Lessons from Mati Staniszewski

Mati Staniszewski, the co-founder and CEO of the pioneering voice AI company ElevenLabs, has shared numerous insights into the future of technology, entrepreneurship, and innovation.

Quotes

"We're slowly realising that there are in fact very few companies in the audio world at this level, and we want to be that voice that delivers. A good way to think about us: something comparable to OpenAI for audio." [1]
"If a company has growing influence, it has responsibility on how this technology is distributed." [1]
"There should be a big input from regulations and public institutions [to create a level playing field for AI development]." [1]
On the use of ElevenLabs' technology in healthtech and education: "So many use cases are coming up where clearly the solutions are missing, where we can do cool things." [1]
Regarding deepfakes and scams: "We think that the world of deepfakes or product scams is going to grow. As a company, we want to and we do take a lot of responsibility for how we can detect this content and prevent the bad actors from using it." [1]
On competition with OpenAI: "Our largest potential competition, less now, but we think it's on the horizon, is OpenAI... They will probably also start to build more and more models in audio." [1]
"Voice will be the future of interactions. The digital interact with interfaces of the digital world." [2]
"It can carry so much more emotion, so much than text." [2]
On the inspiration for ElevenLabs, which stemmed from poor dubbing in Polish movies: "It's like a horrible experience. and it still happens today. and I was like 'Wow.' Um we think this will. change." [3]
On his relationship with his co-founder, Piotr Dąbkowski: "I'm probably in the luckiest position ever. We met 15 years ago in high school... and now 15 years in, we are still best friends." [4]
On the future of education with voice AI: "I think there will be an entire change where all of us will have the guiding voice, whether we are learning mathematics and are going through the notes, or whether we are trying to learn a new language and interact with a native speaker to guide you through how to pronounce things." [4]
On the Turing test for voice interaction: "I think that curing test for voice interaction of like where you cannot really tell that it's AI or or or and it's as good as the human i think this hopefully will make it happen this year." [5]
On the potential of AI-powered agents: "Everybody in the future will have their agent will interact with agents around and many of those agents will will be your personal agent." [6]
On the Steve Jobs definition of focus: “people think focus means saying yes to the thing you've got to focus on. But that's not what it means at all. It means saying no to the hundred other good ideas that there are”. [7]
"Through stories, experience and advice from founders and investors the one thing that people keep coming back to is how maintaining focus is crucial to building a company." [7]
On the moment he knew ElevenLabs' technology was special: "The moment where it hit was when we released this set of samples where for the first time the AI could somewhat laugh it produced laughter and people picked it up and started telling us that wow this is the first time we heard AI actually laugh." [5]
On the importance of data in AI: "Maybe people are too quick to jump to where do I want to realize efficiencies and instead they should start from where do I have really interesting data that other companies maybe don't have and what are product experience I could build with that." [5]
"If you do care about efficiency, you know customer support is the easiest one that we see repetitively whether it's on text level on the voice level." [5]
On the vision for ElevenLabs: "We're trying to create now is a platform for the entire audio AI world which helps publishers and entertainment companies to create audio AI content." [1]
On the entrepreneurial journey: "The enjoyment that comes from creating something new that people love using pushed both Piotr and myself to set up weekend hack projects." [7]
"We thought up many interesting startup projects: a new optimizer, a new recommendation algorithm, a new voice tech product. We first built accent detection software. This then matured into our idea for automating dubbing which stuck." [7]
On the mission of ElevenLabs: "We're now on a mission to make all spoken content available in any language & voice." [7]
On the future of AI models: "We actually think that there will be one or two new model breakthroughs that are required in the audio space to really get it to another level." [2]
"What we think will happen over this year... is going to be multimodel approach which effectively combines the LMS the reasoning side and and the audio state-of-the-art." [2]
"It's not only the model that matters it also matters how you deliver that experience to the user." [3]

Learnings

Focus is paramount for a startup's success. Staniszewski emphasizes that maintaining focus is a crucial lesson he's learned, quoting Steve Jobs on the importance of saying "no" to other good ideas. [7]
A strong co-founder relationship is a significant advantage. He frequently speaks about the strength of his long-standing friendship and professional relationship with his co-founder, Piotr Dąbkowski, as a key element of their success. [2][4]
Real-world problems are powerful catalysts for innovation. The frustration with poorly dubbed movies in Poland was the initial spark that led to the creation of ElevenLabs, demonstrating that personal pain points can lead to significant business ideas. [3][5]
Technological breakthroughs often come from applying concepts from one field to another. His co-founder's experience in image and text-based AI was applicable to solving problems in the audio space. [2]
Start with a smaller, solvable problem to build towards a larger vision. ElevenLabs initially aimed to solve the complex problem of dubbing but strategically started with text-to-speech, a more manageable first step. [2]
User experience is as important as the underlying technology. Staniszewski highlights that the way a technology is delivered to the user is a critical factor for success, not just the power of the model itself. [3]
Embrace a multimodal AI future. He foresees that the next significant advancements in AI will come from combining different modalities, such as language models and audio generation. [2]
AI has the potential to revolutionize accessibility and education. Staniszewski is particularly proud of the applications of their technology in helping people with voice loss and sees vast potential in personalized education through voice. [1][4]
Responsibility and regulation must accompany powerful technology. He acknowledges the potential for misuse of AI and advocates for a combination of corporate responsibility and government regulation to create a level playing field. [1]
The future of human-computer interaction is voice. Staniszewski firmly believes that voice will become the primary interface for our digital interactions, as it conveys more emotion and nuance than text. [2]
Building a successful AI company requires both research and product innovation. He notes that their success comes from not only building their own advanced models but also creating a user-friendly platform. [2]
Don't be afraid to pivot and adapt your initial idea. The initial focus on accent detection software evolved into the broader and more impactful mission of automated dubbing and voice generation. [7]
The entrepreneurial journey is driven by the joy of creation. The satisfaction of building something that people love to use was a primary motivator for Staniszewski and his co-founder. [7]
Identifying a clear market need is essential. Before committing to their idea, they validated the problem by speaking with potential users who confirmed the pain points in audio production. [2]
Sometimes you need to build your own tools if existing ones aren't good enough. Realizing that existing audio models were insufficient, they took on the challenge of building their own from the ground up. [2]
The pace of AI development is incredibly fast. Staniszewski's comments suggest a belief that significant milestones, like passing the Turing test for voice, are on the near horizon. [5]
AI-powered agents will become an integral part of our daily lives. He envisions a future where personal AI agents handle various tasks for us, and voice will be the natural way to interact with them. [6]
Unique data can be a significant competitive advantage. He advises companies to look for opportunities to build AI products based on their unique data sets, rather than just focusing on efficiency gains. [5]
The most impactful applications of a new technology are often unforeseen. Staniszewski expresses excitement about the unexpected use cases that emerge from the community as they explore the possibilities of the technology. [1]
A strong company culture is a key asset. While not a direct quote, the emphasis on their team and the journey of building the company together suggests the importance of a positive and collaborative work environment. [8]
Persistence through early challenges is crucial. He mentions having demos in the first six months that "wasn't very good," highlighting the need to persevere through the initial development stages. [5]
Open-source can be a source of inspiration and validation. The existence of an open-source model, while not perfect, provided a glimpse into what was possible and validated their direction. [3]
Enterprise adoption of AI is moving from novelty to necessity. His background at Palantir and BlackRock has given him insight into how cutting-edge technology gets integrated into large-scale business workflows. [9]
Global ambition requires a multilingual focus. Staniszewski sees ElevenLabs' focus on multiple languages as a key differentiator and advantage over competitors. [9]
The ultimate goal of technology should be to enhance human interaction. He speaks of a future where technology fades into the background, allowing people to focus on learning and connecting with each other. [4]

Learn more:

Lessons from Mati Staniszewski

Quotes

Learnings

Written by Antoine Buteau

Lessons from David Holtz

Lessons from Cristóbal Valenzuela