A Tale of Faith in a Founder
Microsoft announced this week that they acquired Semantic Machines, a Conversational AI based company that is reinventing how computers and humans interact.
We are thrilled that the company has found such a great home, one that can commercialize the technology into a vast array of products. The company deserved a great partner.
We invested four years ago in Semantic Machines for a few key reasons. First, we had gotten to know Dan Roth, the company founder and CEO who was a noted Boston based entrepreneur and someone who clearly had “over the horizon” radar in the speech and language AI space. Dan had previously founded, grown and sold Voice Signal, which was an early leader in the mobile voice interface arena. The company powered millions of early mobile phones with the ability to speak commands into a mobile phone, giving users a brilliant new experience in mobile interaction.
After selling Voice Signal to Nuance, he took a busman’s holiday and started Shaser Bioscience in 2007; a hardware company creating a novel, FDA-cleared, home use laser device for cosmetic dermatology applications. The product was terrific; he sold it to Spectrum Brands’ Remington division in 2012, and then decided to go back to his first love of AI interface technology. Dan teamed up with his former head of R&D from Voice Signal, Larry Gillick, who had gone on to become the Chief Scientist for Siri at Apple. Dan also recruited Dan Klein from UC Berkley, one of the world leaders in Natural Language Processing and Machine Learning who took leave to co-found and build the company.
Semantic Machines team, Berkeley, CA.
The core mission of Semantic Machines was to enable a next generation “conversation” between machines and humans. The current state of the art with Siri, Cortana, and Alexa was all basically unidirectional — a human asked a question, the computer gave the best answer it could. Based on advances in machine learning, new algorithms and the ability to store and learn from massive databases of conversations, the team saw the ability to create true conversational understanding, where humans could talk to a machine; the machine would know or learn the context of the query, and be able to respond appropriately.
Over the course of the next three years, the team began a journey to fundamentally change the way Conversational AI technology works. Part of building the team included getting the best researchers and developers in the space; including luminaries like Stanford Professor Percy Liang.
The company became known as the hothouse of Conversational AI innovation. But it was clear to the team that this was an ambitious goal, and Dan was committed to investing in the research needed to develop true breakthrough technology before they would turn to commercialization. This was at times quite a challenging dialog, as investors we often take market risk but funding a core technology for several years without commercialization is not for the faint of heart. But we saw great results from the team and based on the traction they were seeing among the AI cognoscenti we held our breath and continued to fund the company without hesitation.
Earlier this year it was clear that the company’s technology had reached a strong and viable maturation point, and many potential partners came calling. The team had many choices of whom to partner with but they really wanted to see their technology get into the hands of billions of users and continue its development in the right environment. Microsoft’s interest and partnership was compelling, and it became clear that this was the best partner for a company with ambitious research and development goals.
At GC we look at companies first through the lens of the founder, and we could not have worked with a better founder than Dan. His insight, passion, determination, and team approach made him the perfect person to set forth on this journey. It was not always clear how the company would develop but it was very clear to us that Dan would lead the company to something terrific.
We want to thank Dan Roth, Dan Klein, Percy Liang and the whole team for the wonderful partnership with GC. We look forward to many Microsoft products powered by Semantic Machines. It will be a great conversation.