How AI is transforming startups — for better or worse

0:02Hello and welcome to the Sifted podcast, the show where we help you get to know the brightest and boldest people and companies in Europe's startup ecosystem. I'm Amy, Sifted's editor and your host. And with me for this episode is Miriam Partington, a senior reporter at Sifted and the author of our rather fantastic Startup Life newsletter, in which she gets into the nuts and bolts of company building. Recently, she's been speaking to a lot of founders and investors about how AI is changing the startup workforce, from the new jobs it's creating to the ones it's replacing, or at least significantly changing.

0:35The impact it's having on the software companies use, the way business is higher and even the hours that companies work. No business is approaching the AI era in the same way and everyone we speak to seems to want to know what everyone else is up to. So we thought today Miriam and I would share notes and hopefully give our dear listeners some useful ideas and tips at the same time. Miriam, welcome to the pod. Thanks so much for having me. So this week in your Startup Live newsletter, you were writing about AI slop, all the crappy AI generated content. We're now subject to people posting all over our LinkedIn feeds.

1:12Why did you choose to write about that? I just think that at the moment, we're seeing a lot of AI generated content sweep over platforms. And the problem is, is that it's just very, very obvious that it's created by AI. And I think a lot of founders and investors think they're being really original, really giving their hot takes on LinkedIn, which they think is going to be helpful. And it's just not really, because in the end, it just all kind of sounds the same. And so I started to think of founder branding and how important that's becoming. You know, we're actually seeing that founders who used to really want to get press for their announcements.

1:53So, you know, they'd want an article in TechCrunch or unsifted. we're seeing a lot of them actually bypass traditional media and actually post themselves on LinkedIn. And I think there's just a huge emphasis on being present on LinkedIn, being present as a founder, as a brand and representing your company. And I think that AI generated content actually really damages that brand to put it in a more exaggerated way, I suppose. You know, there's tons of research to show that people kind of tune out or actually distrust content when it's written by AI. So that was the thinking behind it. Yeah, and I've been speaking to quite a lot of investors about this question.

2:33Everyone's asking of who are going to be the AI companies that will sort of win out and what really are the differentiating factors between lots of these different AI companies. And I guess the classic example at the moment seems to be Lovable, which there are a bunch of other websites out there which you can use to build apps and websites yourself but lovable and specifically its founder anton has just done such a good job of his personal brand which is very sort of intimately connected with the company's brand and that's just him posting all over linkedin all over x product hunt all these different platforms just talking talking talking about you know how well the company is doing how its arr is growing and growing all these interesting use cases and when maybe there isn't that much differentiation in the product, if you become the one that everyone knows about, if you are the household name, then that is a really important part of protecting your business.

3:33The founders you speak to, do they seem generally more worried or excited about the changes AI is bringing in all manner of ways? I guess it depends on who you speak to, because I think there's kind of two different sets of companies right now. You've got the established companies that have been around for maybe a couple of years, maybe a couple of decades. And they're having to think about how to rewire their organizations and their products based on AI. And then you've got this kind of new era of what we'd call AI native companies that in the last year or two have been born kind of with AI embedded into their operations and their product.

4:15So I had lunch with a founder the other day who has an AI powered shopping platform. But she said that in the founder community or the people that she speaks to, the WhatsApp groups that she's in, the two primary concerns among founders at the moment are fundraising and AI adoption. So it's pretty high up there. What are you seeing founders actually do when it comes to AI adoption? Are people really sort of jumping on the opportunity Or do you think behind closed doors, everyone's actually having a panic attack about whether they're doing enough and, you know, fearing that they're slipping behind?

4:49I think it's a mixture of the two. I think some of the founders are trying to be quite pragmatic and are just trying things out. And so some of them will probably be doing research into tools and seeing how they can embed them into their organisations, being like, OK, what are the problems in our organisation? Or what are the things that take the most time, maybe even on an administrative level? and then trying to figure out what kind of tools would then suit that purpose. And from what I'm hearing from AI native companies, experts that are helping companies transform, it's actually the better way around to do it, to say, what are the problems across the entire organization?

5:27And when might AI actually be able to help us, rather than being like, okay, there's this new tool that everybody's talking about, let's just bolt it on and just hope that that works, because that leads to a lot of confusion in an organization and ultimately just wastes money in the end yeah i had a few conversations with the people leads at various vcs who are obviously seeing what's going on with hiring across a whole bunch of portfolio companies and they were saying that when it comes to certain key c-suite roles having that sort of vision of how ai will be used across company not specifically saying i'm going to come in and we're going to use lovable for x and cursor for y and but actually sort of having a much more holistic and bigger picture view of how the organization could be rewired it's like a really important part of nailing those those key hires and getting those key roles yeah and actually from what i'm hearing from founders of you know larger unicorn companies they're actually letting go of c-suite hires that don't want to transform with AI, which I found really interesting.

6:37You know, people that have been in their organization for years, some of them have been maybe a bit more resistant to change and have said, well, I'm happy to use AI in this regard, but I don't think we need to completely change everything that we've been doing. And that's led to people actually being made redundant. Because I think there's an understanding among a lot of founders that if you're not willing to jump on the train, then, you know, you're going to be left behind. And any leader that's kind of resistant to that is you know going to get the chop yeah and you've been writing haven't you about tiny teams as this phenomenon and across sifted we've been writing about this a bit and we're seeing aren't we companies of literally two people raise millions of pounds what's your view on that do you think startup teams will just across the board be smaller my personal view is you know you have to grow at some point there's no way you can deliver you know hundreds of millions of pounds of annual revenue with a really really really tiny team you know I think in general teams will be leaned out and we're already seeing that founders are hiring fewer but very kind of higher quality people and I think this means also with fewer people to manage you then have fewer management roles and we're already seeing how like the middle management layer is being thinned out so I think generally teams will be leaner in terms of these tiny teams I think you know lovable of course is the famous example that got to I think it was 17 million annual recurring revenue in three months with 15 people I don't know how long that can actually remain that way it also depends on what kind of customers you're serving if you're like a lovable and you're like a self-serve product, so people can literally just log in and do it themselves, they don't need onboarding help or whatever, then you can afford to keep a leaner team.

8:32But for companies that are serving enterprise customers, there's often much more kind of team members that you need to kind of handle implementation and compliance and all these other topics. And we're seeing how companies like the Lovables of the world are moving into enterprise customers. So then the question is, how lean can you keep a team if you're serving these really huge companies that need a lot of help. Need a lot of help, but they are the ones that ultimately a lot of these businesses want because they are much more likely to be on annual subscriptions or multi-year subscriptions. They're going to probably churn a lot less quickly.

9:07A few weeks ago, we had Emma Burrows, who's the co-founder and CTO of an AI startup called Portia on the Sifted Podcast. And she also said that she expects engineering teams specifically to drastically change over the kind of coming months and years and to get much smaller and specialised. I guess we've been talking about teams getting smaller more generally. What are you hearing from founders and investors when you speak about engineering roles in particular? I think particularly for entry level engineers, we're seeing that there's fewer engineering jobs and there's also data to show that. I think there's definitely a lot more that you can do now with AI.

9:47and of course engineering teams are already leveraging that. I think from what I'm hearing from founders things are changing slower than people actually think they are and what I'm hearing is that with AI tools now being able to do things like write code what that then opens up is more roles needed for example for quality assurance. So I'm wondering actually how this is going to play out long term because as you introduce more AI into your organisation and into your kind of engineering teams specifically, the more team you're going to need to actually monitor that AI and to oversee things like compliance and data security and quality assurance.

10:29So I think it's a kind of wait and see at the moment. But I wrote a piece a couple of months ago, I think, on engineering talent and what kind of CTOs of startups are thinking about when it comes to hiring junior engineers. As we mentioned before, junior level roles are down in Europe. There's not so many companies that are hiring massively for junior talent. But in terms of how kind of juniors should upskill, a lot of CTOs are saying, you still need to master the kind of fundamentals of the engineering job. So instead of being able to kind of go into an organization and just use AI to code and do everything, I think there's still an understanding that you need to, you know, be able to master programming languages.

11:14You need to have an understanding, even if you are using AI, what errors there are, and things like that. So I think in terms of where engineering teams are going, yes, there may be fewer people, but the roles will evolve, and they'll have to be a lot more kind of focus on data security compliance and things like that. Yeah, and that's something you've also been looking at, right? Like there are new roles being created as a result of AI? Yeah, absolutely. We did a piece a couple of weeks ago on the go-to-market engineer. And they essentially kind of build systems to help sales, growth, and customer success teams to work more efficiently.

11:56And this was a role that was kind of popularized by a company called Clay in the US. And it's sweeping over to Europe now. So there's not that many companies in Europe that actually have GTM engineers. But the few that do have them are saying that this is going to become a really really core role in the future a quick one from me the sifted summit is back this october join a whole load of europe's smartest founders investors and operators along with its very best tech journalists for two days of expertly moderated and thought-provoking conversations and extremely valuable networking This year, we've got hundreds of stellar speakers, including Sequoia's Luciana Alexandru, Gusto's Timo Bolt, Seedcamp's Reshma Sojoni, Granola's Chris Pedregal, and Visionary's Club's Judith Dada.

12:45And we're expecting 3 ,000 attendees. We're also launching a brand new CXO stage, which will be laser-focused on what CEOs, chief people officers, chief marketing officers, and chief financial officers want to hear about. I'm personally looking forward to taking to the stage with Balgerton's Saranga Chandratulaka, perhaps one of the nicest and smartest VCs in the land, and listening to my colleague Daphne LePron-Soronguet interview OpenAI's Laura Modiano about the next hot AI startups and trends. Whether you're newish to the startup world and want to get some practical advice and insights, or an old hat wanting to make some new, great connections, the Sifted Summit is where you need to be.

13:25So grab your ticket now at summit.sifted.eu and because you're a Sifted podcast listener, use code pod 20 for 20 % off one other hiring trend I've heard recently I've only heard this from one person and podcast listeners if you've heard of this please get in touch because Miriam and I have been trying to find other examples of this but one of these these the head of peoples who I spoke to recently said that they're seeing founders take management of much more people so I think for a few years there was a trend towards you know founders shouldn't manage more than a handful of people because you know it takes up too much time to be doing all these one-on-ones and this person I spoke to said that she's seeing founders manage maybe 12 people 12 direct reports because everything is changing so fast that founders want to have a really direct line of communication with sort of all their key people who know what's going on so they're making really well-advised decisions which is quite interesting when these sort of perceived wisdoms just completely switch from you know don't have more than five direct reports to have as many of your most important people as possible.

14:38I guess it's all about maybe once again this goes back perhaps to keeping organizations tight and maybe as a founder having that direct line to all of the kind of major people that are owning certain things within the company kind of means less coordination effort in the end and I think that's also part of the mindset behind how these smaller teams are being created is actually when you have fewer people it's less kind of overhead in terms of actually having to coordinate everyone so maybe that's also kind of behind that yeah I find it interesting and and I think this also ties into something else I wanted to talk about which is the the rise of kind of or the return of hustle culture of it feels like it feels almost like we've rewound the clock and we've gone to a kind of earlier stage of startups where there's kind of smaller teams everyone's like working really fast and really hard and we can debate whether we think doing a 996 work schedule is a good idea for a company long term or not but then we're also seeing this rise of like hackathons again and it feels sort of more I you know, more in-person, more roll your sleeves up, more sort of everyone's experimenting together.

15:53And, you know, it's a brave new world out there. Do you see that as well, Miriam? I think it's kind of always been the case that startups want to have people that can get in there and roll up their sleeves. But now this is becoming even more of a thing. And I think like what you hear about now when you look at kind of job advertisements and when you speak to founders about who they're hiring, it's all about being a self-starter, being super proactive being able to kind of spot problems on your own and solve them and also use ai to do that so i think founders are really looking for those kinds of people that can do a lot more with less and i think that's also reflected in the kind of metrics that investors are looking at like i wrote a couple of weeks ago about how a lot of investors are focusing on revenue per employee so how much revenue each particular person on a team can bring in And I think that's a really interesting trend because everything is just about efficiency nowadays.

16:51Yeah. And as you were saying earlier, it's interesting. We've gone from this world of like having loads of employees and having a really high valuation. And both of those things going up quite quickly. We've seen a sign of success. Now it's kind of a lot more focus on revenue. And we can debate whether, you know, the ARR metrics we see bounded around on LinkedIn are kind of true annual recurring revenue or, you know, monthly figures that get inflated, etc, etc. But there is more of a focus on actually bringing in money. And as you said, doing that with as few as people as possible, which is quite an interesting switch.

17:30Yeah, definitely. I mean, I spoke to a team in London called Science Machine and they have a team of five. and they don't really want to grow all that much in the next couple of years so they say but they were saying that the differentiator now is that there's going to be less specialized knowledge and it's going to be much more about bias to action at least in startups and so because the AI tools can help you a lot with a specialized knowledge you as the employee are going to have to be the one that's able to kind of pick things up a lot more quickly so I think we're going from an era of specialism into generalism.

18:07And the way that you're more efficient as a company is if you have people that can work across a variety of things and like a variety of departments and everybody feeds in to different teams in the kind of service of creating revenue. And again, just looking at some of the job descriptions of companies hiring at the moment, they're asking for things like jack of all trades or polymath or these kinds of things. So I think that's also a really interesting trend that now this kind of generalism is coming back to the startup world. So it's quite interesting. On the one hand, you would think with the rise of all of these AI tools, democratizing tech more than ever, and anyone, you know, who has access to the internet can sort of get on board with this.

18:52And yet, I think you will, I sense your view might be the same as mine, this world of like hustle culture work 996 is so exclusive to in particular a kind of young person who doesn't have caring responsibilities you know it just excludes these whole swathes of the tech sector and you know you've got people like podcaster harry stebbings sort of glorifying this 996 all over social media but then you have others like boulderton vc partner saranga Chandratulaka was kind of maybe the most public person to come out against this and say this hustle culture obsession is really silly you know you of course you can really push at a business for periods of time if there's a really important deadline or client you need to win or you know whatever it is to push out but like that is absolutely not sustainable and if you want to run these businesses for decades, then you cannot be working six days a week, such crazy hours.

19:57Yeah, I totally agree. And I personally found the 996 discussion quite frustrating, to be honest, because, as you said, a lot of this narrative is really being peddled out by young men kind of in their 20s, who, as you said, don't have caring responsibilities. And also, I just think that people have different mental and physical thresholds. And some people can't work those kinds of hours because they may have a disability. Some people, including myself, need time to tune out from work to be able to maintain a mental equilibrium at work. And I think a lot of founders get around this by just saying, well, here's how this is going to be.

20:39We're going to be working these hours. And if you don't want to join, then don't join. And in a way, I guess that's fair. but as you say I think now that AI is just kind of so ubiquitous and all companies are kind of going in this AI trend are we creating too much of an exclusive environment where people are really going to be shut out? Yeah and then when it comes to whether companies are rewiring for AI fast enough I've heard some very differing views I've heard some investors at pretty well-established funds in Europe who've got you know been investing over multiple funds now who are more or less saying yep we know that those sass unicorns of yesteryear who we were so excited about in 2021 are not now going to bring the returns that we once thought they were and actually we're super super excited about the new generation of ai companies we're investing in and then i've heard other people say no that's really silly those sass unicorns of you know if your previous generation still have you know excellent relationships with their customers you know they're not full of you know complete incompetent people they will be adapting sure they'll be adapting more slowly but they have a brand they have a reputation they have customers who are already quite sort of you know trapped in a sense in their ecosystem and so they do have a considerable advantage over some of these newer generation of air companies because you're not going to churn as quickly from then which is quite an interesting i think two sides of coin Yeah, I think that is really, really interesting because I spoke to someone the other day who works with companies on their operations and helps them, you know, get to a level of being more efficient.

22:27And she was saying that actually a real challenge at the moment for unicorns that have been around for 10, 15, 20 years or whatever, is that they've got all this amazing tech that they've built up for years and years and years. and now they've got to rewire that to AI. And some of it might be able to be transformed by AI, but some of it won't be able to. So what do you do when you've got all of this legacy tech? So I think that's one big issue. But as you said about the kind of advantages, I also think for a company like Zender, the German freight forwarding unicorn, they've got years and years and years of data on the logistics industry.

23:05And that is really proprietary data that's come about from working with customers, speaking to people and that's all in their internal database so that is a real advantage you can't just copy you can't just vibe code zender's platform in a day because they've got a history so i think that is a key probably differentiator yeah and we and again investors are generally pretty bullish on the sort of vertical ai for specific industries for exactly this point that that kind of very specific niche know-how is not the kind of thing that's just available on the internet that comes from much more sort of long-term relationships with particular industries and particular types of clients and understanding their sort of how that all works.

23:50That I think is a fantastic place to end Miriam. Before you go what other topics are really intriguing you at the moment? I'm just really interested in how teams are changing and also how traditional hierarchies are collapsing in companies, also how departments are changing. You know, I'm hearing a lot that we're going to see hybrid departments. So the traditional silos of how people were working are actually going to merge over time. So really interested in that as well. And then also just what methods companies are using to transform their organisations. For example, I was speaking to Miro the other day, who told me that they're actually trying to develop like an internal search engine so that when an employee has a problem, they can just be like, okay, I want to write marketing copy really quickly for this thing.

24:40And they just write in an internal database, this is my problem. And then it will come up with these are the tools that you can use and the order in which you use them to create this thing, which I thought was really cool. That seems like a really interesting way of getting people to use AI. So anything like that I find super interesting right now. Perfect. And that is all we have time for for today. So if you want to hear more from Miriam about the operational side of building startups, sign up to the wonderful Startup Life newsletter on our website. It's free and arrives in your inbox every Wednesday.

25:12Plus, you can also meet Miriam in real life at the Sifted Summit in October on the 8th and 9th in London. Tickets for startup founders and operators are really very reasonably priced. you can check them out at summit.sifted.eu. And one final thing before I go, we're thinking about ways to make this Sifted podcast bigger and better and would love your input. If you could please take the very, very quick podcast survey, which is linked in the episode description, that would be amazing. And one lucky listener will get to win a rather nice pair of headphones as a thank you. Also, thank you to Tim Smith, who produced this episode of the Sifted podcast, and to you for tuning in.