Today’s widely used virtual personal assistants, while intelligent and useful, are not yet particularly “personal”, and by lacking in depth personalization, they can be a barrier to broader adoption of AI assistants by the public.
At the Virtual Assistant Summit
in January, Ilya Eckstein, Co-founder and CEO of Robin Labs
, gave a presentation focused on the importance on personalization as a means to human-machine bonding, and explained why the notion of a one-size-fits-all assistant is a contradiction in terms.
Robin Labs is a Silicon Valley startup aiming to push the boundaries of conversational AI, a quest that is aligned with Ilya's personal ambition of converging technology, science and art to create superior products that resonate with people and society as a whole. I caught up with him to learn more about his goals for Robin Labs, future developments in AI, applications for virtual assistants and more.
Tell us more about Robin Labs, and why you founded the company.
Robin Labs takes its name from our first and most popular consumer product called Robin. Robin was conceived to be the driver's voice-operated co-pilot (a "Siri for the driver", if you will). My co-founder and I both had worked a number of years in the mobile maps/navigation area and it became clear to us that drivers actually need a minimally-distracting voice assistant for both navigation, communication and infotainment, so we decided that every Batman deserved his own Robin.
Then, as usage grew, we so overwhelming demand for many different flavors of assistants. Realizing we cannot have them all in one platform, we eventually expanded to a platform for conversational assistants, called Robin.AI. We also had an epiphany about the deeper meaning of the name, in hindsight: Robin
can actually stand for Rob
What are the key factors that have enabled recent advancements in virtual assistants?
1. Everybody having an internet-connected personal device in in their pocket at all times.
2. Advances in speech recognition - it mostly works.
3. Ubiquity of APIs that enable low-cost integration of sophisticated services.
4. More generally, computing machines have gotten much faster, enabling all of the above.
5. Finally, Siri deserves the credit for showing the world what was actually possible with today's technology and awakening this space, in a way.
What is your main goal of Robin Labs?
To be a global force that helps create technology that adapts to humans instead of the other way around. Technology that doesn't distract us from the real world, but rather helps us be present in it.
What are the main applications for virtual assistants at the moment?
The applications we are seeing are:
- In-car voice assistants -- this trend is actually growing as "connected car" catches on.
- "Smart home" voice assistants, e.g., Amazon Echo. It's still early, Echo and various voice-operated TV boxes is just the beginning.
- Conversational commerce, "concierge for X"
- Bots for work, e.g., the proliferation of Slack bots.
- And of course, many aspects of customer support.
Which new verticals and industries will this expand to in the future?
I believe most devices / services will eventually have a conversational interface, either via voice or chat. It is just the most natural way of communication known to man, and technology will eventually get there.
What developments can we expect to see in virtual assistants in the next 5 years?
See my answer to the "main applications" - we will see incremental progress in all of the above. To consumers, I think perhaps the most apparent short-term progress will happen in the living room.
What advancements excite you most in the field?
First of all, it is extremely fulfilling when we can impact somebody's life in a significant way, such as enabling a blind person to request and receive guidance to a train station, for instance. These things really get me going. But I am also a computer science geek and I am really excited fundamental advances in the field, e.g., the recent progress in deep learning. However, while impressive, deep learning techniques are still mostly about pattern recognition rather than AGI (Artificial General Intelligence). Important and significant, no doubt, but not quite the holy grail just yet.
At the Deep Learning Summit in London on 22-23 September, we will be holding an additional Chatbots Track to explore how AI and deep learning is being applied to create intelligent agents with conversational interfaces, and the impact they will have on business and society. Discounted passes end on Friday 29 July - book your discounted pass here now.The next Virtual Assistant Summit in San Francisco will take place on 26-27 January, alongside the Deep Learning Summit. Previous events sold out, book your Early Bird pass now to avoid disappointment! Visit the event page here.
We are holding summits focused on AI, Deep Learning and Machine Intelligence in Berlin, London, Singapore, New York, San Francisco and Boston - see all upcoming summits here.
Virtual Assistant Summit
Natural Language Understanding