Yesterday’s sci-fi is quickly becoming today’s reality.
AI and deep learning are disrupting entire industries. Sensors are in everything we use. We don’t just have conversations on our devices - we now converse with them (sometimes with hilarious results).
Michael Zolotov is at the forefront of the AI revolution. As Co-Founder & CTO of Razor Labs, and Axon Vision, he has deployed deep learning systems in a wide variety of industries including traditional industry, medical, security, and Smart Cities. From his perspective, AI is not a niche - it is business-as-usual. It’s here, it’s growing exponentially, and companies in all industries must embrace it or risk getting left behind.
We spoke to him ahead of his session at #Innov8rs Tel Aviv to find out what he’s been working on, where he sees AI heading, and how large traditional organizations can benefit from this new technological revolution - even if they think it doesn’t apply to them.
You're working on all kinds of cool stuff: "Far away or interesting technology." Can you explain what exactly you do, and what the implications are?
I focus on the world of AI: artificial intelligence. In cyber-efficient intelligence, we deal with deep learning, which is a subset of this enormous world. It's a world that has been rising only in the last one or two years. It's very new, very innovative - especially, of course, to larger corporations, who are a little behind.
Deep learning allows you to solve problems that once only humans could solve.
For example, understanding what you see; understanding what you hear; understanding what you read; analyzing patterns in very, very complex data. Just a year ago, these sorts of problems were only solvable by humans. Now, machines can solve them, and even better.
We're excited to be a part of this revolution. We program those capabilities and integrate them into companies so they can analyze their data. It’s probably best if I give an example.
Let's say you are a city, and you want to be able to recognize a complex scenario - like violence - as it’s happening. As of one or two years ago the way that worked, or the way algorithm developers worked, was by manually analyzing what they saw. So to analyze violence you analyze the shapes, the colors, the structure of what you see; but, you're limited to the imagination of the developers, of the programmers.
Our programs at Razor Labs work much like the human brain. When you see violence, you immediately understand what you're looking at because of your experience in life, because of the many other scenes of violence you have seen before. Our systems learn about the environment. They learn using databases that we put in, using many scenarios that train the system. So, after giving the system a few cases of violence and a few cases of terror attacks, it can automatically recognize other cases with a very high success rate. This is something that was impossible only half a year, or a year ago.
It sounds like there's exponential growth; it's only recently started to emerge, and now suddenly it's capable of a lot. What's do you think will happen in the next 25 years? What will these new AI technologies lead to?
The tools we have today are only the beginning of the revolution. Most organizations today do not use those tools or are in the very beginning of understanding their potential.
My vision, the vision of this industry, is that all the data we gather, whether it's customer data or visual data or whatever we need to analyze, will make human life better. It will give us environments that are safer and more comfortable for us.
This is the vision in the end: fitting everything specifically to your needs in all aspects possible. Your hobbies, your workplace, your trips - everything. Not only in the digital world but also in the physical world. The IoT sensors that are connected also to the Internet and, just like in Smart Homes today, automatically turn on your lights or hot water, will also be used to make the whole city or the whole environment smarter. This is the next stage.
What priorities should big corporations have around this technology? It seems like the number one requirement would be to take it seriously and really dive into it?
Yes, they do, but the main challenge for bigger organizations is connecting their business needs to the technology. They need to understand how this technology or how this new AI concept will lead to better business results, and the sooner the better.
For example, if you're the CEO of a food company and you want to optimize your entire manufacturing line to be faster and more economical. How can you use AI technology to do that, right from the beginning? After you already have the product and you want to bring it to your customers, how can you optimize your marketing campaigns, your coupons? In the end, you want to know who your customers are and you want to give them exactly what they specifically want.
Today one of the biggest revolutions is Industry 4.0, which concentrates on two main areas. The first is optimizing industrial pipelines, the manufacturing lines, to be able to produce more with less human power and in less time. The second is being able to customize what you manufacture to the end customer, which also connects to the whole 3D printing concept. In the end, if you know who your customers are and you know what they want, the next stage is being able to manufacture specifically to their individual wants and needs.
Do you have an example of a company that you think is getting this, and is applying it at the best speed and quality possible?
One of the most advanced companies in the world, in this regard, is Google. About five years ago or so, they were saying they were a mobile-first company. Today, their CEO says that Google is an AI-first company. You see AI in all of their products - even the speaker they put on the market, you can actually interact with the speaker using Google Now. It understands your preferences, what music you like, what information you want to receive in the morning when you get up or at night when you go to bed. Even the hardware itself - the Google Home Max uses machine learning to give you the best sound quality using the same AI concept.
What if you are a legacy organization in a very traditional industry?
The first step is to acknowledge that this is a revolution, and it is happening.
It is important, not only to the DNA of the organization but to the end business results. There are companies, like Razor Labs, who come to organizations from the outside and implement innovations inside the company. So we're working with the corporates in order to connect their business needs, their competition, and their market needs to this AI revolution. We're working with the C level and CO's of these companies in order to understand their business needs and connect them to the algorithms that will help them accomplish their business goals. They need to understand how important this is because if they're left behind, in the end, there will be no solution to their situation.
So you're saying if you don't take this seriously and join the revolution, you probably won't be in business in maybe 5 or 10 years?
Exactly. You can see it today in all of the markets. For example, companies that manufacture sensors of all kinds - they understand that they can’t just manufacture the sensor, they also need to make their sensors smart. They need the sensors to actually understand what it sees or what it hears, and then give those insights to the human operator. It can’t just pass on the raw data anymore, that’s not enough.
IoT companies understand that it's not just the device that matters. Sure, they want it to be small and low power and everything. But this device must be smart and understand what is happening around it. In the audio world, in the internet world, the whole social media marketing and e-commerce revolution, it's the same thing. In the e-commerce world, understanding your customers and giving them the exact recommendations that fit them - this is, in the end, the difference between a selling platform and a non-selling one or one with low sales.
AI, deep learning - it's not a niche. It affects the whole industry and all industries.
I cannot imagine a single field that is not affected by this technology. And if you think your industry has no use for it or isn’t going to be impacted by it now or in the next few years, you need to think again.
What do you think - how will this play out for your industry and company? Let us know. We’d love to have a conversation with you about it.
