AI Assistant in the Job of Home Security

by Diane Bédat

We are living in a new social era of living and with that there is a need for smarter solutions and assistants that can recognise and effectively identify friends, from intruders. Like with FLARE by BuddyGuard, an indoor security camera that protects homes all by itself using artificial intelligence and machine learning technology on a small custom-made IoT hardware device.

At the second annual Virtual Assistant Summit taking place on the 26 - 27 January, George Platon, CTO and Founder of BuddyGuard, will be sharing the company's approach to AI assistants in the job of home security.

I asked him a few questions ahead of the summit to learn more about his role, and how AI can be utilised within home security.

What do you feel are the leading factors enabling recent advancements and uptake of virtual assistants?

One of the most important factors is the increasing hunger for consuming online services. Be it flight reservations, entertainment content, news content, sales or shopping. Finding your way around the best offer, or the easiest way to get the content you need is becoming harder and harder, having so many increasing options. The need is there, it’s up to our bots and virtual assistants to fill the need by assisting us in getting the best deal out there.

Apart from that, phones/servers became much faster (thanks to Apple and the gaming industry!), allowing our assistants to finally process the kind of queries that can provide relevant results (see Siri and Google Voice).

In some parts of the world, there are different groups that already proposed chatbots for our learning processes, and in our educational systems. Definitely an important trajectory that is set for the next generations.

Which industries do you feel will be most disrupted by virtual assistants, and AI in general, in the future?

The entertainment and news industry are some of the most disrupted one - finally we can get content suited to our preferences, not based on the A/B testing of a good marketer. Facebook is yielding much of their money though the personalized advertising, and their advancements in aggregating person preferences through the way we share data, it’s been evolving a lot in the past years. Google is targeting advertisement to people whose preferences are extracted from the general search queries they’re making. It’s a lot of automated personalization work that is being done for many small or large services we’re using.

Another industry that felt the touch of virtual assistants quite fast, was support-center assistants. Aggregating information and yielding resolutions following a pre-defined decision tree was one of the first approaches, yet, they provided good results which encouraged the industry. At some point, there were even discussions about guidelines or better Turing tests for identifying real support-center agents from the robots.

Last, but not least, the whole financier sector got a lot of attention from the progress of artificial intelligence. Banks with AI software are able to yield prognosis better than trained financier visionaries, based on huge amounts of data they were collecting over time. This means better investments, which yields to better services to people in the end. It also brings a lot of understanding on profitable economy areas that deserves attention.

I believe in the future, banks will go through a big change, from what we once knew as giants with huge personnel assets, to mostly online banks enjoying the benefits of automated analytics and predictions rendered automatically from user’s live data. One of these banks already got popular in Europe, N26.

What developments can we expect to see in virtual assistants in the next 5 years?

Ability to:

  • Understand user preferences, and offer services suited to their needs. We can’t have one-size-fits-them-all solutions. In order migrate from the traditional manual work, and really benefit from the help of a virtual assistant, they will have to understand the people’s preferences.
  • Gather large variety of data and process it as general artificial intelligence, not only pattern-finding type of intelligence, than currently for example Deep-Learning is famous for.
  • Integrate with most of the online services we’re using - shopping apps, food, financial apps.
  • Learn from us themselves, and adjust the virtual assistant services specifically to each person.

In your opinion, are we ready for emotional AI?

Emotional AI is one of the end goals of AI in general - having the machines understand us as we are. If people ask me from a technical point of view, if we are ready for it - I would say, we still need few years to pass. For a truly emotional AI, we need first to develop its ability to learn and grow itself to a good intelligence level that allows it to comprehend external data, at least in it’s basic forms. Yet, currently, most of the work on virtual assistants involves supervised learning (which are still really good!), hence, there’s no easy way for a virtual assistant to create a comparison term, which is usually one of the factors that generate emotions for humans. Yet, I’m open to surprises, and I am also researching this topic myself as well.

If people ask me from a user experience point of view, I would say in some highly technological places (few in US), people are ready for it, while 50-70% of the rest of the population need to be educated of interacting with such an AI.

What has driven you to work on virtual assistants?

Helping myself and helping others accomplish goals faster and better. Liberating people from repetitive tasks that can be easily learned by a bot or an intelligent system.

From the geek perspective, the fact that intelligent systems, likes the ones we are building, are not so predictable as standard engineering problems (regular programming, engineering, etc). A bot/virtual assistant/AI can yield much more interesting and infinitely more scalable results than a classical problem, where we know that we go from A to B.

Learn more about the opportunities of advancing trends in VAs and their impact on business and society from a global line-up of experts at the Virtual Assistant Summit, on the 26 - 27 of January in San Francisco. There will be an assortment of speakers including: Anjuli Kannan, Software Engineer at Google; Lionel Cordesses, Innovation Project Manager at Renault; Alonso Martinez, Technical Director at Pixar Animation Studios; and (Elena) Corina Grigore, PhD Candidate at Yale University, Social Robotics Lab.

For more information and to register, visit the event website here. Also, don't forget to use the discount code CYBER20 to get 20% off tickets for all summits up until Wednesday 30 November.