Meet Mantas Aleksiejevas, Export Business Development Manager at Google, a tech enthusiast and business development professional with over 5 years of experience in digital growth strategy development and realization across a range of different industries. Mantas holds two science degrees and strong interest in how digitalization, data and automation is driving deep changes in the world around us.

He'll be speaking at Turing Society's inaugural event on November 30 in Vilnius where he will share his personal point of view about the AI industry. And yet Mantas found some spare time to share some of his views with our readers.

What industries/skills are prone to be impacted by AI the most? Why?

Mantas: I think a good way of look into this is through the lens of conditions that are needed for the technology to pick up. Take Machine Learning as an example. Machine Learning models are the most powerful when you have large amounts of good quality data, trained models or expertise to build them and powerful computing infrastructure.

In this context I believe that the current successful tech companies are well positioned to drive innovation through AI-like solutions. They do have the infrastructure, expertise in building products through advanced computing and also vast amounts of good quality data. This is why I believe that with the rise of AI-like self-learning computing we will see an acceleration of change in such industries as travel, retail, media and communications.

Energy and healthcare are another two industries worth keeping an eye on. They are systematic by nature and also extremely large in terms of the people they touch. As soon as good data gathering systems and connectivity infrastructure there will be in place, a vast number of new opportunities to drive major efficiency improvements and better services will arise. There are already isolated cases, for example, when machine learning models applied to data center cooling systems drastically reduce energy costs.

Last but not least, transportation and logistics. Again very systematic industries, which hold huge opportunities in efficiency improvements through data aggregation and self learning models. The major challenge in optimizing these systems have so far been human agents – great logistics on paper were easily affected by human driver mistakes. This will not be the case with the self driving car technology, which is already at the state where it drives better, safer and more efficiently than humans. Once we have self driving technology in place to transport both us and our goods, it will clearly open up huge opportunities for logistic and transportation improvements through AI-like computing.

Don't we live in AI hype? Can you name a few success stories and expectations that haven't materialized yet?

Mantas: That depends on what angle you look through. One could say that in late 90’s-early 00’ the Internet was overhyped. However, very few would deny that since then, Internet-based technologies have touched upon nearly every aspect of our lives. I believe we can look at machine intelligence in exactly the same way – regardless of whether it’s overhyped or underhyped, machine intelligence will penetrate our lives increasingly more bringing both new services and major efficiency improvements all across the board.

Why now? I think we are at the point where computing infrastructure is sufficient to handle processing needs that are required for AI-like models to run. We also have devices that each of us source information from, connect and interact through, use for navigation, entertainment and many other daily tasks. All these devices create signals that AI-like models can use to personalize and optimize products and systems. Also, there has been significant AI-related discoveries in computing with various new models popping up every month and also being successfully integrated into popular products. AI has been the topic since computers were created but never before the conditions were so good to utilize the self-learning aspect of it.

Regarding the success cases, the AI-like technologies are already used in many products. For example, Amazon is running quite sophisticated models for website customization to match the needs of every individual user, Google use machine learning to understand content and organize photos and videos, optimize natural language processing and translation systems. DeepMind, SwiftKey, Magic Pony, and VocalIQ are examples of successful AI start-ups.

Where would you invest when it comes to AI-related solutions?

Mantas: Short term – current tech companies as they are best positioned to drive major improvements through AI-related technologies and do it fast.

Mid-Term – logistics and self driving technology. It right now is constrained by regulation but the technology is there, so once the regulation improves, the field will see major growth.

Long-Term – energy and healthcare because they have relatively been undisrupted but once the data infrastructure is in place, such technologies as machine learning and robotics will drive massive improvements.

Where do you think the biggest AI businesses are going to be created? Where are biggest AI hubs based now?

Mantas: I think that major hubs are not really domain specific so I would rather look in general on what is needed for a successful hub and which locations have got it.

To my understanding a successful hub is a place than has got: a) talent and ability to attract more of it, b) capital and smart investment funds, c) flagship success cases, d) favorable immediate business environment (e.g. big local market).

This is why I believe that US (Silicon Valley and NYC) will remain the frontier because it’s got all of the ingredients. London will also be strong because of educational infrastructure, capital and very favorable local market conditions. Berlin will become the key continental Europe hub. It has done tremendous job in attracting talent over the last couple of years and also has got very favorable local market with good connection to the other DACH countries and France.

What skills will be the most relevant in the dawn of AI?

Mantas: I think the need for computer scientists, data architects, cloud infrastructure developers and other roles that will work on developing computing models and infrastructure will remain in a very high demand. These jobs will also shift heavily into new computing paradigms at faster than ever pace. This is why we will also see the increased demand for educators and education platforms to support the demand for continuous learning.

AI-related technologies will also require non-technical human trainers to teach and tune the models. In the same way as we have processes for training people right now, we’ll need processes and trainers to train the computers. It is hard to say what skills that will teaching computers involve but clearly the ability to at least conceptually understand computer systems, technological development and communicate observations to developers will be important.

The other knowledge that I think we will see an increased demand for is within the domains of psychology, brain science and biology. Deep neural networks, for example, is a paradigm that is built partially in the way human cognitive processes work. As computer science and computing systems will continue to develop, biomimicry will become possible to much greater extent and thus both people on the computing science side and natural science side will be needed.

On the business side, I believe major shift will be happening in operation and business management functions. We will see an increased demand for data-savvy, data-critical, constructive problem solvers who will also be good in managing change and ambiguity. We can already see that successful technology companies change much faster and much more often than traditional similar size companies. Once AI starts accelerating business change even further, the entrepreneurial smart change managers will become a very important part of any organization.

Lastly, we should not forget that we will increasingly more need people who will help to reshape skills and social systems in the industries that will be negatively impacted by another rise automation. I think we will need community organizers to bring impacted communities together, social designers to develop social and tax systems that are of a better fit for a digitalized world than the systems we have now and also trainers/educators that will help people in impacted industries gain new skills and re-qualify.

Thank you for your time, Mantas, see you next Wednesday!