I am a Scientist Entrepreneur with a deep understanding of AI, dating back to the end of the 1980s when I was a Research Fellow at Imperial College and pioneered the machine learning technique known as ‘stochastic gradient descent’ now used in nearly all Deep Learning systems. I was the first to apply it to speech recognition, and went on to develop, in partnership with the UK MOD’s Malvern research lab, one of the world’s first large vocabulary speech recognition systems, and a company, Speech Machines, that provides speech to text services for creating medical records.
Ever since I started Speech Machines, I have been fascinated by the idea of talking to machines, and have watched how the maturation of speech recognition, speech synthesis and natural language processing technologies has created Siri, Alexa and the other ‘AIs’ we can talk to today. And over the last decade, there have been similar leaps forward in Computer Vision capabilities also driven by Deep Learning and the massive increases in affordable computing power provided by modern GPUs. SeeQuestor is at the forefront of these developments, applying latest generation computer vision AI technologies to video analytics and security. Equally important advances have been made in the last few years in mechanical engineering - the structures, actuators, haptic sensors and feedback systems necessary for a humanoid robot that can walk, pick up things, use tools, climb stairs and interact safely in a human environment.
I believe that with the maturation of these three key technologies - conversational AI, computer vision and robotics mechanical engineering - the time is ripe to create the world’s first “Super Friendly Everyone Got To Have One” Humanoid Robot. Much of the best technology in all three fields comes from the UK, and I myself know a great deal about the first two.