Epoch 4 – Outlook: What’s Next?

Welcome to the final installment of our blog series on the history of generative artificial intelligence! So far, we have explored the journey from the earliest statistical models to neural networks and modern applications. But what does the future hold for us? In this concluding part, we will examine the upcoming challenges and opportunities for generative AI.

Interpolating vs. Extrapolating

A central aspect in the advancement of GenAI is the transition from interpolation to extrapolation. While today’s models like GPT-4o and DALL-E 3 deliver impressive performances within the trained data space (interpolation), the capability for extrapolation—creating content beyond the learned scope—is still in its infancy. The next generation of models might aim to surpass this boundary, generating even more creative and versatile content. Whether and how this will happen is currently a hotly debated topic. As of now, there are no concrete concepts on what this new generation of extrapolating models might look like.

Agents

Another exciting area is AI agents. These intelligent systems can operate autonomously, make decisions, and execute tasks without human intervention. This capability sets them apart from ChatGPT and other chatbots, which can “only” provide useful answers to queries. Such agents could, in the future, take on complex tasks in various fields such as medicine, finance, or customer service, far exceeding today’s capabilities.

Ethical and Legal Questions

The growing prevalence of GenAI also brings ethical and legal challenges. Addressing bias—prejudiced or discriminatory outcomes—remains a critical issue. Moreover, ethical standards and legal frameworks for the use of third-party GenAI and proprietary models must be developed to minimize misuse and negative impacts. Currently, intellectual property rights are a focal point. The verdicts in the legal battles between Stability AI and Getty Images, OpenAI and the New York Times, as well as Universal, Sony, and Warner against Suno and Udo, are eagerly anticipated.

From Model to System

A significant development is the shift from individual models to integrated systems. What does this mean in practice? Generative AI will be embedded into complex systems that close security gaps and enhance the reliability of applications. An example of this is that ChatGPT does not directly execute terminal commands but serves a custom API with predefined behavior. This integration allows the benefits of GenAI to be harnessed while minimizing potential risks.

Outlook and Conclusion

The future of generative artificial intelligence is both promising and challenging. The capability for extrapolation, the development of autonomous agents, and the integration of models into secure systems are just some of the exciting developments that await us. At the same time, we must continually address ethical and legal questions to ensure the responsible use of these powerful technologies.

Overall, the history of generative AI shows how far we have come—from the first statistical models to highly advanced, multimodal systems. But the journey is far from over. The next few years promise further significant leaps. It is up to all of us to translate technological advancements into societal progress.

 

This was the fourth and final part of our series on the history and future of generative artificial intelligence. We hope you enjoyed reading it as much as we enjoyed writing it. If you want to learn more about AI, you can find many more blog posts, whitepapers, and interviews on our website. Tarik Ashry, Max Hilsdorf

Welcome back to our blog series on the history of generative Artificial Intelligence! In the last part, we examined the transition from traditional statistical models to neural networks and the first major breakthroughs in AI. In this installment, we focus on the current developments and practical applications that have brought generative AI into the hands of the general population.

Epoch 3 – Transition

Period: November 2022 – Present

Period	Paradigms	Techniques	User Profile	Examples
Nov 22 – present	Plug & Play, text-to-anything, Multimodality, Open-Source Hype	RLHF, APIs, PEFT, RAG	General public using chat interfaces, IT experts using APIs and open-source models	Text: ChatGPT, Bard, Mistral; Image: Stable Diffusion, DALL-E, Midjourney; Video: Runway ML, Pika Labs; Audio: Voicebox, MusicGen, Suno

The Breakthrough of Language Models

Although language models like GPT-3 could already write convincing texts and, with the right prompt, even retrieve information, they were initially not very user-friendly. Besides the technical hurdle that an interface to a language model (API) required programming knowledge, these models were not yet capable of natural conversations.

A significant advancement came in January 2022, when OpenAI fine-tuned GPT-3 to follow instructions rather than merely complete sentences. The result, InstructGPT, can be seen as a clear precursor to the breakthrough of ChatGPT in December 2022.

Not only could ChatGPT engage in natural conversations with up to 3,000 words—it also emerged as a promising assistant for various everyday tasks. Packaged in an accessible web application, the release of ChatGPT marked a watershed moment in AI and technology history. Instead of leaving automation to IT experts, office tasks such as writing emails or summarizing texts could now be partially automated by average users as needed. It is no wonder that Andrej Karpathy, a founding member of OpenAI and former AI director at Tesla, tweeted:

Multimodal Generative AI

But to think of modern generative AI only in terms of text would be to overlook the impressive development of multimodal GenAI models. Since April 2022, DALL-E 2 has been able to generate realistic drawings, artworks, and photographs based on short text prompts. Commercial GenAI platforms like RunwayML have, since February 2023, even enabled the animation of images or the creation of complete videos based solely on text prompts. It is thus not surprising that creating music or sound effects with AI has become quick and accessible to everyone. Early models like Google’s MusicLM (January 2023) or Meta’s AudioGen (August 2023) did not yet deliver studio-quality sounds but already demonstrated the technology’s potential. The major breakthrough in GenAI for audio came in the spring of 2024, when Suno, Udio, and Elevenlabs generated high-quality songs and sounds, sparking a significant debate on copyright and fair use.

Who Benefits?

With all these powerful AI models, the question arises: who benefits from these new technologies? Is it once again only the large tech companies, particularly those with a poor reputation regarding data privacy? The answer is: partly, yes. While major breakthroughs are still often led by the likes of Microsoft, Google, and others, smaller, freely available models—known as open-source models—are increasingly achieving significant successes. The language model from French startup Mistral AI recently outperformed OpenAI’s GPT-3.5 in standard test metrics—and did so with a much more resource-efficient and faster model than its Silicon Valley competitor. With Meta being one of the leading open-source developers, particularly with their Llama models, even one of the world’s largest tech companies is contributing to this trend. Those who wish for generally available, private AI assistants can look forward to a bright future.

Challenges and Opportunities

The third phase in the history of GenAI is characterized by the widespread availability of high-performance AI models, either through commercial web applications and platforms or freely available open-source models. Companies are increasingly realizing that the value creation through AI is not solely tied to the availability of highly qualified IT experts. Rather, it is essential to maximize the benefits created by AI through the broad application of existing technologies. Nonetheless, numerous challenges remain, including the security of input and output data or the fairness of AI decisions.

How Do We Move to the Next Level?

A key to the success of this epoch is the paradigm of “Plug & Play”. This means that models like ChatGPT and DALL-E 2 can be used easily and without deep technical knowledge. These models are easily accessible through “Reinforcement Learning from Human Feedback” (RLHF) and API interfaces. Andrej Karpathy’s statement that “the hottest new programming language is English” underscores the democratization of AI usage.

Another crucial aspect is the fine-tuning of models to human preferences, which has significantly improved user-friendliness and applicability. At the same time, open-source models are experiencing a boom, as they can run on regular computers, making them accessible to a broader user base.

Ethical and legal questions are also in focus, especially in dealing with third-party GenAI and proprietary models. Issues such as bias and fairness are not to be underestimated, as they significantly influence the acceptance and integrity of AI applications.

What’s Next?

In the next part of our series, we take a look into the future of generative AI. Don’t miss out as we explore the upcoming challenges and opportunities in the world of generative artificial intelligence.

Don’t miss Part 4 of our blog series. Tarik Ashry

Welcome to our four-part blog series on the history of Generative Artificial Intelligence. Our journey through history will highlight the significant milestones and show how the entire concept of generative AI has fundamentally transformed with each step of development. From the early attempts of sketching probability distributions with pen and paper to today’s sophisticated algorithms that generate complex and creative content – each of the four steps marks a revolution, not just an update.

Why is the history of Generative AI so exciting? Because it demonstrates how each technological advancement has not only changed the methods but also the assumptions, usage, audience, and interaction with the models. What began as a tool for statistical analysis is today a creative partner capable of producing art, music, text, and much more.

Join us on this journey through the history of GenAI.

Epoch 1 – Foundations

A well-kept secret: If you rearrange the letters of “Data Science”, you get “Statistics”. Just kidding. But it is true that the roots of data science go back to the 18th century. Back then, α, Θ, and other mathematical symbols had more of a mothball charm than venture capital appeal.

Mathematicians like Gauss, Bayes, and a number of clever Frenchmen recognized the value of counting early on. They counted, recounted, and compared the results – all by hand and very laboriously. Yet these methods are still relevant and proven today – a true evergreen!

With the invention and availability of electricity, a new era began. Data could now be processed and analyzed much more efficiently. The idea of an “electronic marble run” for data emerged – a system with switches and paths that triggered various actions based on data input, such as lighting a bulb or executing a function.
An early, actually functional form of Artificial Intelligence (AI) was born: algorithms based on observations and derived rules.

Period	Paradigms	Techniques	User Profile	Examples
1700-1960	Pen, soldering iron, punch card	Counting, sorting, making assumptions	Engineers, manufacturers, researchers	Accounting, assembly lines, natural sciences
1960-2010	Programming application-specific code	The same as before, but automated	Statisticians, computer scientists, early data scientists, and machine learning researchers	Spam filters, (sentiment) text analysis, Optical Character Recognition (OCR)

What Makes These Early Models Generative? Well, the “electronic marble run” could also be operated in reverse. Forward, it was a statistical model that assigned a category or value to an observation. For this, the model needed to have an idea of the data. Backward, however, high-probability examples of mushrooms, marbles, data – in other words, images or tables – could be generated through random draws. The generative capabilities of the models were often underestimated, as the forward function was the focus.

This methodology is called the Naïve Bayesian Classifier. “Naïve” here is not meant derogatorily but refers to simplifying assumptions that make modeling significantly easier. With naïve methods, one does not have to assume complex relationships between variables like mycelium, stem, and cap of a mushroom. One simply says: If the average quality of all three parts is good enough, then the mushroom is good.
Some of the first applications of these models included handwriting recognition (for example, at the post office, known as Optical Character Recognition or OCR), as well as spam filters and general text analyses up to this day.

This was the first insight into the foundations of generative Artificial Intelligence. In the next part of our series, we will dive into the world of neural networks and machine learning, which laid the foundation for modern AI systems. Stay curious and don’t miss the next milestone in the history of generative AI!

Don’t miss Part 2 of our blog series. Tarik Ashry

Welcome back to our series on the history of generative Artificial Intelligence. In the first part, we explored the basics and saw how early statistical models like the Naïve Bayesian Classifier paved the way for today’s AI. Now, we take a significant leap forward and delve into the second epoch—a transitional period where neural networks and GPUs take center stage and revolutionize the world of AI.

Epoch 2 – Transition

From 2015 – Awe in the Pre-Stage

The AI winter is over, and neural networks as well as GPUs (Graphics Processing Units) have made their grand entrance. However, these new technological marvels are largely confined to tech experts. But that doesn’t mean impressive products and applications aren’t being developed—quite the opposite! StyleGANs (Generative Adversarial Networks) deliver unprecedented image quality, and transformer models like BERT (Bidirectional Encoder Representations from Transformers) capture text with remarkable detail.

Despite this, direct operation of these models remains out of reach for the general public due to their technical and specific nature. One must choose, expand, link, and train specific models and architectures. Nevertheless, applications such as chatbots, customer service automation, generative design, and AutoML solutions make their way to the market.

Period	Paradigms	Techniques	User Profile	Examples
2015-2019	Latent Spaces, Embeddings	Masked Language Models, GANs	Programmers, Data Scientists	BERT, StyleGAN
2019-2022	Text Prompts	Few Shot, Prompt Engineering	Programmers (API), End Users	GPT-3

From 2019 – Lift-off

“Bigger is better” becomes the new credo. Open source is sidelined, and the world of Natural Language Processing (NLP) is turned on its head: Large Language Models (LLMs) are here! However, the first model, GPT-2, is not released in 2019 due to concerns over potential misuse:

“The Elon Musk-backed nonprofit company OpenAI declines to release research publicly for fear of misuse.” (Guardian, 14.02.2019)

The words “Musk,” “nonprofit,” and “fear of misuse” in one sentence – almost surreal in hindsight. By the end of the year, GPT-2 is eventually released. It finds significant use in research to explore the fundamental properties of LLMs and later serves to understand the implications of advancements by comparing it to larger models.

In 2020, GPT-3 follows—with ten times more data and a model a hundred times larger. In 2021, DALL-E is introduced, followed by DALL-E 2 in 2022. Texts can now be processed and created using natural (written) language, although not yet in the now-familiar dialogue form, but through Few-Shot Prompts. For images, however, this was not the case; in DALL-E and DALL-E 2, example images could not be provided in the prompt. In this paradigm, which is still common in non-chat variants of GPTs today, the model was not trained to conduct a conversation but merely to complete texts. This means it requires examples, such as question-answer pairs, to understand how to continue the text.

An example of a Few-Shot Prompt: After three provided examples, the actual user input follows up to the word “Label:”, expecting the model to grasp the task or meaning and continue the text by giving the correct answer.

The public, as well as developers, are impressively confronted with state-of-the-art technology, for example, through the first articles written with GPT-3.

In the next part of our series, we will look at the latest developments and the revolution of generative Artificial Intelligence. Read on to discover how we transition from Few-Shot Prompts to practical applications, making generative AI accessible to the general public!

Don’t miss Part 3 of our blog series.  Tarik Ashry Tarik Ashry, Max Hilsdorf

AI chatbots are quickly becoming essential in businesses, but not all chatbots are created equal. Some can be set up swiftly with ease, yet they often lack the customizability and advanced features that could truly elevate performance, like enhancing customer service responses. Tailor-made solutions that do offer those capabilities can become complex and costly, particularly when they include sophisticated, use-case-specific technologies like Retrieval-Augmented Generation (RAG) that enhance the precision and reliability of generative AI models, allowing them to converse using a company’s own databases and produce verified facts.

 

statworx’s CustomGPT stands out by combining the best of both worlds: high customizability with quick implementation. This tailor-made solution offers secure and efficient use of ChatGPT-like models, with interfaces that can be designed in a company’s brand style and easily integrated into existing business applications like CRM systems and support tools.

So, what’s crucial when companies seek the ideal chatbot solution?

Requirement Analysis: First, a company’s specific needs should be pinpointed to ensure the chatbot is perfectly tailored. What tasks should it handle? Which departments should it support? What functionalities are necessary? 

Model Training: A custom GPT-chatbot needs to be equipped with relevant data and information to assure high accuracy and responsiveness. If the necessary data isn’t available, the technical effort might not be justifiable. 

System Integration: Seamless integration of the chatbot into existing communication channels like websites, apps, or social media is crucial for effective use. Different solutions may suit different infrastructures. 

 

Ready to deploy and adaptable

The CustomGPT-chatbot from statworx is notable for its quick setup, often within weeks, thanks to a mix of proven standard solutions and custom adjustments. It allows file uploads and the ability to chat, extracting secure information from the company’s own data. With advanced features like fact-checking, data filtering, and user feedback integration, it stands apart from other systems. 

Moreover, CustomGPT gives companies the freedom to choose their chatbot’s vocabulary, communication style, and overall tone, enhancing brand experience and recognition through personalized, unique interactions. It’s also optimized for mobile displays on smartphones. 

 

Technical Implementation

On the technical front, Python is the core language for CustomGPT’s backend, with statworx developers utilizing FastAPI, a modern web framework that supports both Websockets for stateful communication and a REST API for services. CustomGPT is versatile, suitable for various infrastructures, from a simple cloud function to a machine cluster if needed. 

A key feature of its architecture is the connection to a data layer, providing a flexible backend that can quickly adapt to changing conditions and requirements. The frontend application, built with React, seamlessly interacts with the backend, which, for example, leverages the powerful Azure AI search function. This configuration allows for the implementation of custom search solutions and efficient fulfillment of specific requirements. 

 

The benefits at a glance:

Data Protection and Security
CustomGPT ensures all data is stored and processed within the European Union, with full control retained by the company, setting it apart from other GPT-based solutions. 

Integration and Flexibility
Its flexible integration into existing business applications is supported by modularity and vendor independence, allowing CustomGPT to adapt to various infrastructures and models, including open-source options. 

Features and Customization
CustomGPT’s customization includes integration with organizational data, user role adaptation, and the use of analytics to enhance conversations, offering flexibility and personalization for corporate applications. 

Personalized Customer Experience
By tailoring to a company’s specific needs, Custom GPT-chatbots can provide personalized and effective customer interactions. 

Efficient Customer Support
CustomGPT chatbots can answer questions, resolve issues, and provide information around the clock, increasing customer satisfaction and efficiency.  

Scalability
Companies can effortlessly scale their customer support capacity with GPT-chatbots to maintain consistent service quality, even during high demand. 

 

The time for your own chatbot is now. Therefore, statworx focuses on upstream development with quick deployment and easy implementation. This means all users of a statworx CustomGPT benefit from patches, bug fixes, and new features over time. CustomGPT remains versatile and flexible, meeting the specific, changing needs of companies and addressing complex requirements. Contact us now for a consultation.

  Tarik Ashry

Imagine it’s Friday and instead of the usual office routine, you find yourself in the middle of a hackathon for a pro bono project, planning a trip to the Dialogue Museum at Frankfurt’s Hauptwache, or asking passers-by on the Zeil about their thoughts on AI and sustainability. What do all these diverse activities have in common? They are part of the 4:1 week (“Four to one”) at statworx – a work model that provides space for personal development, innovative work, and social engagement.

Why did we introduce the 4:1 week?

At statworx, we faced a challenge: Our employees had great ambitions for further education but found little time for it. The workweek was filled with projects and appointments, and personal development often took a back seat.

To solve this dilemma, we introduced an innovative work structure in 2022: the 4:1 week. The principle is simple and effective. Four days of the week are dedicated to intensive work on projects. The fifth day, Friday, is all about further education. On this day, business goes on at a reduced pace, and employees can fully dedicate themselves to their professional and personal development.

This new structure has led to a noticeable change. It allows every individual at statworx to pursue further education and learn new skills, even during busy times. This not only promotes personal development but also the company’s innovative strength. While customer projects receive full attention from Monday to Thursday, Friday becomes a space for learning and inspiration. This way, we collectively maintain a balance between customer needs and employee development.

What training opportunities does statworx offer?

statX: Once a month, employees voluntarily meet to exchange insights from projects, personal knowledge, or new models and approaches. Topics range from deep learning with audio data to the AI Act and anomaly detection.

Clusters: The self-organized workgroups are the breeding ground for the development and promotion of expertise at statworx. Currently, there are fifteen clusters where employees can delve into topics that are particularly close to their hearts – thereby strengthening statworx’s overall innovation power. Here are three examples from our cluster portfolio:

• Bio Medicine Cluster: Development of AI applications in the biomedical and pharmaceutical fields.
• NLP Cluster: Implementation of state-of-the-art models, best practices, and software for NLP, as well as multimodal applications of NLP models.
• Explainable AI Cluster: Engagement with methods to make black box AI models transparent and explainable.

Training Budget: Each team member at statworx has an annual training budget. This budget can be used for individual training measures such as online courses, external trainings, participation in conferences, certifications, and much more.

Technical and Non-Technical Trainings: Throughout the year, we offer a variety of trainings that develop both soft and hard skills. From effective communication and constructive feedback to software engineering and Scrum, numerous exciting topics are covered.

How do we know that our 4:1 concept works?

Through regular pulse surveys and our biannual employee satisfaction surveys, we collect anonymized feedback on the 4:1 model. This gives us insights into our team members’ personal experiences. Additional direct conversations with employees provide us with further valuable impressions to decide how we will shape and develop the program in the future.

Our interim conclusion is that with every certificate earned and each personal success story, the 4:1 week at statworx proves its effectiveness. We look forward to continuously refining the concept and supporting even more employees on their individual learning journey.

Colleagues confirm our preliminary conclusion:

“The 4:1 week gives me the opportunity to delve deeper into topics and areas of knowledge that I’m passionate about, alongside my project work. That’s why I, along with my colleagues, initiated the Bio-Medicine Cluster. And I contribute my software engineering skills to the Cluster for Technical Delivery to continuously improve the technical deployment of our solutions.” – Benedikt Batton, Consultant Data Science, AI Development

“The 4:1 week promotes individual professional development and provides space to create value in innovative ways beyond existing departments and hierarchies. In my time at statworx, work on Fridays has always been a source of motivation, inspiration, and self-fulfillment.” – Max Hilsdorf, Consultant, AI Academy

Would you like to be part of a work culture that makes space for personal development and a balanced ratio between project work and further education? Then visit our career page and apply now! Ida-Marie Trieba

Have you ever imagined a restaurant where AI powers everything? From the menu to the cocktails, hosting, music, and art? No? Ok, then, please click here.

If yes, well, it’s not a dream anymore. We made it happen: Welcome to “the byte” – Germany’s (maybe the world’s first) AI-powered Pop-up Restaurant!

As someone who has worked in data and AI consulting for over ten years, building statworx and the AI Hub Frankfurt, I have always thought of exploring the possibilities of AI outside of typical business applications. Why? Because AI will impact every aspect of our society, not just the economy. AI will be everywhere – in school, arts & music, design, and culture. Everywhere. Exploring these directions of AI’s impact led me to meet Jonathan Speier and James Ardinast from S-O-U-P, two like-minded founders from Frankfurt, who are rethinking how technology will shape cities and societies.

S-O-U-P is their initiative that operates at the intersection of culture, urbanity, and lifestyle. With their yearly “S-O-U-P Urban Festival” they connect creatives, businesses, gastronomy, and lifestyle people from Frankfurt and beyond.

When Jonathan and I started discussing AI and its impact on society and culture, we quickly came up with the idea of an AI-generated menu for a restaurant. Luckily, James, Jonathan’s S-O-U-P co-founder, is a successful gastro entrepreneur from Frankfurt. Now the pieces came together. After another meeting with James in one of his restaurants (and some drinks), we committed to launching Germany’s first AI-powered Pop-up Restaurant: the byte!

the byte: Our concept

We envisioned the byte to be an immersive experience, including AI in as many elements of the experience as possible. Everything, from the menu to the cocktails, music, branding, and art on the wall: everything was AI-generated. Bringing AI into all of these components also pushed me far beyond of what I typically do, namely helping large companies with their data & AI challenges.

Branding

Before creating the menu, we developed the visual identity of our project. We decided on a “lo-fi” appeal, using a pixelated font in combination with AI-generated visuals of plates and dishes. Our key visual, a neon-lit white plate, was created using DALL-E 2 and was found across all of our marketing materials:

Location

We hosted the byte in one of Frankfurt’s coolest restaurant event locations: Stanley, a restaurant location that features approx. 60 seats and a fully-fledged bar inside the restaurant (ideal for our AI-generated cocktails). The atmosphere is rather dark and cozy, with dark marble walls, highlighted with white carpets on the table, and a big red window that lets you see the kitchen from outside.

The menu

The heart of our concept was a 5-course menu that we designed to elevate the classical Frankfurter cuisine with the multicultural and diverse influences of Frankfurt (for everyone, who knows the Frankfurter kitchen, I am sure you know that this was not an easy task).

Using GPT-4 and some prompt engineering magic, we generated several menu candidates that were test-cooked by the experienced Stanley kitchen crew (thank you, guys for this great work!) and then assembled into a final menu. Below, you can find our prompt to create the menu candidates:

“Create a 5-course menu that elevates the classical Frankfurter kitchen. The menu must be a fusion of classical Frankfurter cuisine combined with the multicultural influences of Frankfurt. Describe each course, its ingredients as well as a detailed description of each dish’s presentation.”

Surprisingly, only minor adjustments were necessary to the recipes, even though some AI creations were extremely adventurous! This was our final menu:

• Handkäs’ Mousse with Pickled Beetroot on Roasted Sourdough Bread
• Next Level Green Sauce (with Cilantro and Mint) topped with a Fried Panko Egg
• Cream Soup from White Asparagus with Coconut Milk and Fried Curry Fish
• Currywurst (Beef & Vegan) by Best Worscht in Town with Carrot-Ginger-Mash and Pine Nuts
• Frankfurt Cheesecake with Äppler Jelly, Apple Foam and Oat-Pecanut-Crumble

My favorite was the “Next Level” Green Sauce, an oriental twist of the classical 7-herb Frankfurter Green Sauce topped with a fried panko egg. Yummy! Below you can see the menu out in the wild 🍲

AI Cocktails

Alongside the menu, we also prompted GPT to create recipes that twisted famous cocktail classics to match our Frankfurt fusion theme. The results:

• Frankfurt Spritz (Frankfurter Äbbelwoi, Mint, Sparkling Water)
• Frankfurt Mule (Variation of a Moscow Mule with Calvados)
• The Main (Variation of a Swimming Pool Cocktail)

My favorite was the Frankfurt Spritz, as it was fresh, herbal, and delicate (see pic below):

AI Host: Ambrosia the Culinary AI

An important part of our concept was “Ambrosia”, an AI-generated host that guided the guests around the evening, explaining the concept and how the menu was created. We thought it was important to manifest the AI as something the guests can experience. We hired a professional screenwriter for the script and used murf.ai to create several text-2-speech assets that were played at the beginning of the dinner and in-between courses.

Note: Ambrosia starts talking at 0:15.

AI Music

Music plays an important role for the vibe of an event. We decided to use mubert, a generative AI start-up that allowed us to create and stream AI music in different genres, such as “Minimal House” for a progressive vibe throughout the evening. After the main course, a DJ took over and accompanied our guests into the night 💃🍸

AI Art

Throughout the restaurant, we placed AI-generated art pieces by the local AI artist Vladimir Alexeev (a.k.a. “Merzmensch”), here are some examples:

AI Playground

As an interactive element for the guests, we created a small web app that takes the first name of a person and transforms it into a dish, including a reasoning why that name perfectly matches the dish 🙂 You can try it out here: Playground

Launch

The byte was officially announced at the S-O-U-P festival press conference in early May 2023. We also launched additional marketing activities through social media and our friends and family networks. As a result, the byte was fully booked for three days straight, and we got broad media coverage in various gastronomy magazines and the daily press. The guests were (mostly) amazed by our AI creations, and we received inquiries from other European restaurants and companies interested in exclusively booking the byte as an experience for their employees 🤩 Nailed it!

Closing and Next Steps

Creating the byte together with Jonathan and James was an outstanding experience. It further encouraged me that AI will transform not only our economy but all aspects of our daily lives. There is massive potential at the intersection of creativity, culture, and AI that is currently only being tapped.

We definitely want to continue the byte in Frankfurt and other cities in Germany and Europe. Moreover, James, Jonathan, and I are already thinking of new ways to bring AI into culture and society. Stay tuned! 😏

The byte was not just a restaurant; it was an immersive experience. We wanted to create something that had never been done before and did it – in just eight weeks. And that’s the inspiration I want to leave you with today:

Trying new things that move you out of your comfort zone is the ultimate source of growth. You never know what you’re capable of until you try. So, go out there and try something new, like building an AI-powered pop-up restaurant. Who knows, you might surprise yourself. Bon apétit!

Impressions

Media

FAZ: https://www.faz.net/aktuell/rhein-main/pop-up-resturant-the-byte-wenn-chatgpt-das-menue-schreibt-18906154.html

Genuss Magazin: https://www.genussmagazin-frankfurt.de/gastro_news/Kuechengefluester-26/Interview-James-Ardinast-KI-ist-die-Zukunft-40784.html

Frankfurt Tipp: https://www.frankfurt-tipp.de/ffm-aktuell/s/ugc/deutschlands-erstes-ai-restaurant-the-byte-in-frankfurt.html

Foodservice: https://www.food-service.de/maerkte/news/the-byte-erstes-ki-restaurant-vor-dem-start-55899?crefresh=1 Sebastian Heinz

statworx at Big Data & AI World

From media to politics, and from large corporations to small businesses, artificial intelligence has finally gained mainstream recognition in 2023. As AI specialists, we were delighted to represent statworx at one of the largest AI expos in the DACH region, “Big Data & AI World,” held in our hometown of Frankfurt. This event centered around the themes of Big Data and Artificial Intelligence, making it an ideal environment for our team of AI experts. However, our purpose went beyond mere exploration and networking. Visitors had the opportunity to engage in an enthralling Pac-Man game with a unique twist at our booth. In this post, we aim to provide you with a comprehensive overview of this exhilarating expo.

Fig. 1: our exhibition stand

Tangible AI Experience

Our Pac-Man challenge, where we provided booth visitors with an up-close encounter of the captivating world of artificial intelligence, emerged as a clear crowd favorite. Through our arcade machine, attendees not only immersed themselves in the timeless retro game but also witnessed the remarkable capabilities of modern technology. Leveraging AI, we analyzed players’ real-time facial expressions to discern their emotions. This fusion of cutting-edge technology and an interactive gaming experience was met with exceptional enthusiasm.

Our AI solution for emotion analysis of players ran seamlessly on a powerful M1-chip-equipped MacBook, enabling real-time image processing and fluid graphics display. The facial recognition of the players was made possible by a smart algorithm that instantly detected all the faces in the video. Subsequently, the face closest to the camera was selected and focused on, ensuring precise analysis even amidst long queues. Further processing involved a Convolutional Neural Network (CNN), specifically the ResNet18 model, which accurately detected players’ emotions.

Functioning as a multimedia server, our backend processed the webcam stream, facial recognition algorithms, and emotion detection. It could be operated either on-site using a MacBook or remotely in the cloud. Thanks to this versatility, we developed an appealing frontend to vividly present the real-time analysis results. Additionally, after each game, the results were sent to the players via email by linking the model with our CRM system. For the email, we created a digital postcard that provides not only screenshots of the most intense emotions but also a comprehensive evaluation.

Fig. 2: Visitor at Pac-Man game machine

Artificial Intelligence – Real Emotions

Our Pac-Man challenge sparked excitement among expo visitors. Alongside the unique gaming experience on our retro arcade machine, participants gained insights into their own emotional states during gameplay. They were able to meticulously observe the prevailing emotions at different points in the game. Often, a slight surge of anger or sadness could be measured when Pac-Man met an untimely digital demise.

However, players exhibited varying reactions to the game. While some seemed to experience a rollercoaster of emotions, others maintained an unwavering poker face that even the AI could only elicit a neutral expression from. This led to intriguing conversations about how the measured emotions corresponded with the players’ experiences. It was evident, without the need for AI, that visitors left our booth with positive emotions, driven in part by the prospect of winning the original NES console we raffled among all participants.

Fig. 3: digital post card

The AI Community on the Move

The “Big Data & AI World” served not only as a valuable experience for our company but also as a reflection of the burgeoning growth in the AI industry. The expo offered a platform for professionals, innovators, and enthusiasts to exchange ideas and collectively shape the future of artificial intelligence.

The energy and enthusiasm emanating from the diverse companies and startups were palpable throughout the aisles and exhibition areas. Witnessing the application of AI technologies across various fields, including medicine, logistics, automotive, and entertainment, was truly inspiring. At statworx, we have already accumulated extensive project experience in these domains, fostering engaging discussions with fellow exhibitors.

Our Conclusion

Participating in the “Big Data & AI World” was a major success for us. The Pac-Man Challenge with emotion analysis attracted numerous visitors and brought joy to all participants. It was evident that it wasn’t just AI itself but particularly its integration into a stimulating gaming experience that left a lasting impression on many.

Overall, the expo was not only an opportunity to showcase our AI solutions but also a meeting point for the entire AI community. The sense of growth and energy in the industry was palpable. The exchange of ideas, discussions about challenges, and the establishment of new connections were inspiring and promising for the future of the German AI industry.
Max Hilsdorf

What to expect:

The konaktiva at the Technical University of Darmstadt is one of the oldest and largest student business contact fairs and is organized annually by students. With over 10,000 visitors last year and more than 200 booths, konaktiva is the ideal opportunity for companies, students and graduates to get in touch with each other. 

This year statworx is again represented at konaktiva with a booth and several colleagues. We will be at the fair on one of the three days (May 9th to May 11th) and will announce the exact date here as soon as we receive the information.

We are looking forward to meet interested students and graduates to inform them about different job opportunities – from internships to permanent positions – at statworx as well as to share our experiences from our daily work. However, getting to know each other does not only take place at the booth – it is also possible to get in touch with us in pre-scheduled one-on-one meetings and to clarify individual questions.

Participation is free of charge for visitors.

As anyone who works at statworx knows, a party always comes at the right time for us. After our exhilarating summer party high above the rooftops of Fankfurt in August 2022, we invited everyone to a big Christmas party at the end of the year – and opened our huge and beloved kitchen on the conference floor for the entire team and all partners. And as always, it was a great party. But first things first.

Die Agenda

This year, besides the drinks at the bar, there were also some nice, contemplative items on the program agenda. As part of the micro-event “Decorate the Christmas Tree”, employees had the opportunity to decorate our bare Christmas tree with individual ornaments. From unicorns, rainbows and Christmas painted Paracetamol packets, everything was hanging on the tree in the end, underlining the individuality of our team. We thought it looked really nice.

Special mention must also be made of this year’s award ceremony for our value carriers of the year. The entire team had the opportunity to vote for special carriers of our corporate values. The winners of this voting were then entered into a draw for special prizes and each winner received an individual, very personal address from the management.

statworx set its official corporate values  in 2021 already, after several rounds of voting and discussion among management and the team. Ever since these values represent the maxim of our work:

 

• We run on data
  Data is our fuel. We are united by our passion for AI technology and data-driven innovation. From it we draw the strength and inspiration to take new paths and grow beyond ourselves.
• We thrive together.
  The power lies in the team. Always. Without exception. Mutual trust and the knowledge that we can accomplish more together are at the core of our collaboration. This is the only way we can achieve our ambitious goals.
• We grow through challenge.
  The curiosity and desire to consistently face new challenges and grow through them are deeply enshrined in us. We see opportunities in change and learn from new experiences.
• We embrace individuality.
  We value the uniqueness of every person and always treat each other as equals. Different backgrounds, mindsets, and ideas enrich us and build the foundation for our success.
• We do what matters.
  We focus on what truly counts. In our projects, we work on solutions that create long-term value. We use data and AI to shape the future for the better. For people, economy, society, and the environment.
• We own our game.
  We take responsibility, execute ideas, and think big. If you want to change the status quo you must do so with full conviction. We believe in ourselves, set ambitious goals, and make the future happen.
• We care for the crew.
  We are more than just colleagues – we are a crew. We are our own most valuable asset. We look out for and support each other, and we create a harmonious working environment where everyone feels valued and supported.

Our special Value Carriers this year are:

• An Hoang / Alexander Müller – We run on data.
• Eva Engelhardt – We thrive together. (Winner)
• Stephan Müller – We grow through challenge.
• Markus Berroth – We embrace individuality.
• Jan Fischer – We do what matters.
• Stephan Emmer – We own our game.
• Andreas Vogl – We care for the crew.

Gifts, gifts, gifts

But it is not only through special performance that you can win at statworx. With our big Christmas raffle everyone had a chance to win. Already a few years ago the raffle was a popular addition to the Christmas party and caused big eyes for the winners. Therefore, it was just the right time to stir the lottery drum again and to raffle off gifts with a total value of 2.500 Euro within the team. From cool prizes like the Airpods Pro Max, interior from vitra and artemide, donation vouchers, merchandising kits and the handmade “Trashy Treasure” (a card game with embarrassing and funny photos from the company history). This year’s well-deserved main winner was our colleague Jannik Klauke, who not only proved to have a lucky hand, but also showed a particularly high level of commitment and thus contributed to the success of the company in 2022 – we would like to take this opportunity to say thank you.

And as if there weren’t enough presents, this year we once again celebrated the traditional Christmas Secret Santa, in which the team gave each other funny, useful and useless gifts. Again, we had great fun! Thanks to the organizers.

Finally, the official part was rounded off by the annual Christmas speech of our CEO Sebastian, whose inspiring content was listened to devoutly by everyone present.

Party, food and good vibes!

A proper sound system is part of a kitchen party. That’s why we had a professional sound setup come and a large DJ booth mounted in the kitchen. Finally, around 9pm, our DJane friend Elisa Cielo arrived (she is very well known in Frankfurt), who had already given us a good kick at the summer party. While we had a lineup of 6 DJs at the summer party, this time we made do with Elisa and her DJ colleague Anton, which also turned out to be a stroke of luck. Cool electronic beats were blasting throughout the conference area and kept us awake until half past four in the morning. We were nourished by numerous sweets brought by the team, tons of pizza and a sensational midnight snack by the girls from “Gudrun Kocht”, who specialize in homemade, natural soups and stews here in Frankfurt – just the thing to fight the hangover.

When the last guests had left, a colleague simply stayed right in the office and did – more or less effectively – night watch.

On Saturday and Sunday, our clean-up team finally arrived and by the time this article was written, our kitchen already looked as if nothing had happened.

We had a great party, are now looking forward to a contemplative time with our families and are already curious when, where and how we will celebrate again in 2023.

With this in mind, we wish all our readers a Merry Christmas and a Happy New Year! Cheers!

Julius Heinz