Can GenAI really be useful? |🎙️#46

Promotional image for "DEVOPS ACCENTS EPISODE 46" featuring an illustration of a smiling man with text asking "CAN GENAI REALLY BE USEFUL?" and an invitation to "LISTEN NOW!" against an orange background. Promotional image for "DEVOPS ACCENTS EPISODE 46" featuring an illustration of a smiling man with text asking "CAN GENAI REALLY BE USEFUL?" and an invitation to "LISTEN NOW!" against an orange background.

In this episode, we dive deep into generative AI with Álvaro Sáez , creator of an AI tool Current Anima that transforms text prompts into 3D scenes. Álvaro explains how his technology works and what skills you need to have to build something like this and how difficult it is to build an AI startup in 2024.

  • Alvaro's AI tool that generates 3D scenes from text prompts.
  • The role of Unreal Engine in creating high-quality animations.
  • The oversaturation of the AI market with fake solutions.
  • Competing with AI giants like OpenAI.
  • AI as a tool for enhancing, not replacing, creativity.

You can listen to episode 46 of DevOps Accents on Spotify, or right now:


Becoming an AI engineer

Becoming an AI engineer often starts with curiosity and passion for technology. Álvaro Sáez’s path to AI engineering began in childhood, tinkering with robots made from paper boxes and electronic cars. Despite a non-technical family background, he pursued mechanical engineering before discovering artificial intelligence (AI) through an online program in 2018. From there, he delved into AI concepts, combining automation and business applications before pursuing further studies in machine learning and deep learning. His story illustrates that a strong foundation in problem-solving and curiosity, even outside of traditional computer science fields, can lead to success in AI.

I was a little kid when I realized that I liked a lot to this automation topic. When I was a young kid, I was making robots with paper boxes and electronic cars [...] I studied mechanical engineering [...], later I encountered a course from a university here in Spain, and they had a program about artificial intelligence. It was the first time I heard about AI. It was so shocking. After that, I went for a master's degree in AI, and later, projects in machine learning and deep learning."

Generating 3D scenes from text prompts

One of Álvaro’s most innovative contributions is an AI-driven tool that generates 3D scenes from simple text descriptions. The tool interprets natural language input, such as a man picking up a bottle and throwing it, and transforms this into a 3D animated scene. It combines natural language processing (NLP) with cutting-edge video game technologies like Unreal Engine to create environments, characters, and even their movements. This tool is a game-changer, particularly for industries like gaming, marketing, and cinematography, as it automates the creation of complex scenes, which can typically take days or even months to produce manually.

Obviously, there are real problems. What we do is combine video game technology with Natural Language Processing AI... The NLP algorithm understands the main characters, actions, environments... We then go to Unreal Engine, state of the art for video game quality, and with AI, we generate 3D animations... For example, if a man is taking a bottle and throwing it, the AI generates key points of the character’s pose in real-time.

Unreal Engine: Powering high-quality animations

Unreal Engine plays a crucial role in creating the high-quality animations that Sáez’s tool produces. Initially, his team worked with Unity, a popular game development engine, but the demand for cinematic-level detail required a switch to Unreal. Known for its state-of-the-art graphics capabilities, Unreal Engine provides the perfect foundation for generating professional-quality animations and immersive 3D environments. By programming directly into the engine, Sáez's team is able to seamlessly integrate AI-generated scripts with sophisticated visual assets, producing scenes that rival those of video games from the PlayStation 4 era.

Unreal Engine is like the state of the art in video game developing and in quality for video games... At the beginning, we were working on Unity because it's easier to program, but Unreal offers far better quality. Our product, focused on marketing and cinema, needed the quality Unreal provides. It’s harder to program in Unreal, but the results are worth it.​

Navigating a market flooded with AI hype

Álvaro faces significant challenges in an AI market oversaturated with "fake solutions"—tools that promise much but deliver little. While many companies rush to label their products as "AI-powered" to ride the hype, few truly innovate. For Sáez, this environment presents both obstacles and opportunities. Competing against inflated claims from startups that often lack genuine AI development, he focuses on building real, practical solutions. Tools like his, which are rooted in solid technology and solve tangible problems, are distinguishing themselves in a marketplace filled with noise.

It’s so hard when you're not in the hype... You need to be really good to have visibility... But when a lot of people are using it, the market starts taking it seriously. Before, you were just a freak in your garage doing software... But now, when OpenAI launches something like ChatGPT, people suddenly take it seriously.

The AI startup struggle

As an AI startup, Sáez admits it’s difficult to compete with giants like OpenAI, especially with limited resources and data. His startup lacks the financial and technological backing of major players, making it crucial to leverage creativity and niche applications to stand out. Nonetheless, he emphasizes that the upside to this challenge is the increased attention the AI sector is receiving, making it easier to find an audience. As big names push AI technologies into the spotlight, smaller innovators like Sáez can ride the wave while offering more specialized, high-quality solutions.

When you're a startup, it’s so complicated to compete with big companies. You need a lot of money, data, and resources. When you see companies like OpenAI launching ChatGPT, it’s tough, but it also validates the market. You need to work hard to take advantage of the moment.​

AI as a tool for enhancing creativity

Far from seeing AI as a threat to human creativity, Sáez believes it enhances it. His philosophy is that AI automates repetitive tasks, allowing creators to focus on the more imaginative aspects of their work. For example, his tool streamlines the technical process of generating scenes, but the vision, design, and storytelling remain in the hands of creative professionals. By removing the burden of time-consuming tasks, AI can amplify creativity rather than replace it. This shift allows industries, from marketing to film production, to focus on innovation and artistic expression without being bogged down by logistics.

AI is an enabler. It’s about lowering costs and increasing possibilities... Creativity isn’t in the time you spend on technical tasks like merging shots or recording scenes. It’s in how you conceptualize and structure the project. AI allows you to focus more on creative decision-making, leaving the technical grunt work to the machines.

In conclusion, Álvaro Sáez’s journey into AI, his innovative tools, and his perspective on AI as an enabler, not a replacement, reflect the evolving role of artificial intelligence in enhancing both technological capabilities and human creativity.



Show Notes


Podcast editing: Mila Jones, milajonesproduction@gmail.com

Previous EpisodeAll Episodes