Why Shipping Faster Isn’t Enough in the Age of AI | OpenAI’s Laura Modiano

If you truly are able to scale across EMEA, then you're probably more enabled to scale to both the East and the West faster than someone who's just building for Silicon Valley.

This is Melanie. And this is Christian. You're listening to Follow the Gradient, the weekly podcast with actionable insights about how to build a business while staying sane. you

What happens when the cost of curiosity drops to near zero? Today's guest, Laura Moriano from OpenAI, is seeing European startups cut development sprints from three weeks to just three days. As she points out, with language models and instant prototyping, the only remaining bottleneck is your intention. In the next 45 minutes, you'll discover why Europe's market fragmentation is a strength and not a weakness. How wipe coding is democratizing deep tech for non-engineers, And the exact framework Laura uses to tell if an AI product is truly disruptive or just a temporary wrapper. Laura leads startups EMEA and OpenAI and works directly with founders, VCs and developers across the region. She previously held leadership roles at AWS and is an active angel investor. One more thing before we jump over to the conversation with Melanie, Laura and myself. You guys ask us often, how can you help us? One thing we want to do is to bring you inspiring guests to the show which unpack real life insights about business building. The best way you can support this mission is to subscribe to our show. Subscribe and follow us on our channels. This will already mean a lot to us. And for more hands on content, head over to followthegradient.io. And now enjoy the conversation.

Laura, you spend your time with founders across Europe, the Middle East and Africa, across stages, markets, but also different levels of ambition. Before we get into the numbers or narratives, when you look at European founders right now, what are you feeling most optimistic about, but also what worries you the most? First of all, think I have a real privilege of working with the best founders in the world. And I think some of the best companies and future unicorns that are really going to define what technology in the future looks are building their companies right here in Europe. So I'm extremely honored to be working with this ecosystem. What I'm most excited about is that we have so much diversity in Europe. We have founders with different background, different levels of operating experience, education that tends to be extremely technical as well, and are solving problems across the industry. Whether we're talking about FinTech, healthcare, robotics, education, manufacturing, every single vertical that has opportunity for disruption, improvement, solving new problems with new technology that have emerged recently or old problems. that can just be disrupted and rethought of because of how the technology has changed with first principles. I think founders in Europe are addressing all of those. And it is a joy for myself and my broader team and us at OpenAI overall to be supporting the journey of these founders right here in Europe. And is there also something that worries you or are you like just 100 % optimistic about the European landscape at the moment?

I think always being cautious and providing advice and support so that people can stay on course is important. But yes, I am 100 % optimistic of what the potential in Europe is and not just in one market, but across all of Europe. I think there is no problem, big or small enough, that founders here in Europe cannot address and solve successfully. That is the perfect line to also mention what you recently wrote is that 22 % of all new unicorns in 2025 are European and that this is not just a trend, but due to an activated ecosystem. write with 2026 being the year of execution, expansion and value creation, as you call it. When you look across EMEA, where do you see founders? maybe then still underestimating themselves because you seem to be super optimistic. And when we look at it, sometimes we feel like it would be great that the founders themselves would also be so optimistic to not underestimate themselves. Yes. So when I see founders who are working with OpenAI right now, the ones that I have daily conversations with and the ones that I meet in conferences and meetups and dinners and events and demo nights. The one thing that I consistently hear back as positive feedback, especially in the relationship that we have is that OpenAI has launched multiple models and released capabilities for people to launch faster, launch cheaper, prototype and go from idea to scale. in the most affordable way they were ever able to in history with the best tools. And it feels like you now have a coworker sitting with you. So what is possible from the models and technology? What is the intelligence that exists in the models? What is built for founders and developers specifically? And then how do you put that all together? If you look at the cost of these models and the speed that they have, it has never been more possible.

to go from idea to generation and it has never been more affordable. So you can start fail and iterate over and over again with a much lower risk. And that's why I now say to everyone, my number one advice is just try. This is a year of really doing and then coming to the community, building in public, building in the open, working with people to test ideas out and then finding your path to growth much faster. So it's basically if I conclude what you were just saying is that the thing that potentially founders still underestimate is like the level of speed and impact that you can have with the tools that are around. Yes, I think it is understanding just how much is possible at OpenAI. We ship new features every single day, every single week. You just have to look at our accounts and then you're going to see how much we're doing for developers specifically. We deeply care for developer tools and for the ability for developers to not just have an idea and prototype, but really bring something to life and have working, working prototypes, working models, working companies. that can fundraise that they can bring to customers that they can iterate on, that they can have very fast cycles. And when I compare customers that I work with now in startups to a year ago or three years ago, we used to have sprints that were three weeks, six weeks. Now those are cut to three to six days. Having tools like Codex and then the continuous improvements and capabilities of, know, GPT-5.2, which we recently launched as well, means that whatever it is that you dream of, you can actually build and you always have a coworker with you that keeps you accountable and makes you move a lot faster. So there is no excuse not to ship, not to try, not to build. The only limitation is your intention to actually put time on. So if you want to do it, first of all, we will make it possible for you. And second, we will support you along the process from ideation all the way to fundraising and scaling.

And I think this part is super important because there's this saying that ideas are worth nothing. It's really the execution that matters. the idea and prototype, yes, that's super, super important, but then also the other part of execution. is there maybe also something, because you work with so many startups also in your, your, with your hat as an angel, what is maybe an execution gap that shows up again and again once European startup hits real traction, so to say? First of all, I wouldn't say this is European versus American. When I said, thank you for referring to my LinkedIn post, that this is a year of community. I think last year, it was incredible to see how many hackathons and events and meetups were happening, conferences that really upped the level of really heart hitting at global scale best of. But a lot of it was happening. in one city here, one city there. We were saying this conference is happening in France and hence it's a French conference. No, all of these things, what we're seeing towards the end of the year is that you have hackathons in Berlin or Stockholm or London that people are traveling across the region and sometimes even from beyond the region. You have hacker houses like Freight in Finland that had more than half of the attendees coming from outside of Europe. to Finland to spend three months to actually build something in deep tech that has impact. So that's what I want to continue to see that you're not building just for your market, just for your friends, but you're building something that is a standalone company. And you're not building just for the hackathon to come up with something cool at the end of it, but you're putting steps in of iteration and trying to get a company that actually can fundraise, can be used by customers. and can be relentlessly obsessed with solving a problem, creating value, being a company that has a platform for growth because it is operating on the application layer and understands what the customers that this company is going to need. So my recommendation to everyone is don't aim small, but aim big and continue to shoot further and further, higher and higher, because that is how growth is a choice.

What kind of backgrounds do you see the people have who come fly into these hackathons? Are these the typical hustlers or rounders who've been hustling all their lives? Are these like students in STEM degrees, social sciences? Like, do you see a pattern there? Well, what is great is that first of all, these models have become so good and then the tools around them as well. So we don't have a single profile. I think the AI founders who want to build true AI application layer companies need to be technical and having a STEM education really, really helps with that. So when I look at the hacker houses, yes, most of them have a technical background. However, and very importantly, when I look at hackathons. It is not always the case. And some of the great solutions that are built, some of the winning projects are not always necessarily from engineers or physicists. It's from people who understand how to use the tools. And that is why Vibe coding has come up. This is why people who spend time with codecs to know how to interact with the technology, whether they had a formal university education that was STEM or just really focused on upskilling themselves. And I think Given the tools that we have, you could literally spend as much time as you want on ChadGPT and other tools to just focus on upskilling yourself in a technical area. And when I see people really spending time with tools like Codex, there is no limit to how technical within the area that they want to be, they can be. So again, the self drive and motivation is something that has to come from the individual. What we as a community need to do across Europe, is to recognize when people are focused and let them fly. So put every element in place to let them access capital, to have co-working spaces, to have a community come together, to share ideas, to meet investors, to meet other founders who can mentor them, so on and so forth. Demo nights to show what they're working on in a safe space where people aren't judging it as if it's going to die tomorrow, but knowing that if you are building in public,

then people are gonna support your long-term. And on this point, and I want to come back also what you said, that especially like a background in STEM really helps. Also, we know all the white coding and you can upscale yourself, still STEM background is still super, super valuable. And you reminded us in previous conversations that this really nice stats that most of the people I think are not aware of that Europe produces more technical STEM graduates. per capita than the US or China, and that with AI, many of Europe's historical disadvantages around capital, around proximity and scale matters less than they used to. So if that's true, I'd really love to dig a little bit even deeper into this, like unpacking this tension. Like where is Europe's real competitive edge today and where do founders maybe misinterpret the edge in a way that actually slows them down? So let's start first with why it's important to be technical. If you're building a startup today, I'm going to assume you're trying to build an AI startup. You can build it on your own or you can build with community and then you can use every tool to your advantage. So we at OpenAI want to work with founders who really want to build for scale and growth. And there is an incredible way of working with us, which is partnership and feedback. And I often say you need to have a chief feedback officer And the more technical you are, it means that the faster you can understand how these tools can be used, how they can be applied, and where there is room for improvement. So having a close relationship with us at OpenAI means that you can give us feedback, you can iterate. Every time we launch something, you understand where to plug it in and then what is possible with that. You're also not limited into just tech for tech, but tech for everything, whether it is, again, construction,

manufacturing, robotics, fusion, how to create better performing batteries, transportation. There's never a limit to how much innovation, improvement and ideation you can have and how much disruption because there will always be, it is a fact of human life that we are always striving for improvement. So if we start from that, it means that A, by being more technical, you understand where technology can be applied. B, you can work better with us because you can stay on top of the models. and see as new things come in, whether you have a super, super close relationship with us or not, you can always be the first one to integrate it. You can work on evals. You can test different ways of applying things, different architectures, different prompts, different system prompts, how different users are using it. You can do continuous A-B testing. If you've understood in your formal education how to do experiments and hypothesis testing, then you can do that with your customers. And iterating in that model is where I see some of the best startups succeeding because they're always on an A-B test and they're always looking to work with us on how it is that they're talking to us about what they are experiencing from our models, what it is that they want to see more of and how to work with us on guiding what the future roadmap is going to look like. And that is for me, the ideal relationship that we can have with a founder or a startup. And so I totally get why this is important and also where the STEM background can actually help a lot in order to give that feedback. Now, because we really focus on the European ecosystem and you mentioned a few, like what is now really important that we don't talk about France or like Finland or Switzerland, but that we see ourselves. do want to call out Switzerland for a couple of reasons, other than the fact that you are based there. I've been to ETH Zurich a few times in the last year, and I've been floored by how incredible the energy has been in Switzerland. And again, if we look at new unicorns created by year, Switzerland has had one of the biggest growths. So it went from a small country to now in the top five new unicorn creators in Europe. if I'm not mistaken, I think it's fourth right now.

Germany, and Switzerland, and then France. So it is definitely punching above its weight, partly because there is extremely good technical education. It also is because there is an industry around it where people can prototype and partner with that is also actually very technical. So robotics are coming out of Switzerland, healthcare and life science are coming out of Switzerland, manufacturing is coming out of Switzerland. So traditional industries that can benefit from AI in a way that they couldn't necessarily benefit from SaaS are now seeing the advancement and the investment that is actually paying off. Because students who have for years been learning from their local ecosystem and having a technical education can match the two and are not building in the abstract or theoretical, but are building things for the community and customers and companies around them. And I think this is what having an ecosystem looks like. So you match a great set of universities, because obviously ETH Zurich is incredible, but there's also multiple others, know, St. Gallens there, TUM is not far, Munich, et cetera. So you have a fantastic ecosystem of technical education and educators who are advisors who bring in experience. You have industry around where people can intern or be operators in and then have that community relationship with the coming talent. You also have investors. So now we see multiple funds that are investing locally and also cross regional European funds and international funds, both in the U.S. and other areas in the world that are looking into what the innovation is happening there. And that means a vertical integration. The system works where you have students that have job opportunities that have other startups to look into and up to to say, I have something to dream of and I know it is possible, reachable and attainable. You have investors that are supporting that so you can tinker, can try, can experiment, and then you can get funding and you can have global exposure. And the thing that gets you from it is possible to actually succeeding and scaling is I'm not just building for my local community, but as a community, we're celebrating these successes and taking them to the global stage. And what I've seen in the last year with the ETH Robotics Club, the Agents Systems Lab,

Jay floor that came up from, the founder's house. Now all of those communities have come together to say, we want to have a global ambition. And then you scale it out across all of Europe in every country. hearing similar stories of. are thinking global first. We're all working in our local language, but also in English, because we want to think of customers everywhere. We're looking at problems here, there, everywhere in our country, in our city, in Europe. in the US, in India, in Africa, in Latin America, and we want to see how that can be applied. So if you're building for the world, then you're not limiting yourself in the architecture that you're putting in place on your go-to-market and strategy for just one market. And I think that has been a huge shift in the last couple of years compared to... I really love that you're such a cheerleader. I think whoever listens now, it's just nothing else is missing. We just need to get going because basically back to my question about the competitive edge also of Europe, basically we have everything here. People are working on stuff like we're building these communities. We're more closer than ever. So what is still missing? It's just like... I'll you, I'm coming over in a couple of weeks for another robotics hackathon that Mimic is organizing. And then there's multiple robotics companies that are there and I'm very excited to meet all of them. Love it. Really, really love the spirit.

It's what we're here for. Absolutely. You and you alluded to the ecosystem across the continent. So if I put myself now into the shoes of an aspiring entrepreneur who wants to join the ecosystem, who is free to roam across Europe, what are the go-to places to really join and, and visit like tomorrow or next week? Again, I think it depends what it is that you're building. And you should remember as an aspiring entrepreneur that where you build is what you should be optimizing for yourself and your team. But we now have the ability to collaborate on everything remotely as well. So hire the best talent wherever they are and optimize for them being productive and happy because happy people are harder workers. So if it happens that you are building out of Switzerland and you have a founding team there, but you found a great engineer in, let's say Spain or Estonia, then you shouldn't stop, let that stop you from working together. If you want them to be co-located, you know, absolutely up to you, but it is possible to work on everything remotely as well. First of all, second, I would say, look at what industries are around you as well. So if you are building in FinTech, probably the UK or Netherlands are great places to be building in. If you're building in

gaming, maybe the Nordics are a great place to be in. Or again, I think the UK is good for everything. And then there are hubs that have a lot of excellence. So what that excellence looks like, and why I mentioned Netherlands, for example, with fintech is that you have Adi in there. You now have Duna. You have multiple very experienced entrepreneurs that have succeeded previously that you can learn from, that you can have as mentors and advisors and other operators who have worked in that industry. Or in the UK, you have large banks and financial institutions that you can then work with as prototype design partners. You can have a community of events that are happening within that industry. So if you're thinking of the two main things that you need as a company, as you're growing, forget the funding for a second. It's you need to have a team that is technical, that understands the industry, and you need to have potential customers and people within the industry outside of what of the people building so that you can test and sell to. So where you can maximize for those two, depending on what it is that you're doing, is probably the right home for you. I think Germany has incredible manufacturing, healthcare, and other areas that are now growing. I think the Eastern Europeans are extremely strong at being very hardworking and actually solving multiple problems, especially on consumer. Right now I'm seeing there's a company called Sintra in Lithuania that is doing fantastic. Super scale similarly in in creative space, which has founders both in the Netherlands and Germany So yes, you have some really big hubs, but then across the region you'll find certain areas of specialization So there is no one answer. I think the most important thing is finding a community around you that supports you along the way I want to go back to something you've alluded to earlier, which is that obviously now with AI, idea to prototype cycles are collapsing rapidly. And you've said before that execution is now the differentiation. It's not the access to the technology itself. For founder who already ships fast, like what does great execution actually look like? Or what have you seen on the ground, which actually works out if you're, I don't know, five people, 25 people, a hundred people.

So first of all is understanding the technology. So it's not shipping to ship, but shipping for a reason with the technology and understanding where that technology sits and how you can get the best out of it. I see some of the best founders across Europe. My team is very distributed. We have people across EMEA and we work with companies across that they, again, as I was saying earlier with the chief feedback officer, use us as an extension of understanding what is possible. So not just passively waiting, but really understanding the technology and engaging with us at OpenAI to wherever they can in person and online so that we can help them along the way. So that's the first thing. The second thing is having a clear view of what your customers want and staying very, very close to customers. So feedback is not just how you're giving feedback out to your providers, but also continuously collecting feedback from your customers. of having a structured way of always on listening, always understanding where changes are happening and how each specific feature is responding to what it is that your audience wants. And then understanding where within the four options do you fit in. And the four options that I think innovation comes from are, are you doing something better than your competition? Are you doing something cheaper than your competition, but equally as well? Are you doing something that wasn't previously possible, so opening a new capability, or are you doing something that is entirely new? So for example, are you giving a capability that couldn't have existed before? So is it better? Is it cheaper? Is it different? Or is it new? So focusing on those four, I think, is where, again, you can test the technology and there is a different level of risk and competitive profile within each of those. As you scale, you might end up trying to do more than one of those things, but you should always have a driver for that. And then understanding from your audience what matters to them. So what is the elasticity of the decision-making of the audience that you have, the industry that you're working with, et cetera? Does it have regulatory constraints or not and how you should be working with that? Always coming back and working with us so that we can help you along the way. And then having a clear vision. And I think really good founders today.

have taste and opinion. So really sticking to your own taste and your opinions that are informed by your audience. So you don't have to be super hard and fast, but continuing to have a point of view, I think is what people are also now appreciating very much in the market. There is one follow up question I have on this point. Do you see any patterns in adoption on the customer side? So you're supporting obviously a lot of companies and they're working very closely with companies who work with AI, utilize AI to leverage their own products and their business model. Do you see any patterns on the customers of these startups, for instance, in certain industries, certain age cohorts, countries maybe, which are quicker to adopt these kind of AI native technologies than others. Sure. this is again, going back to very basics, what AI opportunity looks like. I often refer to this two by two. I think AI innovation happens either, or AI tools are there to help either on the enterprise side, your internal staff, so your workforce or your customers. And then either to have new ways of finding efficiency, productivity of existing workflows or giving the capabilities to do something new. entirely new. So if we take gaming as an example, which I often use, if you say we as a team, we're always designing characters and this would take very long doing all the different characters. Now you can use AI to say, drew a troll with a, an axe and automatically I want to change this axe to a sword. That is internal for my workforce force and efficiency, productivity, existing workflow, just improving that.

That is where I think a lot of people on the enterprise side think AI comes in and a lot of the application layers until now, startups in the application layer until now have really focused on. What I think has great opportunity is the other three quadrants. So how do you look at what your customers are seeing and what the new use cases are? So going back to the gaming example saying, I'm playing a game with multiple characters. There's the playing characters and the NPCs, the non-playing characters. With AI, I can now give them all a voice. And every player can interact with every single character there. And I can use AI to create a whole new universe within my existing game. So that's customer-facing, new use case. That is actually a pie that we don't even have the size of right now and can continuously expand, both on the enterprise side, but also, obviously, on the consumer side as well. So if we think of the creator space, code generation space, This is where we're seeing really big transformation and focusing again on what is new use cases, I think is something that will always have more opportunity if you get it right. But to get it right, you need to have an opinion. Thank

One thing that I also found super interesting that you said when it comes to founders is that you basically highlighted the unlearning as a core skill of great founders. And you were very concrete about it saying unlearning processes that no longer serve users, stage specific behaviors that stop working, tools that create drag, or designs that hardens too early and even a vision that stays too small. And I think you were alluding to that earlier in our discussion as well. When you look across founders who've raised real money and have real traction, which one of these today almost always unlearned too late. And why is that so also from a psychological perspective? Why do you think this is hard to let go? By the way, this is true for startups as it is for mid-market all the way to very large enterprises. Unlearning is a skill we should all have because yes, you can learn things, but what you know should always be context relevant. And that's why, know, at OpenA.io we think a lot about context, but this isn't just about your conversation within ShadGPT or Codex remembering what it is that you like in terms of how you code. It really is in everything that you do professionally. Just human.

potentially also personally, but professionally, you should always have adaptability and context in how you operate. So a company that, or a founder that starts a company and is looking to hire an immediate team to just build an MVP and go to market is very different than I now have a hundred employees and a thousand customers. And that will be very different to I have 10,000 customers. So what processes work at each stage of your company's life? is going to be very different. And I think the more we across industries only look to promote a certain way of operating without allowing new culture to come in and new ideas, I think that could be one risk. The other thing is as companies grow and they hire more experienced leaders, making sure that those leaders bring in some of their experience and skills and tools without necessarily bringing in a specific culture from wherever it is that they came from. Because what makes companies great is also the culture that they have in each culture is unique to that company. The same way that the same system prompt would not work for five different companies and you have to adapt it to the audience that you have and the product that you have. I think the same applies for internally how you're thinking about your own culture as a company. It reminds me of something that Noah from Neko Health, she's the chief people officer said, and she's very, very much into culture. And she said, if your culture can be copied, then you never had a real culture. So it really is what you're saying. And I completely agree with you. Now, my question would of course be, have you seen any frameworks, any ways, be it coaching or anything that have worked for the founders, but also for the operators? how they can actually adapt to that.

So I think what startups have, which maybe larger companies don't always maintain as much, but is incredibly beautiful to see in startups. And I'm so proud of the startups that have that is they give everyone agency. So everyone has a real voice and they give a lot of collaboration between operators. Traditionally, you have a research team and a engineering team and a product management team and a marketing team. And the bigger we get and the more political. the structure of organizations get the more separate these teams are. But actually, the closer you bring them together, and I see this when people use OpenAI's tool, when they have shared canvas and shared projects, and especially now with capabilities where you can collaborate on things, it allows for voices with different perspectives to really influence the product for the betterment of the product and for the success of the company. So saying a... product manager and an engineer and a marketing person should be separate or a comms person, actually, I don't think is always right. I think bringing them in the same place and saying, well, all of you are here for the same mission of the success of this company and the growth of it and getting more customers and being more visible in the market and really understanding our audience. So what are you seeing from your point of view and how can you combine these voices is what allows startups to move a lot faster. And especially at a younger, stage of a company, everyone gets to see so much. And I think it's, it's where a lot of, as someone who's worked in startups, this is the happiest part of feeling like you get to contribute to different parts of the company and having the agency of saying, this is my baby, this company and its success and its mission and its customers and their needs are what drive me and working together in a very collaborative manner. is what is all going to drive us all as a team to success. And I say that by the way, as someone who's worked in very large organizations from the NHS and AWS to a startup to consulting. And as a consultant, I had seen multiple different cultures working at OpenAI. I have never ever seen a better culture because we are so mission aligned as a team. And whether you take someone in...

you know, go to market or research or marketing or comms. Everyone is here because they truly believe in the mission and that we're here to do something important in the world, but also for the people that we work with. Like we all just want our customers to succeed, to do well, to have every advantage and tool in their arsenal to go as fast as possible. We're just here to enable others. So when I look at my team and I say, hand raise, I need help. Can someone jump in because there's something I don't know? Everyone's going to try and help me. We're all celebrating each other. We have Slack channels that are literally celebration channels where our wins are shared. And we all try to find ways to be the voice for each other. And if, by the way, there's something that is important that I see that could help someone else, I go and I say it. And we all appreciate this collaborative nature. So maybe OpenAI is now a a very important company in the world and we're bigger now, but we still have that startup mentality of be lean, be fast, go as much as you can towards a mission aligned, speedy, collaborative way of operating. And that is what drives us to success. So I'm extremely fortunate. be part of this team. And on this point of speed, because I'm totally with you and I think we both like we work more than we should and we love it. And, we, we particularly say it in our, like when we build up this podcast, like it's about how to scale from Europe while staying sane. So the sanity part is also super important because we're very, very convinced that if you're also kind of trying to make sure that your health is not completely off, that you can actually reach even higher stars. So my question for you would really be like, what is it that you have seen that also helps beat founders or just personally yourself to not confuse intensity with effectiveness? And really what are maybe boundaries or stuff that helps you to also find this balance, so to say, to actually get enough energy to do all

I can't speak for other people or other companies, but I can say at least how I see my team operating. I some of really good startups have a similar model is that you need to not always just always be on. are different, there are right times for different things. So there are sprints, there's learning time, there's testing time, there's building time, there's shipping time. And again, different members in the team. might have different intensity at different times. So if you think of the holistic entity of the company and then taking the time at the end to celebrate, to learn, to understand what happened, to do retros, those pauses also really, really matter because I think how growth doesn't happen linearly. The same way that, know, there's not a new thing that's shipping every day. We have to create a model. We have to test it. We have to iterate on it. And then you have to ship it. Then we have to see how the market deals with it. And then when we have a new model, it's specifically to have improvement on things that we think matter. So this is a learning process. And again, you could say the same thing for individuals, teams, companies, whatever it is. So I think instead of thinking of growth as linear or going like this, it's you, there's a jump that happens, then there is consolidation, and then there's another jump for your own personal learning. So taking those pauses, to refill however brief they are matter. And then the other thing is that understanding that different individuals in your team have also a different situation. So I'm a mother, I have a 10 year old daughter, I love my child, but we often have conversations and we say, know, there are times where I just need to focus on something really important at work. When this is done, I will be a hundred percent here for you and we're gonna do homework and math tests and go skiing on the weekend or whatever it is. And that is equally as important because being able to show up as a mother means I can also show up as a whole person at work. But when I'm focused at work and we're launching something or I'm here talking to you, I'm only doing this and I'm a hundred percent doing this. So for me, this is the way that it works. But I know that if I'm doing this, then someone else might be doing something else and having the flexibility to give people the space to always show up in their best self.

And this is also, as a mother, think what really has helped me is also to put in the life in your calendar. Because we tend to just put in the business stuff in our calendar, but just to also put in whatever, be it a ski trip or be it whatever it is, also in your calendar. So it also gets the priority it deserves. So just thinking of the Christmas holidays, we had a few days off obviously for Christmas, New Year's. And I really try, I think it's very important for people to again, continuously upscale themselves. So I decided to spend a lot more time on codecs and then test tools like cursor and different vibe coding tools just for me in a quieter time when I didn't have a deliverable every single day to say, test different things. I selected four different application tools. so lovable was one of them. Cursor was one of them. I was on canvas on Chad GPT and I was saying, well, if I say I want to build an application and I just write my prompt and actually personally, my workflow is I don't write, I use voice mode on Chad GPT and I literally dump all my thoughts and I say, okay, well, before writing something down, I have to organize my thoughts and sometimes writing, leave some things out. So I just speak. So I, you know, I'll speak for like three minutes and I'll ramble them. I have this and I that. actually. You know, maybe, and then I'll have a full conversation with Chad to iterate on my idea. Then when I get to, it knows what I want. And this is really from where I started in my mind to what is now on screen is really what I was imagining. Then I started saying, okay, create a PRD for me, or let's now wrap this up into like one prompt. And then I, what I did was, you know, I went to the different IDs and I was like, okay, well now take this. prompt that I had on ChadGPT and build me the app. Now take the PRD that ChadGPT wrote and write it out. And then I would compare what the different output was like, because that way I had my own eval's process of saying when I explain what I want first time versus when I've worked through it and reason through it versus when I've worked with ChadGPT, what is it that changes and how does the system then understand me and which tool do I like the best? And I think that's something that you need to do.

for many things as you're building a company, because again, different audiences will want different high-end front-end and back-end, et cetera. So having that A-B test always on matters, but you should know what works for you. And again, the customers that I work with the most, the reason why they love our tools and work so closely with us is because we make it possible to do this. We have an applied evals team that creates a harness and says, we'll do applied evals for you as a company. So what I did for me as an individual, to see if the canvas output on Jadgbt versus the lovable output was better, or if I went straight on cursor, what it would look like, you should do that as a company. And that's why we spend so much time thinking as OpenAI, what works best for whatever your company is trying to achieve. It's pretty interesting because basically it also goes back to what you said about iteration and learning and unlearning, right? Like you can apply that to your own life as well. One question there for also founders or startups. mean, a lot to what you're describing or what we've talked about, it's about making like super uncomfortable trade-offs. It's what do I have to learn? What do I have to unlearn, stop doing, continue doing? And so if we take a lens. Or if we want to give the founder like a decision lens for a moment where they feel, okay, everything seems urgent. Everything seems high stakes. Everything seems to be the right direction. And I could be going from idea to prototype tomorrow. What kind of questions can they ask themselves to move down the right path? So first of all, you the example I just gave of like on the holidays, I tried this, I was able to do all this extremely important learning for me and how I use my tools because I had a quieter time. If I was in full execution mode and I was like, I'm launching something next week and I have, we're prepping dev day, whatever, then I wouldn't be able to do this. I wouldn't take the time to go deep and really get that full understanding and personalization for me, which is essential for me to get to the point where I'm fully executing every single day.

So knowing when to do what is extremely important. Carving out time. So as you said before, having a calendar to say, this is a learning time, this is a sprint time, this is a launch time, this is like full execution, focusing on team, focusing on product, focusing on launch, focusing on marketing, whatever it is. A, not everyone should always be on in one thing. So keeping some people in the team who, if you need to... If you need a team to be dealing with many, maybe like problems or something, you know, went wrong, then you shouldn't be taken away from execution, but you should always have people who can help on everything else that comes in. Like you should always anticipate unknowns and unexpected carving out, as we said, the times and then always being aware of the fact that not everything goes to plan. Like plans are there to guide you. It is like a, a map and a path, but there might be a little bit of deviation. You never know when there's going to be traffic. You never know if there's going to be an accident on the road. So as much as a plan is important, being able to continuously adapt to whatever the situation is, given the context, as we said earlier, is critical. Also, you don't always have full information completion. So if you're working on launching a product and suddenly we say, there's a new model that we're releasing next week, you have the choice. Do I launch my product in a week? Or do I launch it today? Do I want to wait and see how this new model is going to give me further capabilities and how it's going to get better or not? Or do I launch today, focus on everything I have? And then when the new model comes in, I'm going to embed that because I am building in public and I'm building on trust and shipping fast, shipping often with my customers. So what matters to you and your company and the type of company you have, what your competitive environment looks like, et cetera, I think. or continuous unknowns that you have to dynamically adapt your plan on.

And to follow up on that or go deeper, you said it, the model makers in general are releasing so many different features and with every feature you have the feeling you go to X or to LinkedIn and it's like another hundred startups have been killed. Another YC cohort has been killed. Given that you guys are working on so many things, what should startups be focusing on? What makes sense? Yes. So first of all, shouldn't be focusing on exaggeration or only one news source because concentrating voices that are wanting to push a specific message isn't necessarily what's accurate. think there's many startups that are doing extremely well and our job here is one and one only. I have one job and that is to make startups successful. It is what I wake up with and I go to sleep with. So all I want is for startups to be successful in everything that they do and in every vertical that it is that they ship in. What I want them to also know for a fact is that we are going to continuously improve our models and our capabilities to make it possible for them to build on the application layer solutions that are verticalized, that are for every industry, that are really creating value, but our models will get better. So if you're working at something that is at the edge of the model, if you're working at something that is very much a capability that is extremely horizontal, then maybe switch your bet to OpenAI will continue to improve for the community and focus on what is application layer really focused on what goes on top of the model. How do you take what's in the model and create something that provides value for a cohort or an industry or a type of user? We are closing in on our conversation as a last section. Let's call it. We have what we call the rapid fire action, which is I'm just going to ask you four different super short questions and you can answer very shortly to whatever comes to your mind first. So first question, a European founder myth you'd like to kill in 2026.

that they can't be the best in the world. one thing European founders should stop copying from Silicon Valley. doing tech only for tech around them. So I think you can build solutions, not just for your friends or coworkers or your immediate community, but think of a problem as it touches a whole industry. the most underrated advantage European founders actually have. This is a challenge, which is an advantage, and that is the fact that our market is fragmented. The fact that we have market fragmentation means that by definition, if we're looking in our own neighborhood, we have to understand different regulatory, different ways of purchasing, different ways of operating, different cultures, different language. So we're building for international from day one. If you truly are able to scale across EMEA, then you're probably more enabled to scale to

you know, both the East and the West faster than someone who's just building for Silicon Valley. Last question, if you could mandate one habit for every scaling founder, it would be... Talk to OpenAI, give them feedback, ask them what you want for. And now for the founders listening who are right now in the middle of building the Altyg listing, what would you want them to feel urgency about in 2026? Keep shipping. Just keep shipping, show up, keep learning and give me a call. I'm here to help you.

Awesome. And on this point, I'm sure that anyone that is listening wants now to know how do they actually get in contact with Laura? So how do they get in touch and how can they reach out to you? So we have a very complicated email. It's startups at openai.com. Just email us. If you have a technical question, especially, we will try to answer it. But also we have openai.com slash startups. And you can find information about where we have events, how to attend build hours, which are virtual. And there's a library to see build our videos that we previously recorded and your resources if you want to apply for credits or anything else. we're here, we're friendly, we're available. and we're here to fight for the European tech ecosystem. Thank you so much, Laura. It was such a pleasure talking to you. I love seeing you cheerleading all the founders in Europe and beyond. I'm looking forward to seeing you next week. Thank you so much. Bye.

Thank you. Bye. And this was another episode of our Follow the Gradient Podcast. And now we would love to hear from you. How did you like it? What should we improve? Do you have any suggestions, ideas, or dream podcast guests? Please let us know. Thank you and see you again next week!