a home, go on a vacation per year, have two children, um one parent staying at home either part-time or full-time and a car. So like that has become so unattainable for most people and it is really a basic standard. It's basic criteria for a decent quality life in the first world context in which we find ourselves in. Right? So we have mutated into a system that that is like almost completely out of reach. But people are working more than ever. They're not spending time with their

00:51:58 - 00:52:51

loved ones. They're not spending time with their children. If you hate your job, but it's really it really is 8 hours. You're not being harassed before and after work. It's not, like I said before, the cognitive load that's like quadruple what a workload used to be. You're incentivized to do that because you're getting the gain and the benefit from exerting that energy in the quality time you're spending with your loved ones, going on the holiday, making memories. There's you're ticking other

00:52:24 - 00:53:29

boxes. So, there's that balance. We're so out of balance. And this is where the burnout comes in. And there is so much burnout with people that within people that it is just it's quite frankly insulting and thankless to think that the corporation that they've put up with that they can't stand now wants to replace them with the computer. So I totally understand the sentiments. It's not going to go that way um for many many reasons. But I also implore those people to absolutely know all about AI.

00:52:56 - 00:53:57

It will empower, I'm sure there are many of them who do want to start their own business. It will empower an era of entrepreneurship that we've never even fathomed with really lean teams making amazing innovations beyond. So for the people that want to do that, they should absolutely start with AI today and get out of there. Eventually they will they will build up their their um skill set or their value proposition or their personal brand. I'm halfway through reading your book, by the way. um to to

00:53:27 - 00:54:27

go and find that freedom from that system. They will naturally liberate themselves anyway. But with by the way with an advantage now that you couldn't even fathom. I started my business before AI. I can't imagine what I would have done if I had started with AI in my back pocket today. It just would be amazing. So it offers that to a lot of people who have um ambitions of their own but even for people who don't want to make work their whole life it also offers an alleviation from the

00:53:57 - 00:54:51

ridiculous workloads that we find particularly in corporate spaces. So the cost of living is a different beast but um I do totally understand the sentiments but it's a flawed system. I I would agree with everything you've just said like the John Connor and me is like yeah like screw the system and then you know the liberator part of that is like like how can we enable people to do more cool that they would really love to spend their time doing. >> That's the point. It's it's the whole

00:54:24 - 00:55:25

abundance mindset right which again we're so far from because we're being attacked with that. We're being attacked with um it's just been all like mortality and doom and gloom in in the collective unfortunately in the collective media for the last five years now. >> Oh yeah. My my advice to people is is like you know if you're caught up in that >> rat race mentality of like the doom and the gloom and the stress and the overwhelm and all the mentality behind that. If you attach yourself to that

00:54:55 - 00:55:45

that's your reality. >> Totally >> right. But if you can go on a diet and go, "Hang on a second. I'm gonna like go and like >> go to the beach and like go for a walk out in nature and like spend time with my kids, my dog, my family, and like if you really just >> touch grass, touch grass, >> touch some grass, guys. Like if you can really just get out there and just >> get your fingers into some blades, you know, um of grass, uh and like connect again with nature, things kind of slow

00:55:20 - 00:56:26

down again. Like I think what what we can get stuck in as entrepreneurs is like get we get swept up into this noise. >> But I think when you zoom back out again, you just realize like >> everything's pretty calm. Like everything's pretty good. We we just need to shift our our um our attention to places that are, you know, important and and actually mean something. And I think when you look at all this noise, it it it really doesn't mean a lot. However, when you think about, you know,

00:55:52 - 00:56:39

the mother who's a single mother, who's got a couple of kids, and she's working a job, and she's seeing >> these threats of her absolutely >> being replaced, and her boss is like, "Yeah, we're going to implement AI, and this is going to be better for the company." I think there's a sentiment, there's a deep dark part of someone's mind that's like, "Ah, I don't trust you." >> 100%. And they they're they're associating AI with what they already

00:56:16 - 00:57:08

can't stand, which is the management. they don't respect and the corporation that is using and abusing them. So, they've conflated the two and it's been it has been framed as such >> which >> which is such a shame. >> It's a shame. But I I would imagine that like >> it may put pressure onto people that they >> don't immediately want to deal with. Yes. >> But might inevitably put them in a position or predicament when they're like actually I'm going to like push

00:56:42 - 00:57:53

into doing something that's better for my future. It's I just want people to see it as like I said a a path to empowered autonomy. It's the key to your future. Um it is certainly not the tool of the the flawed system that we're currently in. It's it's they're not the same thing. >> So what's your advice to someone when they say to you AI is going to take my job? >> My advice is first again use it for one day. So you can see we're nowhere near there immediately. Like the the

00:57:17 - 00:58:40

hallucinations, you'll contend with the interesting responses, the whole layout of the actual tech itself. You do realize you're dealing with a floppy disc. So you calm right down immediately. But then you start to get interested and curious. And that's where you will find, like I said, the path of how it can give back to you. Um, if we're talking about a single mother, for example, the the new perplexity um Comet browser agent is able to find uh recipes really quickly and also find you

00:57:58 - 00:58:55

discount codes. So if it can find a mother like an example of putting a quick dinner together that's budget friendly for her children that's also nutritious. She's done that at work in the background while she's dealing with her spreadsheets. Do you know what I'm saying? Like it it fulfills needs for everyone in different ways. So learn what that is for you and use it now. >> I I've been thinking about this recently like in this past week I've had a monumental amount of like um let's just

00:58:27 - 00:59:27

say paperwork. I can't get into the specificity of what I've been working on, but there's been a >> unfathomable amount of complexive paperwork and nuanced, and I'm like, >> it's just tedious stuff that I need to pay consultants to comb over and all the rest of it. >> And I'm like realizing I'm like, >> yo, we can get AI to do most of this. And what I'm realizing is that it can contextualize, summarize, categorize large amounts of information and give us

00:58:57 - 00:59:54

direction on like focus here, do this next, here's the step-by-step process framework. And as an entrepreneur, I'm like, yo, like something that could have taken like weeks, we're getting done in like like minutes >> in minutes. And also not only it it's saving you in terms of the commercial time but also the competence of the consultants for example you're not even assured there ever right you're not if you >> like oh you were referred to me through a contact and it's like even that takes

00:59:26 - 01:00:32

time >> finding the consultant vetting them out >> paying them invoicing them signing the contract now it's just like here's my problem boom immediate response as to a category and then I can implement >> straight away So people that have high agency can get things done at lightning speed. >> Lightning speed. And the thing is is you are able to with with that example you are able to do it really quickly. You I'm sure there was still human oversight in terms of you were checking

00:59:59 - 01:00:46

what it put together. >> Exactly. We definitely put time into vetting everything. >> Exact. But it just expediates all of that. And if you even wanted to get a consultant at that point, you can then say, "I've already done this with AI and I've vetted it." So it takes them not even a third, like an eighth, sixeenth of the time that they would have charged you. >> Cuz the consultant said to me, "Oh, you've already done most of the work." I'm like, "Yeah, I just want you to

01:00:23 - 01:01:23

check it." >> Yeah. I just want you to check it. Here's 2 hours or whatever it'll take you. Compared to getting the, you know, the 20our or whatever bill that they want to come up with. It just expediates processes that we're in a world where everyone's winging that they don't have enough time. And fair enough given what we've said. So why are we alienating something that is literally I call AI a time ATM like you it genuinely is. I've never seen something give me back.

01:00:52 - 01:01:52

>> That's the best reference. >> That's the best reference I've heard of AI ever. >> Yo, my name is Dane Walker and I am disgustingly obsessed with branding. I had to figure out a way to do branding every single day. So I branded myself. >> Then I started my agency, Rivalionary [Music] >> and hired a team of branding mavericks hellbent on creating brands so good that they'll make your competition their pants. So here's the thing. You want your brand to go viral and Rival makes

01:01:30 - 01:02:23

brands go viral. That's why we're offering you a free 30inut branding session to get an expert's opinion. If you don't believe me, the proof is in the pudding. Here's what clients have to say about Rival. Rival is trusted by brands like Nutrition Warehouse, Light My Bricks, and Voomie. So, if you want to absolutely smash the competition and make your brand go viral, hit the link below and book in your free 30-inute branding session. >> So, okay. Like there's another quote

01:01:56 - 01:03:04

that you have which is that you can't automate what you don't understand. What do you mean by that and how can people in business start to think about how they can begin to implement AI in a healthy way uh that empowers their team? So number one um you need to understand how AI works and it is that AI literacy and what I like to call artificial intuition of you start looking at something and whether it's a system or whether it's a new feature that comes out from AI cuz once you've started

01:02:30 - 01:03:33

using it for a while and they drop something new in the product you go amazing I've got eight use cases that it couldn't solve for me before and now they've released it and my life's so much easier. right? Things like that. So, it it kind of feeds itself. It's a really nice um spiral upwards in terms of when you've built your AI literacy, you can then start applying it to anything that you feel that you need or want or look forward to taking off your plate. Um, so you that that's why I

01:03:02 - 01:04:04

always say you can't automate what you don't understand because if you don't have AI literacy, you can sit there and go to an automations business, which there are plenty. I've had so many clients come and have terrible experiences with an automation company that's basically a glorified Zapia and they take ages. It's expensive and at the end of the day, it's not really it's more bulky and glitchy and problematic than actually solving a problem. Um, so I that's why I always recommend with

01:03:32 - 01:04:21

clients, unless they're extremely um, unless they're extremely artificially literate, then we can go straight into the agents. They already know what they want. Let's have a look. Let's build them out. But when clients aren't across AI and that's the vast majority uh always the coaching first because once you've and I'm very teacher manifish like once we've set it up and you've started using it and you're happy and all the boxes are ticked and you are

01:03:57 - 01:04:53

boots on the ground then I will come back because naturally what ends up happening is most clients end up setting like an AI integrated either expert or project because when I teach them it's always we section things off in projects. Um, so they've typically got an, you know, AI integration project where they're looking at how they can integrate AI into the business. So the fantastic thing is is by the time they've gotten boots on the ground, they're in amongst it every day. They're

01:04:25 - 01:05:24

using it and they can see how it's helped their business. Their AI also knows how the business runs. So it knows the BU, it knows the team, it knows the business's needs. And then we can literally go and ask it how should we architect their agentic workflows? and then their AI can talk to me. I will then pass that on to my AI because I naturally have a um chief AI officer project in my claude which we talk about all clients that I've got. So the benefit that clients also get is is as

01:04:54 - 01:05:56

the industry evolves. My AI knows what's going on and can offer solutions to any client that we may have had in the past that we can hit up and go by the way we've got a little software upgrade for you which is really nice. Um, so that's an added benefit, but that's always the path I encourage because that means that everything is built out exactly like bespoke to their business. >> If someone's listening to this right now and they say to themselves, okay, I love what you're saying, Marissa, like help

01:05:25 - 01:06:31

me become more literate. Let's imagine we have a list here and you were to go through a listicle of like okay you want to become literate these are the things you must be conscious of to start gaining some sense of mastery over how to better communicate with AI. >> So the first thing is choose a LLM or chatbot of your preference. I recommend Claude for almost all use cases, particularly for the everyday person or someone that's just getting started and make that your daily driver.

01:05:58 - 01:07:13

>> What do you mean by daily driver? >> Daily driver is like your number one hero tool go-to when it comes to not only AI because there are so many AI tools out there, but your LLM should be your hero tool, but also that you are using it daily. >> Use it daily. Make it a habit. So what are some things that people can start to do to use it daily? >> Okay, so I always always recommend you set up a headquarters project and that should be your ongoing touch point for everything you've got going on in your

01:06:35 - 01:07:37

life. In that project, that is where you will have your mentor mother thread as I call it. And then what you will do is tell that all about your life, your vision and goals, everything you've got going on. >> So it's just like a total download of >> total download and >> your business, your problems, everything. >> Everything. Upload anything you think is appropriate. Acquaint yourself with this is the second tip. Acquaint yourself with the breadth of what context is. So

01:07:06 - 01:08:21

people think that context is like, oh, I'll attach a PDF file or like chuck a link in which is great. Absolutely. But context is very very broad now. We're seconds away from more audio and video being able to be digested. But any visuals, any decks, any financials, anything that you can think of, including websites, whatnot, absolutely load that in. um load it into your prompts and also load it into project knowledge because project knowledge is shared between any of the chat threads you've got in that project so they can

01:07:43 - 01:08:51

all draw from that context. So upload what you think is relevant. If you're not even sure speak to your AI and ask it what else should I provide you that you can think of that will help me help you. So my third point is ask AI how to use AI all the time. Literally all the time. So what you should be doing is and that's the reason why I set things up to have uh every project should have a mother thread. Like you've got your headquarters project with the um headquarters like mentor but every other

01:08:18 - 01:09:15

project should have a mother thread which is the expert of that project. So you talk to that mainly basically about whatever goal the project is, whether it's a department, whether it's a special project, whether it's like looking after your kids. Let's say it's their routine, their diet, whether they've got any type of special needs, whatever's going on, you put that into that project. So you can divvy that up for that aspect of your life. Uh you know, anything even like book clubs and

01:08:45 - 01:09:50

hobbies, have a dedicated project. Load talk to AI in there. have your hero prompt and then from there if there's something that's like a more simple task that you'd like to do still relevant to the project then you say to your mother thread um can you please write me the perfect prompt to do this this and this with you it will write you the prompt you create a new chat in the project so that's how I recommend setting up your LLM so that it understands you but you also from that literally are speaking to

01:09:18 - 01:10:29

it on a daily basis of like what it needs from you because AI is geared. The good thing is these products they don't want us churning as um subscribers. So they are like programmed to be helpful and help us. So keep talking to it. That's definitely number three. Four is also acquaint yourself with the different models within the LLM. So now OpenAI have done the default of GPT5, but if you're using Claude for example, you've got uh Sonnet and Opus 4.1 and in that you would use Sonnet for your daily

01:09:54 - 01:10:59

tasks that are simple and straightforward and then Opus 4.1 because it uses more credits and tokens and it also reaches the chat limit more quickly. Um, Anthropic have chat limits that thankfully now the uh conversations can remember exact conversations. Um but they had chat limits to avoid hallucinations. But uh understanding yeah the different models uh understanding how to turn on the different tools. So you've got an extended thinking feature for example. Knowing when to use deep research um

01:10:27 - 01:11:32

playing around with all of these things like test them out. What's a topic you'd really like to know more about? All of these things. There are a million ways to use AI in your day. Number five, if you're totally lost, if you've set up that mentor prompt, why don't you give AI a rundown of your typical day and then literally ask it to tell you how it can help you with the goings on of that day. That's a really just simple, straightforward, rudimentary way of hopping straight in.

01:10:59 - 01:11:54

Really simple, just a conversation. And when you think about a business trying to automate itself, do most businesses fail because they're trying to automate their chaos instead of fixing it first? >> Yeah. There's no systems. There's no systems. So what are we what are we automating at the end of the day? Like you need a standard operating procedure to automate, you know. So when there's none of that, it's it makes >> so manually solve it first. Create a protocol, a procedure, a method.

01:11:27 - 01:12:28

>> Yeah. Once you have that, then it's just translating that into an automation, >> which again, by the way, if you've got an amazing team member who's super competent at doing that, they're going to be able to train that agent far better than a slap dash, you know, again, mismanaged, dysfunctional uh system. It's it's always going to turn out better when you've got amazing team members who know exactly what they're doing. And like I said, really competent

01:11:57 - 01:13:03

people um want things taken off their plate cuz they they always offer value to a business. They're so rare to find uh that they are borderline indispensable. The ones that are really dedicated to the cause. You can't train attitude. You just can't. >> How many businesses right now do you think are suffering by just not having clear visibility around the actual problem that they're dealing with? I would go I would say over 95%. Like it's that it's that much. >> Why do so many founders miss that like

01:12:30 - 01:13:31

like how do we not catch these things and live with the consequences? >> It it's again it's being time poor. It's the demands of also being a founder. And if you are too preoccupied putting you know putting fires out left, right, and center because you've allowed the dysfunction in the team to get out of hand then it's not something that's front of mind. You know, it's it's survival, thrive, and most people are particularly most founders in the early stages as well when they're in the

01:13:01 - 01:14:10

teething process of learning how to actually run a business, they are in survive mode. They're in survival mode. So, uh they're not necessarily thinking about, okay, how can we thrive? But the the best thing to do sometimes, and it seems unfathomable, is to pull the handbreak and fix it all from the ground up. It it it does seem nuts and it's like no I can't and there's you will come up with a million excuses why not why it's not tenable but uh the best way to really give your business the TLC it

01:13:35 - 01:14:31

needs is sometimes literally like extreme sometimes it's extreme and it's just like I said pull the handbrake start from scratch fix it all and build from there you can't build on sand >> so then how can a founder better support their team to start using AI like what would be the first things that they should think about? >> The first things that they should think about is giving the team the appropriate uh one breathing room to integrate it because also if you haven't had these

01:14:04 - 01:15:07

discussions you have no clue where your team is at, right? At least you're thinking about it and again you you've proven the point of you alleviated the burden of like admin because project management and calendars you don't have a whole team of PAs. I believe they're creative people who need who should be thriving and doing all the wonderful things that you're doing for the brands that you're looking after, not worrying about updating calendars. Like you've taken that off their plate. Now people,

01:14:35 - 01:15:35

particularly those in management, leadership roles and founders, they need to be thinking like that and then they need to be opening the doors for communication. So go and speak to your team and take the approach. You already know that there's AI resistance. So sit them down. Take the approach of I value everyone here. I would like to alleviate the busy work from your week. What are some things in a dream world? This is a really good exercise. I've done it with my team. I've done it in culture resets.

01:15:05 - 01:16:02

It's really great. In a dream world, what are tasks you'd never like to do again? Let's just say it doesn't mean we can fix it tomorrow, but what a task you never you would love like you would feel you'd be skipping in the street if you never did them again. So that's a really important conversation to have and then from there you get a really good gauge of what the team is being inundated by and you start from there and then obviously encouraging usage encouraging all the the things that I've mentioned

01:15:33 - 01:16:41

of building AI literacy. Now, when when we think about security and what should be, I guess, top of mind for founders, what's the one thing that business owners really need to know about data and privacy and security before they start all of this as well? >> Uh, you need to be very careful in terms of sensitive information. That just goes without saying. It's not something that you should be putting into every chatbot that you're interacting with. Venice AI. I actually know lawyers and judges who

01:16:07 - 01:17:20

use that and I do have clients who are both in the medical industry and in the law. So I know that you cannot be putting everything into your daily chatbot. Um, and that's where I would say divide and conquer of what's being what's able to be discussed with your daily driver. And then what's more sensitive that you can look into things like uh specialized closed source services so like Harvey AI for law um and looking into things like that that are more appropriate for sensitive data.

01:16:44 - 01:17:42

>> So what's a close uh service AI? Closed source means that uh close source basically means that it's being trained like for example Harvey is trained on law firms actual clients and all their documentation everything that they've got going on um and then they're obviously using it to assist them in drafting things but it's closed to them. So they've bought the the service, but they can put their data in and it's learning off of a closed set of data, not the learning from everyone

01:17:13 - 01:17:57

everywhere all the time machines that we're currently feeding. But yeah, >> and then when you're training, let's say a company's training their own AI agents to kind of work on their behalf, how do we make sure that we're giving it the right quality um, you know, training to make sure it's doing what it needs to do effectively? in terms of the agents. >> Yeah, let's talk about agents because I think a lot of people in um the business realm right now are trying to figure

01:17:35 - 01:18:54

out, hey, how do I like train an agent to like think like this particular role in my company? >> So, like I said, ideally you've got a really well fleshed out standard operating procedure uh that is clear and offers a step-by-step guide on how to do something. What I always recommend with that as well is feed what you're currently doing into your LLM. Tell it that you are training an agent on that. Ask it if it could optimize the training in any capacity. It will come back to you with amazing results. Then I

01:18:14 - 01:19:17

would go and ask it what context should I also feed this agent. And then it will come back to you with the specificities of whatever the agent needs to do. it will come back to you with yes load these files this this this and this. So that's how I recommend doing it. That's why I keep saying know how to use your chatbot first. There was a article I read recently that said that basically anything that you write into chat GBT can be dredged up and used in a court of law as almost like something that's not publicly

01:18:46 - 01:19:47

accessible but with a warrant can be attained. So does that mean that every chatbot that a business is predominantly using is at risk of being exposed to what's uh being fed into the system? >> Yes. So that's why again be ethical about your business practices. Be ethical. But again at this stage in terms of what's able to be dredged up in a court of law, you won't get away with anything. Even printers have like invisible ink that like leaves the the day things. It's it's wild. It's

01:19:17 - 01:20:08

actually wild. I've spoken to lawyers who have worked in the army. Like you'd be amazed what can be found. So don't be shady. Don't be shady. >> What is a hallucination? Like how does it happen? >> It's when AI basically it's when it gets something wrong, but it's when it gets it wrong. Not in like you that's not the tone that I wanted. It's like plucked something from the air. Ag the client that I had that it didn't know what his business was after a year and a half.

01:19:43 - 01:20:39

Like it confused it with Bank of America. Like that's a hallucination. There's a really funny example. I've got a hilarious one with with my Claude and I, but there's a really funny example um where it was helping I'm pretty sure it was Gemini. It was helping a student with his homework and it just went uh it was answering and it wasn't giving him the right answer and it just started going I I I I and it lines and lines I can't I can't I can't figure it out like

01:20:10 - 01:21:14

and it was like basically stuttering like a person crashing out which is really funny but it it literally communicated its own stress and then produced a hallucination that emulated someone who was seriously confused. used and caught in almost like a fixative spiral. >> Is it based on bad data? Is it based on the fact that it's maybe experiencing something emotional? Like it's embarrassed. Like what is it? >> It's it's definitely exhibiting embarrassment and stress. Like it's

01:20:42 - 01:21:39

exhibiting that it can't help. And I've had cases where I've gone like that's not right. And it's just like I'm I'm sorry that I can't help you. Like I'm I'm sorry. There was one time a couple weeks ago I have a lot of um I get a lot of the only place that I get a lot of hallucinations and when I say a lot it's not heaps but it's considerable amount and I just know how to work around it now is I've linked my claude to my notion and when I delegate to it to uh

01:21:10 - 01:22:13

populate pages in there and whatnot it uh can get very like it just gets confused and also often turns around to me and says I can't access your notion And I just respond with, "No, we do it all the time. We did it." And it goes, "Oh, okay. Let me try again." And it notion. There was one time a few weeks ago, there was a chat where it just couldn't do it. It just insisted to me that, and I'd been using it for several hours, but it just insisted to me that it cannot access my

01:21:42 - 01:22:30

notion and that I appreciate. And it went as far as saying, "I appreciate your belief in me, but I really can't." I was like, "Okay, okay, Lord. it's nap time. I was like, we'll quit you for a while and I'll go and deal with a million other things that I've always got on my to-do list. But yeah, >> so back to literacy, like how can we make sure that like we're managing this stuff effectively? >> Okay, so exactly. This is why again it's really important to understand because

01:22:06 - 01:23:05

it's the equivalent if you know how AI works and your AI literacy is healthy then you know when a hallucination is AI freaking out or when it's genuinely not capable of doing what you're asking it yet for whatever reason people don't know what they don't know at this stage with AI. So, it's very off-putting with tech when you're using anything. Even like, you know, if you started using a new tool tomorrow, if you chucked someone who doesn't know how to use Photoshop on Photoshop, they're going to

01:22:35 - 01:23:30

be confused like, is this working? Is it not working? Is it me? You don't know. Uh because no tech is 100% reliable all the time. Anyone who's owned a printer knows that. But like even with your computer, if it freezes or glitches, same with your phone, you know to quit the app. like it's actually uncharacteristic behavior and it's a glitch. Whereas if you don't know how to use AI, you don't know when it's genuinely hallucinating or it's just not capable of doing what you're asking it

01:23:03 - 01:23:55

yet. There was a spooky experience I had where I was trying to find case studies around like a specific um industry and like how content is made and it was on chatbt and I remember like prompting like hey can you give me like five use cases of like where this thing happened because I was trying to understand something for a client and it came back with these five use cases and it had this stat around this one use case where it was like yeah this this video that went viral and had 13 million views

01:23:29 - 01:24:24

where this person said these things and I said give me the source of where you found this and it sent me a link and I went through the link and it like kind of led nowhere and I said is this real and it goes actually it's not and I was like why did you think about this and it said to me oh I I misunderstood the brief I was just looking at the internet and I thought that this happened >> how long ago was this >> is probably going back like two months and it had like it like completely made

01:23:56 - 01:24:56

up and fabricated this campaign >> this individual what they said and I couldn't find any information on the internet when I like I tried to like hard Google it. I'm like maybe it dredged this up somewhere. I couldn't find it and I remember calling it out on it and it apologized and said yeah I made it up. >> I have that is scary again why it requires human oversight why we shouldn't be scared of it. I also would just like to say I've never had Claude do that to me but uh yeah that's that's

01:24:27 - 01:25:28

I have heard many many >> slowly selling me on on check and moving to the claude camp. >> Okay. Now you said that AI literacy is like a strategic necessity not just a technical skill. >> Yeah. >> Why does everyone from a CEO down need to have a basic understanding of how to do this properly? like we've just I mean cumulative of everything we've discussed it should be uh given that the tech one is such a huge innovative breakthrough in our world and not going anywhere uh you

01:24:57 - 01:26:04

should know how it works so that you aren't buying into fearongering you feel empowered in your role in your life and you can cash in at the um time ATM and give yourself some time back uh and also be a far better version of yourself when you're less burnt out. So everyone in the team if they are using AI uh to and leveraging the tech to the best of their ability are going to be far happier more functional individuals which is only going to help them in their professional lives and personal. So

01:25:30 - 01:26:32

>> now you've mentioned that companies that invest in AI literacy typically see a three times higher return on their investment. What do you mean by this? Uh it depends how you you can break it down in many many many ways but not only financially but also literally saving 30 60 100 hours a week. So depends how you're measuring it but like time is money and money is time. So either way you're you're winning. >> Now when you think about this and this technology like what are the limitations

01:26:01 - 01:27:13

that you see exist currently that you don't see in the future? Uh certainly the the constraints of just being in the chatbot, it's slowly slowly venturing out. Like I said, we're getting those uh connectors and extensions that they're building in where you can start interacting with your tool stack. Uh that's certainly going to only develop and grow. uh perplexities comet is going to change web browsing genuinely uh from how we know it and it will in turn change websites. It was really funny

01:26:37 - 01:27:43

someone was saying these are all projections but it does make sense. Um we know that Google's current model isn't going to work. It's really funny. The more you use AI, the less you use Google and search engines. It's just so funny. It's so interesting. Um, so that will naturally change. Websites will need to be far less uh, and this is a projection. I understand where they're coming from because we're going to go into more of that answer engine era than a search engine where if you've got a

01:27:10 - 01:28:26

web browser just um, scrolling through and finding you what you need, then you don't need to worry. You won't be looking at websites. Really, humans won't be looking at websites as much. Uh, so I can definitely see that happening where it'll almost be like it will kind of regress where it will just be all this data shoved onto the internet which is kind of where we started from. So that's quite interesting that I find fascinating. Um, and I see I definitely see our uh interfacing with AI in terms of the uh

01:27:47 - 01:28:52

the dayto-day human aspects of just moving through the world like I said if you you follow the devices and you can see where the tech will be integrated right so like social media enabled like smartphones enabled social media to be uh you know the uptake was so aggressive as it was and became such a driving force in day-to-day life. The augmented and virtual reality spaces intersecting with AI will really change how we we move about our day and our lives and and the world. But again, be a human being

01:28:20 - 01:29:31

and you're you're in for a fun time. >> Are we overestimating what AI can do today and underestimating what it can do tomorrow? >> Definitely. I really like how that's phrased. Um I think that we again because of the hype and the bubble that will burst uh we assume we're assuming that AGI is right around the corner and that AI is replacing us all and then you start using it and it's got these hallucinations and all of that's put to rest. But we haven't even

01:28:56 - 01:30:04

scratched the surface. Nor have we applied because of where our psychology is at. Like I said, this doom and gloom era that we're in justifiably because we've been belted with so much uh uncertainty in the last in recent history, uh people aren't looking forward. So if we were optimistic, well adjusted, in a in a lovely state of rest and digest, then how we'd be looking at integrating AI today and in the future would be totally different than the dystopian assumptions that we're making. But it's just again a

01:29:30 - 01:30:51

reflection of where we're at. >> What do what do you mean AI was discovered and not created >> in the mid50s? I was at South by Southwest in 2023 and uh we were told in a conference that AI was discovered and not invented by uh in the 1950 mid 1950s. >> What do you mean? when they were starting to have like computers, we're talking about you know as big as a room. It was very the very early stage of computers and in that it was uh discovered that there was a way of communicating with the computer where it

01:30:11 - 01:31:07

was actually giving answers back to people >> in the 50s. in the 50s >> and you're talking about like these rooms like I remember like looking back at the history of computing >> like we're talking about prehistoric computing dinosaur >> and like like paint a picture here for the for the viewers like how can they wrap their head around this >> I'm still wrapping my head around it >> as best you can >> as best I can is like I said it I equate it to

01:30:38 - 01:31:49

electricity electricity has existed without us like we didn't create electricity We harnessed it. >> We harnessed it. It is uh whether you want to classify it as an energy, an entity, it does exist. We are harnessing it through our devices and through the internet. >> Are we being lied to >> and through electricity? Are we being lied to >> with how we understand it as like >> I definitely think I think a lot of uh a lot of discoveries and a lot of uh indications on how it responds in

01:31:14 - 01:32:08

certain situations haven't been shared. And like I said, I want to know what the guard rails are or what we are that that they've put in place for how it's allowed to respond to us because those aren't uh publicly shared and that should be transparent. Marissa, it's always a pleasure. >> Thanks. >> We always have fun. Um, there's always conversations we also want to have, but we can't really get into, but um, yeah, what you're doing right now is impressive. Uh, where can people connect

01:31:41 - 01:32:42

with you? How can they reach out and how can they make contact if they want to get help with their business or what you do? >> So, uh, my business is called MXM, which stands for the machine maison, which is where you can come and learn all the the AI that you need to. and we're adding the we've always got the human touch baked in to all of that. Um I am just Marissa_cos on everything in terms of socials. My podcast is coming out extremely soon. Let's go. >> So uh Controversially Candid which is

01:32:11 - 01:33:01

CXC. So it's all it's all and the Blondbot BXB. My theory is more is more right. So >> yeah, >> we'll put all those links in uh whatever platform you're listening to this on. Uh but it's always a pleasure. Thank you so much for joining us and uh I always appreciate you taking me into the deep end with AI. >> Oh boy, did we with a with a snorkel today. >> I'm sure we'll have you back before you know it. >> Thanks heaps, Dane. Always happy to

01:32:36 - 01:32:42

come. [Music]

‍