252. [Out] 7 ways AI has changed the world of STARTUPS

0:00AI has changed the game forever for tech startups. So many things are different to what was considered just truth even a year ago, two years ago. And while business fundamentals are the same, you still want to make money, keep your costs low and keep growing. The way you do that today is so different. And it's something I struggled with. I don't know how we keep growing, how we maintain an advantage in this world. So I've been thinking hard about it, I've been researching, I've been using all my podcast guests in order to get you seven different ways to pay attention to so that you can win in the AI world.

0:32These are seven ways you need to rethink building a startup today. Number one, it's faster to build today than it's ever been before. The speed of build is up. Now this is great as a founder, right? This is starting a positive, is that now you can have an idea, prototype using AI tools, and get something somewhat workable very quickly. This is great because before you might have to take months, you might have to buy a dev team, whereas now you can really get your idea down, do it yourself for as long as possible, and also ship products quicker. So we look at it as two different phases. One is that initial phase, if you're an aspiring entrepreneur trying to test out an idea, but also if you already run a startup, before when you think about your product roadmap and you're thinking, okay, we're going to release this feature in three months time or six months time, now there can be three days or six days.

1:21So that whole thing has been compressed. So when you're building now you've got to adjust the timelines you can't compare yourselves to startups of the past where canva or figma where they took months or even years to build out their products now you've got to move faster in order to stay ahead and in order to win if you haven't tried vibe coding yet try it you can use level you can use bolt.new there's so many different tools out there just get in the game work out what you can do yourself without being technical or needing other technical people. Second point is that the time to duplication is down significantly too.

1:56Now, while the last point is showing about how great it is, it can build things really quickly, the problem is so can everybody else. What takes you days to build all takes everybody else days to build, which is scary, right? Because before you could build something, but it takes months, then if somebody else has your idea, it'll take their months too. And we always used to say to people that ideas are cheap and execution is everything. Now that execution is becoming quicker and easier, maybe you do have to protect your ideas more. Maybe you can't tell everybody your idea in the same way in terms of the building aspect, because another thing can build it before you can, if you're not on top of things.

2:30And there's two elements of this, right? When it's about time to duplication, what do I actually mean? It means that if somebody sees your idea, how quick can it take for them to build it? And you've got two different risks here. One is a new upstart, right? So you've got the small side. So somebody else can take your idea and steal it and try and build it themselves. And if they can now compete you on brand and other elements, then your idea is gone, done, dusted. The second element is the big companies. We saw recently where Google and Gemini announced massive updates, where ChatGPT announced massive updates, where Claude has announced massive updates.

3:04Now, what other people are building, so many startups just got wiped out whenever one of these larger companies releases new features so before you could think okay google is not going to bother with my product or anthropic is going to bother with my product now they can build things so easily and so quickly though you're at greater risk and when you've invested so much time and effort and you're trying to build a brand you may be giving up your job to do it this is a really significant risk and that time to duplication for anybody to do and the ease of doing that is a really big risk you need to take into account and if you're going to spend your time building something and giving up your job to do it or taking on financial risks to do it you've got to really consider the realities of how quick is this to duplicate and do I have something which can protect me in that case what a lot of people forget about AI and building in this world is that if you are doing something which is really data intensive then the Googles of this world have got way bigger data centers and can spend way more compute power than you can.

4:06If you look at data the larger companies can farm data so much quicker to train the models better than you can. So if you're trying to compete in that public sphere it's very difficult and you need to work out how you're going to do that. That brings me to point three. Brand loyalty is more important than it's ever been and as somebody owns a community right Now communities are a massive fad. Everybody's trying to build a community but the reality is it's not easy. So when you look at different products, if the actual technological side becomes more of a commodity, unless you're building something really deep and really in-depth, if the actual build is so commodified and everybody can just have the same idea as you and launch it quickly, then it's why are they going to use you over the competitor when the products are quite similar and when you're both using AI to make the design and all those different things.

4:52So how do you build brand loyalty? Is that how can you use your mission, the way you think about things? Do people like you? And that becomes really difficult, right? Because look at LinkedIn today. When you go into LinkedIn, how many of those posts are clearly AI generated? But also, the sad thing is how many of those posts are AI generated, but also working and they're getting loads of engagement. So when people talk about, okay, yeah, brand is now really important because execution is easy. The challenge is brand is also becoming easier because everybody can use these AI tools to also create huge amounts of content.

5:25And we've got to work out how is this going to adjust over time? Because with so much noise now, right? How do you make your brand stand out in all of that noise? What we're investing in and why we're doing things the way we're doing them is those in-person side of things. But again, that's becoming more competitive because everybody is realizing if the online space is so saturated then okay you need to go offline but when there's so many people now trying to host events who are only doing it to sell their product it becomes really obvious and becomes less authentic so this is a huge challenge in the age of AI is how do you build a brand that people care about and love when everybody else has access the same tools that you have access to as well the fourth point is that you can do so much more now with smaller teams so before that hiring process was such a huge part of a startup journey and in a way it was almost like a badge of honor the more people you have then the more successful you are right so proxy you don't necessarily know how much revenue somebody's making but okay wow they've got 100 employees they're doing really well the metrics are changing now you have vcs who are openly saying that they look at the metric of revenue per employee to determine how efficient somebody's being and you have some of the companies such as Lovable, Cursor, Windsurf etc who have been able to do this and grow really rapidly their revenue while keeping very small teams.

6:46Now when you look at that it creates a huge amount of pressure on those people internally but also when you think about it too is that all those admin tasks that founders used to have to do before they could hire is so much quicker to do that now. There's so many different tasks where before you might think about okay I need to hire somebody for this or I need to work loads of hours to do this now you can actually just use AI to do it much quicker for you and that enables you as a founder to focus much more on the actual underlying bits that are going to make the difference the actual putting the foot to the ground getting in the room you need to get into so what you need to do now when you're building a startup is really look at what are the jobs to do so jobs to do and then for every job to do can this be automated can it be done by AI or do you need a human being to do it and if you need a human being to do it it's looking at which skills can we merge here right can we have somebody who's overseeing this and using AI to do it quicker or do you need a person who's really specialized in that area and obviously for some roles say for highly scientific roles you're going to need somebody very experienced who really knows what they're talking about to have to assess properly what AI is generating but for other tasks you might not need the same level of depth and what's incredible is just how AI is permeating across so many different levels of a startup.

8:03You can use it to speed up the coding, the marketing and all other different areas and use it for meeting notes etc etc right. So huge change now but when you're talking about how do I build a startup it's just about who should I hire and more about what tools do I need and for the existing team you have and the people you do decide you need how can you get the most out of them by giving them AI tools and AI capabilities they need to make the most out of their time and their effort. And that brings me to my fifth point about data-driven decision making. Before for some elements of building a startup you were just guessing and hoping for the best right and I still do that tonight sometimes but what you can now do with AI is the models can be trained on so much external data that it can tell you which way to word your sentences, how to do a good hook, how to build in certain ways.

8:56You can spar with it, right? You can ask, okay, based on how other companies and the data you have in your model, what decision should I make? Okay, these are my financials. Should I hire this or should I hire that or what's the best way to spend my money? You can now process so much data so quickly with AI and especially different tools, which are proprietary tools and have access to that proprietary data use it to make decisions so much quicker and so much more informed than you want able to do i think personally there's always going to be a space for that element of taste of when everybody's using the same data then you've got to have somebody who takes the risk into something that's unknown and obviously chat gpt or these different tools they don't know about the unknown right they don't know how fashion is going to change in the future they don't know how consumer behaviour or the way people think or what's going to get popular or unexpected events.

9:52So there's always going to be that element of those visionaries where they can change and see something that nobody else has seen yet. AI can be such a useful tool to just spar your ideas across, right? So you can say pretend you're Leonardo da Vinci or pretend you're Elon Musk and have a chat with me about my idea and see what it says. And that's unheard of, right? That access to information, access to insight you can get today is incredible you just have to remember that everybody else has access too so you want to make use of it as much as you can because your competitors might be doing the same thing.

10:26Point six the giants are awakening unlike almost any time that I can remember the large corporates and governments are open to new software because before you think about it they're very risk averse they have to think about what's the knock-on effects software was improving and updating but it wasn't necessarily going 10x in a short amount of time whereas now that's happening right now if you're using software that's a year old it's so out of date and other people are doing things much quicker and it's become existential right if you're a large company and all your competitors are using ai effectively and you're not their margins are going up they're doing better service to their clients than you are and you've got to adapt very quickly and you can see it right now where larger companies maybe they're taking on less junior employees maybe they're taking on less staff and they're seeing how they can do things with AI that's a very odd situation to be in because and it's a sad situation too for so many people that they're not getting this opportunity once was and as a founder you've got to work out if they're now looking and they're now open to ideas that maybe they weren't before how do you get your idea in there and how do you make sure that they're paying attention to what you're trying to build because it's going to have a huge impact on their business and almost make them very aware of you talking to their competitors their competitors are interested maybe their competitors are going to use it before they do and create that FOMO and looking at governments it's no secret that so many governments are in huge amounts of debt and they're now looking at how to make things more effective even for example with hmrc and different areas like that there's these huge processes in place which is so built so long ago and now they're trying to be updated it's a huge task but if somebody maybe you can build an ai or a tool or a startup that's being able to make a huge difference at that giant scale the contract sizes are massive and this is the time now to seize the opportunity and see while these people and these large companies are getting even the governments are getting that permission from the top to go and spend money to go and test new ideas to go and make things happen can you capitalize and finally product market fit is no longer reliable for so long in the startup world product market fit was a holy grail product market fit is where you built something and the market wants it and it's going to grow and investors are going to back you vcs going to put money into you and that was the aim right and the challenge was is that so much of that relied on what the economy is called setter as paribus everything else staying more or less the same because then you know the market is going to be where it's going to be it's not going to change dramatically it's not going to move in different ways and your product can stay fairly similar you're going to have updates you can have small changes but broadly you know that that product is going to meet the demands of people they're going to be loyal to you and they're going to continue what do you think has happened since open ai and chat gbt so many startups that a couple of years ago were flying high were doing so well have just been wiped out because the way AI and the way it's interfaced with how people do things and how the whole system works has completely changed.

13:56If you look at for example companies which are very heavily reliant on SEO who are now losing out because people are using chatGBT instead. People are changing their preferences. The whole strategies about how you go about everything is fundamentally changing at so many different levels and problems that once existed just don't exist anymore in certain regards and what was entire startups and now just features. AI has made some problems obsolete and this makes building a startup fundamentally more risky than ever before. The reward is still there, you know that right, because you see these AI companies were growing massively and getting huge raises but you might not be seeing because it doesn't make the headlines how many companies are folding how many companies are now just zombies of where they're not really competing anymore they're not really getting anywhere in this world that's changing so rapidly we don't know what's going to happen in ai regulation we don't know what's going to happen in so many different areas when you're trying to build a company and there's all this huge amount of uncertainty around you it's a bigger challenge right and we've got to just be open and admit that.

15:06So even when you look at, for example, the platform risk, so many companies are wrappers, right? They're built on ChattyTBT or Claude or whatever it is and they can do really well, they can grow really quickly. But at any point if those companies start increasing the calls of APIs or how much it costs for a token, then all of a sudden the entire business model is gone and it's what we call platform risk right is that if for example in the past you might build a company that's dependent on facebook where it's an add-on for facebook so many companies now are add-ons in a way for these large language models and you saw it recently with winter which won't get into too much detail about what happens when somebody shuts off from that so when you look at product market fit and whether or not there's a good business model all of these different elements now in this faster-paced world it can just disappear so I know that's a scary final point right and the truth is I don't have all the answers and the point of this video is to make you think about these different challenges and how you're going to adapt to them and what you're going to do and how your strategy is going to change because you can choose to ignore what's happening and the way that things are changing or you can choose to make a plan that might not work but that's the life of entrepreneurship right you make a plan things change you're wrong and then you go again it's an exciting time right if you're a founder and you're doing this and you don't like the uncertainty you're in the wrong industry but there's different positive elements and everything's speeding up and everything's moving faster and that means you can win so quickly like you can see people like lovable and win stuff are done but it also means that everybody else is moving fast too and you've got to keep running and keep sprinting to keep up