Paris AI Startup Illuminates History and Culture: Ask Mona

The discussion featured Marion Carre and Valentin Schmite, co-founders of Ask Mona, alongside OpenAI's AI policy lead, Martin Signeux. The conversation explored the intersection of AI and culture, particularly how AI can enhance museum experiences, democratize knowledge, and expand accessibility to cultural heritage. Ask Mona has been using AI for nearly a decade, well before AI became mainstream, to create interactive, personalized experiences for museum visitors. The company’s work spans major cultural institutions like the Louvre and the Center Pompidou, as well as educational applications.

The discussion also touched on broader themes, including AI’s role in creativity, how AI is evolving, and the ethical considerations in using AI for knowledge-sharing. Concrete examples, such as AI-powered interactive guides in museums, augmented school books, and "smart magnets" that enable conversations with historical figures, highlighted the impact of AI in making culture more engaging. Additionally, the conversation addressed AI’s ability to facilitate trust in cultural institutions and how AI could play a role in education beyond traditional settings.

In this event, we are welcoming two guest speakers, Marion Carre and Valentin Schmite. Both are co-founders of Ask Mona, which is a pioneering company using AI to revolutionize knowledge sharing across cultural and educational sectors.

Ask Mona is collaborating with institutions like the Louvre, Center Pompidou, and Marion and Valentin will be joined in this chat by my colleague, OpenAI's AI policy lead, Martin Signeux.

I have a lot to say about these special guests today, and their bios are absolutely incredible. To say a little bit more about them, Marion Carre, she is the CEO and co-founder at Ask Mona. Marion is an entrepreneur, she's an author, speaker, and artist exploring the intersection of art and AI.

In addition to co-founding Ask Mona, she also serves on France's strategic committee for generative AI and culture, evaluates projects for the European Commission, and teaches AI and art at Sciences Po and Sorbonne. Marion has authored several books and exhibited artistic work, receiving honors such as Forbes 30 under 30 and Chevalier in the Order of Arts and Letters.

Her co-founder, Valentin Schmite, is the managing director at Ask Mona. Valentin is an expert in AI and culture, co-founding Ask Mona to help cultural institutions engage audiences through AI-driven tools. Valentin has taught courses on AI's impact on culture and has explored AI's role in creativity through his work bridging technology and arts, making culture more accessible.

I could go on saying so many more things about our two amazing guest speakers today, but I'll leave it to them to dive deeper into all this work, so welcome, Marion, Valentin, and Martin.

Thanks a lot for the kind introduction, Ben, and good evening, everyone. I'm really, really excited for this conversation and to be joined for that by Marion and Valentin to explore this topic about technology and culture and the intersection of the two.

I think it's a topic that is really, really exciting with many dimensions to explore. For example, how technology can reinvent experiences in museums to enable audiences and visitors to have a deeper connection with the artwork they are viewing and trying to understand. But it's also about, for example, individuals being empowered to get a better grasp of the knowledge they are trying to acquire and have a more personalized experience in education. And at a deeper level, it's also a question of the interplay between technology and creativity and whether AI is creative in itself or if it's just a tool for creativity.

So I'm really, really thrilled to have the opportunity to explore these topics with our two guests, Marion and Valentin, who co-founded Ask Mona a couple of years back. And I think this is like an interesting point to start with. How come Ask Mona started to use AI so long ago? Because before AI went mainstream, you were already using and embracing the technology to basically give greater access to culture and knowledge. And we're really curious to learn more about the history of Ask Mona and how come you were using AI so long ago already.

Well, thank you very much, Martin, for this question. And I'm so glad to be here with you guys to share about our journey at Ask Mona. It's been almost 10 years now since we started Ask Mona. So it's been a long time in startup years. It's been so long for us, 10 years of working with AI and working at the intersection between AI and culture.

And 10 years ago, approximately, we had a dream with Marion. Our dream was to be able one day to talk to a work of art. And this is where our name came from. Ask Mona is obviously Ask Mona Lisa.

So how you can ask questions directly to a work of art. And we worked relentlessly in this field, first with the first NLP models and trying to build a chatbot to understand cultural preoccupation and to give our audiences some recommendation around cultural venues. And then progressively turning this device into an audio guide that you can talk to.

And so this is one of the products that we are building at Ask Mona. We are creating AI solutions to create interaction between visitors and museums.

And maybe to add on top of that about the why, like why did we decide to do that? We have this feeling of the museum experience, which can be sometimes a kind of one size fits all in terms of contents, where everybody comes inside the museum with like different background and sensitivity and wants to learn and know about different things.

But on the labels, most of the time, everybody has exactly the same contents. So we wanted to explore and with the new relative AI, it became a possibility how we can use this technology for everybody to be able to still consider to ask any question. Because sometimes people tend to be shy asking questions when they're not comfortable in the museum. And so the AI can be here to answer their question in a more like personalized way. And of course, they can talk with guides.

And I think the best experience possible is to have one guide for easy visitors. But as you know, it's impossible. So AI is a good complement and complementary solution.

Yeah, I think this is fascinating to see that the technology can be used to like have a more personalized learning and experience, especially in the cultural sector. So maybe why don't we talk a bit about like some concrete examples and maybe some of the examples you've been the most proud of in terms of like realization with museums, in terms of like concrete experiences you've built over the past few years.

Do you have like some things to basically illustrate the mission that Asmara has been delivering on?

Yeah, go ahead. Yeah, maybe before we show anything, I can talk about what we wanted to build. Because as Valentin said, we were in the field for years, but choosing an LP algorithm.

So it means that when someone was in front of a museum, not wanting to ask questions, we have to think about all kinds of questions they could ask. And it was so tough and difficult.

So with the rise of generative AI, we saw it as an opportunity for us to be able to bring real conversation, but we had to find a way for the generative AI to use only contents that was validated by the museum team.

So we worked hard to be able to use LLM and open AI LLMs, but for the museum to be able to control the data they put inside it. So that's how we started working on this.

And maybe a quick more technical question on this, how have you seen the evolution of the technology since you've been starting to use it?

When you think in retrospect, when you started to use these tools, and the state of the technology today, what is your feeling?

Have you sensed like a true acceleration of some rapid progress?

Are you seeing like incremental improvements over time?

What is your sentiment about that?

Well, every week there is a new tech to try for us, and there is new ideas that is becoming possible.

In 2022 when we started experimenting with the LLMs of open AI, we understood that was a huge possibility for us and we tried some small tests on our project.

and now everything is becoming bigger and with more features. We are developing things with real-time API, so it's really easy for end-user to have an interaction. At the beginning, it was only text. Now it's vocal, but it's also visual. So the multi-modality of the AI is becoming something truly meaningful for cultural experience. So we are expanding every day our experiences in museums thanks to this acceleration of technology. It's a bit difficult sometimes to keep it up because there are so many new things and we need to be on the verge of each technology, but it's truly rewardful when we can integrate those technologies in our experiences. Even as an insider at OpenAI, I feel sometimes the vertigo of the pace of technology. So yeah, I definitely feel that.

Marion, you wanted to say something?

Yeah, maybe to start talking about a concrete project so it will be more clear for you. We have a video for you, but just to say a few words before. It's a project we developed with the National Museum of Fine Arts of Quebec in Canada. The way the museum team developed this project with us is that they gather some contents they have about the artwork. So for example, text, images, videos, and various contents. And they put it into our system so that the visitor can... So you will see it on the video, but they can take a picture of an artwork and ask any question they have and the AI will answer. So we can let the floor to the video.

♪♪♪

J'ai cru voir des personnages bouger dans l'oeuvre. Est-ce possible?

Vous croyez avoir vu bouger un personnage?

C'est normal, puisqu'il s'agit d'une oeuvre vidéo invitant le spectateur à ressentir de l'empathie pour les personnages qui luttent pour maintenir le repos exigeant.

Sur la toile, je vois du soleil et de la pluie.

Votre observation est tout à fait pertinente. C'est un bel exemple de la manière dont Fortin utilise les éliminerelles pour créer une ambiance et transmettre des émotions.

Quel modèle est la voiture?

Il est le modèle de Mercedes décapotable, mais pas d'une voiture ordinaire. Elle représente un symbole de réussite économique et dit compte de la société de consommation, tout en étant un simulacre artisanal et rustique de la voiture luxueuse réelle.

N'hésitez pas à m'interroger au sujet de cette oeuvre.

Je pense que c'est une très telle illustration de ce que vous avez réussi à atteindre et de la façon dont elle crée une connexion significative entre le visiteur et les œuvres et les arts qu'ils voient.

Pouvez-vous nous expliquer un peu les institutions avec lesquelles vous partagez, au-delà des musées, ou même entre les musées, quels sont les musées avec lesquels vous partagez? Et peut-être aussi sur les défis que vous avez rencontrés dans les différentes partenariats avec lesquels vous partagez.

Pour répondre à vous, Martin, je peux vous donner une réponse courte. Mais avant ça, je vais essayer d'expliquer les résultats concrets de cette vidéo à l'intérieur du musée. C'est quelque chose que nous voulions vraiment stresser. L'objectif du musée était de permettre aux gens de rester plus longtemps devant chaque œuvre d'art. Ils avaient beaucoup de visiteurs, mais ils passaient juste par, et ils ne s'arrêtaient pas devant chaque œuvre d'art.

Donc, nous avons essayé de créer ce système pour créer une plus longue connexion avec l'œuvre d'art et une plus sentimentale connexion avec l'œuvre d'art. Grâce à l'IA, nous avons multiplié le temps qui passait de cinq secondes devant l'œuvre d'art à cinq minutes. Maintenant, chaque personne à l'intérieur du musée qui utilise notre technologie, passe cinq minutes devant chaque œuvre d'art, et pas cinq secondes. Et c'est quelque chose de très et vraiment significatif pour eux, parce que c'était l'objectif principal de permettre aux gens de rester plus longtemps devant chaque œuvre d'art.

Et c'est clairement un objectif que nous avons accompli avec eux grâce à cette technologie. Et pourquoi ? Parce que cet aspect de questions et réponses fait une expérience plus interactive et immersive devant chaque œuvre d'art, et vous pouvez très bien s'intéresser à des sujets spécifiques autour de la vie de l'artiste ou de la technique de chaque œuvre d'art. Et c'est un bon moyen d'experimenter un grand champ de questions autour de ces sujets.

En plus, et peut-être d'un point de vue plus technique, il n'y avait pas d'application d'enregistrement. C'est quelque chose qui est vraiment propriétaire du musée. Vous scannez juste un code QR et vous entrez dans l'expérience. Et c'est en train de fonctionner comme une solution purement web où les gens peuvent prendre une photo de chaque œuvre d'art et avoir du contenu, et pouvoir poser des questions autour de l'œuvre d'art.

Et désolé, c'était en français, québécois, mais ça fonctionne, évidemment, dans 27 langues différentes en ce moment. Donc, on peut opérer en chinois, en anglais, bien sûr, en espagnol ou en italien. Donc, on peut travailler dans différentes langues et à chaque fois, on peut choisir une tonalité spécifique pour la voix, mais aussi pour le ton éditorial de chaque projet pour faire une connexion significative et une solution propriétaire pour chaque musée.

Et peut-être juste pour donner à notre public un sens de la profondeur des expériences nécessaires.

Combien de temps prend-il pour créer une telle expérience, par exemple ? Entre le moment où vous commencez à travailler avec le musée, le moment où vous devez construire la technologie, obtenir les données du musée pour s'assurer qu'ils restent en contrôle de ce qui est dit ?

Ça dépend du nombre d'artworks disponibles, mais s'il y en a quelques-uns, on peut avoir la première version en 3 à 4 mois. En fait, on aime commencer petit, et c'est toujours le feed-back que nous partagions avec la communauté du musée quand ils veulent faire un nouveau projet d'innovation, c'est de commencer petit, faire des expériences, mettre ça dans la main de votre public, et puis vous pouvez ajouter plus d'artworks et passer plus de temps sur ça.

Mais nous sommes vraiment orientés et concentrés sur le feed-back des visiteurs et sur ce qui se passe quand vous mettez ces outils à l'intérieur des mains des visiteurs.

Et pour revenir à la question précédente, nous travaillons beaucoup avec des musées, de grands noms, comme vous l'avez mentionné, le Centre Pompidou, mais aussi des musées plus petits. Nous travaillons avec des sites d'héritage, par exemple, le Colosseum de Rome, l'Arbre Triomphe de Paris, et d'autres. Nous travaillons aussi avec des joueurs de la salle d'éducation, parce que, comme nous l'avons dit dans l'introduction, notre objectif est d'utiliser l'IA pour partager plus de connaissances.

Nous avons eu des gens de la salle d'éducation qui sont venus nous aider. Par exemple, nous avons eu le premier texte augmenté par l'IA. Donc, vous avez un texte, désolé, je n'en ai pas un avec moi ce soir, mais vous pouvez scanner un QR code sur la page d'entrée et poser des questions à votre texte.

Et le texte vous répond à vos questions, mais aussi vous pose des questions pour voir votre examen. Donc, c'est le genre d'outils d'utilisation que nous avons dans le domaine de la salle d'éducation.

Oui, et nous multiplions nos projets dans le domaine culturel et aussi dans le domaine éducatif afin d'utiliser l'IA pour partager plus de connaissances. Et je vais vous donner peut-être un autre exemple. Nous vous avons montré le premier texte.

Sure, here is the improved transcript with appropriate line breaks between speakers:

audio guide that you can talk to. Mayo talked about the augmented school books with AI and there is something maybe a quicker object that I want to show to you that what we call the smart magnets. So this one is a Van Gogh. So you know magnets are those small objects that you can find in museum shops. So this one there is Van Gogh on this one and at the back of it there is a QR code and if you scan the QR code you can have a conversation with the AI of Van Gogh and the idea behind it is to create a database with all the history and the exchange of Van Gogh with his brother and a huge database around the life and the work of this personality and then you can create an interaction via AI to create those forms of interaction and so we did it with Van Gogh but we did it with plenty of other characters so you can talk to Napoleon for instance, Louis XIV, Marie Antoinette and plenty of other ones. Even Abraham Lincoln we did one for US museums so there is this idea that AI is a great way to impersonate also some specific characters and to create a meaningful conversation in order to share knowledge not just from scientific point of view or an educational perspective but also from a more incarnated experience through those smart magnets that we developed.

And actually it makes me think there could also be ways to have cross-cultural or cross-historical perspective. If for example you take two devices with the real-time API and two different people Van Gogh talking to Monet or Van Gogh talking to Bonaparte and you make the two of them discussing together it would be really funny to try this experiment I guess.

And actually I want to ask because you mentioned also that you were exploring bringing to life also more like daily life products for example like bottles of wine this kind of thing but then also it's like bringing perspective to like the heritage that is also part of our daily life culture because culture is not just something that is in museum it's also something that is all around us.

Can you maybe tell us more about like these ideas?

Yeah most of the time we say that we operate in the market of curiosity which means that it's where people are curious and want to know more about something and actually you take this example of wine bottles which is also something in France which is part of our heritage in some way and so it's the kind of daily life object that we augmented in a way that you can better know about how this wine is made and how you can drink it with which kind of food and those things where people are curious and wants to know more about that's where we think we can bring value to the France.

Yeah and of course we are French so it's obvious that we will work with wine bottles and museums and education because it's something that is defining French culture and so in the end with the wine bottle you can just ask any questions and have your small sommelier in your pocket thanks to OpenAI API and then you can ask any questions and get the knowledge of sommelier in your pocket and this is truly powerful for winemakers.

Yeah and we focused a lot on the digital experience for interesting but what is also interesting for the organizations we work with is that they give feedback and insights from the conversations people have with the AI so it's a good way also for them to better know their audiences.

And do you have any like stories of feedback to tell about this for example a museum that all of a sudden discovers that they are certain pieces of art or certain topics that were more explored by their audience compared to what they were expecting?

Yeah well we had plenty of cases and maybe one of the anecdotes that we have is it's maybe a bit trivial but it's interesting. We worked for years with this huge and fantastic space in Paris called Grand Palais. This is where actually the AI summit took place weeks ago in Paris. This is a museum and an exhibition place in Paris and one of the questions that people were asking the AI a lot because we created an AI companion for them is where were the restroom and then the company had to close, to shut down, to have a renovation and they transmit this information to the architect and now there is a new way of a new signalization for the restroom based on the feedback that the the user gave to the AI. So they can make meaningful and maybe a bit trivial change in the process because we gather some data and some preoccupation about about these specific kind of things.

You know to have a more artistic anecdote for instance we created also a tool that is accessible for the Louis Vuitton foundation and some of the people where we asked them some testimony some of the people say we didn't want to to ask a question to a physical and human guide because we were thinking that the question could be a bit dumb and we don't want to and people can feel illegitimate to ask those questions but with the AI it's easier to ask any form of questions without any form of judgment. So it's a great way also to expand the access and to democratize this mediation between an artwork and a visitor.

Yeah this is I think really interesting. I think two things that you mentioned are really interesting to me. The first one is oh you mentioned the AI being like a driver of curiosity too. I think this is like something that is truly important in the sense that it expands the horizon and the depth of things people can get access to. So I think this is really a good thing to see like AI being like basically an enabler and like a maximizer of curiosity. I like the way you presented that.

And the second thing is like only AI can basically have a meaningful impact at the end of the day and can redefine in certain locations the way they are basically building experiences and these kind of things. Something I want to ask maybe is like beyond the different partnerships you've had with museums about like the experiences that they've built, have you tried to build bridges between different museums on certain collections and on certain like era of or movement of art and so on?

Yeah well we are trying to work with museums you know because sometimes an exhibition can can travel between different museums and now we are working for instance with the Centre Pompidou on that kind of matter. So we created an experience and now we are moving this experience from one museum to another because the exhibition is moving too so we can the device can move on that matter. But we also worked on how to understand the collection of not a museum but a group of museums and how can we make links between those artworks from different museums thanks to AI. And so you can ask your questions to a huge database and this question can make questions to a huge database and this question can make links between different artworks. We worked with the groupment of the museum in Paris that is owned by the city of Paris and we worked with them on this collection website that is gathered artworks from nine different museums in Paris and you can ask questions and make links between those different artworks and it's a great way to make meaningful connections and connections that were not expected before. So if for instance you want to see something blue and something

new. Well, you can have different artworks that can emerge, and thanks to AI, we can analyze different metadata that can bring you connections between artworks that were never made before. This is interesting. And also, I think a question that everybody is like, that is top of mind today when you speak about like the intersection of technology and culture, is a question of creativity, and a question of like whether AI can be creative in itself, or whether for artists, this is only a tool. Because I think we've talked a lot until now in this conversation about like using AI to basically get access to like existing arts and existing like artworks, but we haven't talked much about like AI being used maybe to create more or to add new dimensions. Do you have like strong views about this very like live and hot topic in the community? Do we have the night to talk about it? Joke aside, this topic is fascinating, and it's also been years that we've been thinking about that.

On my side, I would not call the AI creative, because I think that somehow we tend to anthropomorphize the artificial intelligence by using like human quality to define the AI. And the AI is great, and brings a lot of interesting things, but we could find other ways to qualify what the AI is doing. But of course, AI is a great way for artists to expand their creativity. And maybe I can take various examples from projects I've made. I've worked on the more like inspiration side of AI as an artist. For example, thinking of AI with some visual inspirations I had to generate new composition and then reinterpret them in the way I wanted to. So, the AI was huge for me.

And there is another project that I've done a few years ago during a master's in Scotland, where I was working with the archives of the National Library of Scotland. And my goal was to create fake archives. But those archives were in like kind of old English, tricky to reproduce when you're not like an old English person, I would say. So, I used the AI to generate some real and fake archives, because I wanted people to face those kind of real and fake contents, and be able to decide whether they want to archive or destroy the documents. And so, a lot of them destroyed real archives and archives fake ones. But I wanted to ask people to be able to add themselves questions. So, the AI here was more like a technical help for me.

I think AI is a fantastic tool for artists. And I think it's as powerful as the beginning of photography, for instance, for artists. And there's many parallels that we can draw between the beginning of photography in the 19th century, and what's happening right now with AI and artists. Because at the beginning of photography, people were saying, well, you just have to click a button and it's not really art. And there were so many questions about right and how it's going to destroy all the painters and everything was not true, actually. And photography was a great way to create and to expand a new field. And it was a new tool to create a whole new artistic field. And I think it's the same with AI. People are saying, well, you just have to write the text to create a work of art. It's not true. It's really hard to make true and meaningful work of art with AI. And you have to work through it with passion and with an artistic vision to make a true work of art. And people were saying, well, it's going to destroy the work of artists. Well, I'm not sure it's true either, because it's a new way to create a new field of artistic expression. And I'm pretty sure that AI, as we know it today, it's going to be a new artistic field in the future, as photography took 70 years to be a proper artistic means of expression. I think it's going to take maybe seven years for AI to do it.

And actually, I love the parallel you're making with photography, because you're making the point that the new technology is becoming an art in itself and increase the depth. But it makes me think about something I heard from H. N. Klein, the philosopher in France, that was making the parallel between when photography was introduced, of course, it expanded the field and the creativity using photography, but it also kind of forced traditional painting that was figurative to reinvent itself in abstraction. And I think this is maybe something that AI is going to force us to a certain extent to reinvent ourselves and maybe reinvent our intelligence, reinvent our creativity in light of what can be done here. So I really love the parallel.

Yeah, I'm pretty sure that Monet or all the impressionist field will not have existed without photography. It's because of photography that they needed to change the way they were painting. And nowadays, digital artists that are not using AI, some of them are refusing to use AI. Well, they have to position themselves in a certain way and reinvent their craft. So it's a great way also for traditional artists to reinvent their craft. Using AI or not, it's a way of positioning yourself and it's moving a whole field in this sense.

Sorry, I was also going to talk about Mario Kingerman, which is one of the AI pioneer artists. And he said that we've generated AI, images are not enough for artists, which means that they have to work on something else. They have to find new ways to express themselves. And also, there are a lot of artists which are working with or without AI, but also highlighting the impact of artificial intelligence. And I think this is super interesting because there is a lot of things that are intangible that you can't see. And the artists are here to be able to highlight things we can't see with AI and to help everybody understand what is this technology about. And it's also something we were talking about using AI in museums, but there are also museums doing exhibitions about AI. And I think this is super interesting because AI, I think in the coming years, we are going to study it at school, but what happens after school? And I think museums are great places to learn about artificial intelligence.

Do you think, on that note, that visitors can maybe be like triggered by AI to become creators more easily? Because, you know, if you're giving to someone like a pencil or brush, it can be maybe a bit daunting at first to try to be creative. Do you think that AI can help them be more creative or can nudge them in the right direction?

Yeah, well, I've seen a very interesting exhibition in Helsinki in Finland, where it was designed for children. And the children will just have to draw a picture and then AI will, based on their picture, will try to redraw something in the style of a specific artist. And I think that was great because it was a great way to learn to the children how AI can work with them and how they can co-live with this form of technology and expand their creativity. And I'm pretty sure that what Marion said is very important. In this field, museums have a fantastic role to play because they are an actor of trust. We are trusting museums to deliver some knowledge. And in a field where we have distrust about politicians or media or all those big spheres, well, museum is still a safe haven for trust. And the question of AI is also a question of trust, how you can trust this technology and how you can transmit the way this technology is working and how you can make understand people how this technology is working. So if you are empowering people with AI inside the museum, it's a great way for them to make them understand how this technology is working and how they can use it in their daily life to be more creative.

Yeah, and I think this is something that is mostly overlooked, the role of cultural institutions for AI literacy, so I really like that.

I know we're almost on time, so maybe to close off, why don't you maybe give a few recommendations of things that you've seen made with AI, or tools to experiment with AI to our audience, things that have truly inspired you, or projects, or things that you've been really, really moved by recently?

It's tricky, because there are so much things that I would recommend. Perhaps I can talk about an event that happened a few months ago at the Tate in London, where the artist Stephanie Dinkins was there, and I love her approach, because one of the things she did is to host dinners with people talking about artificial intelligence, and the way they experience the technology, and the impact it has on them, and I think it was great to have this kind of citizen approach of talking about the technology, and this was an interesting project I want to highlight, but just like a top of mind project, and there are so many of them, I don't know, Valentin, if you want to drop one.

Yeah, well, I want to talk about a project that has not happened yet, but it's going to be happening in May, and it's something that we are building currently with the OpenAI team. We partner with OpenAI and the Chateau de Versailles to create the first purely audio experience in order for you to talk to a sculpture in the garden of the Chateau de Versailles. So in front of each sculpture, you can ask them directly some questions and have answers, and it's a way for the visitors to make those sculptures live and to have conversation with them, and so it's a project that we are currently building with the OpenAI team and the Chateau de Versailles, and it's going to be live in May, so if you are passing by France and if you want to see something in the Chateau de Versailles, please try this experiment, and it's going to be live at the end of the month of May.

Yeah, that's a very good opening, and I think it's really, really an exciting project also because it's taking advantage of the new multimodality, and as you mentioned, I think, in the introduction, the pace of development with the new dimension that multimodality opens is truly exciting for the field of art and the intersection with the technology. So yeah, thanks a lot, Mario and Valentin, and maybe to close it off, I will invite Ben back to wrap up the conversation.

Wow. What to say? That was amazing, and maybe just to repeat something that I saw in the chat, c'est incroyable, so amazing talk, and I'm really excited for the Q&A session.

But before we move there, it would be helpful to give a few closing remarks to those listening in in the live stream and for other folks to know what we have upcoming on the calendar.

So we have three events I just wanted to make you all aware of. We have an event on how AI is changing European higher education next week on March 27th, and on that same day, on March 27th, we have an exciting research talk with two of my amazing colleagues who will be leading contributors on OpenAI's deep research, so they will be giving a presentation on all things deep research and that product.

Later on April 3rd, we're having an event with SOAR artists, so if many of you in the audience are interested in art, interested in SOAR, so if you want to hear more about how folks are using SOAR, focusing on the next chapter in AI storytelling, definitely check that out on April 3rd.

These are not events, but just three things for you all to be aware of. We had the Paris office launch a few weeks ago, so the Paris office launch, if you haven't checked it out, you'll hear from some incredible leaders at OpenAI, some of the people who you may recognize who are on the event today, and also, given that many of you are in France or in Europe, we've created a Europe local interest group for you to join on the OpenAI forum, so you can connect with other forum members in the forum network and exchange contacts, and then also for the creatives here, you can also join the creative and cultural production interest group, so make sure you check those links out.

That was a lot, so you have a lot of material to check out, and we hope to see you back on our next event.