Aleksandr Kotelnikov is a senior engineer with nine years of experience in diverse electronic engineering, including focusing on fields such as Nuclear Physics, High-Frequency Trading and Machine Learning.
His current professional interests include high-efficient deep neural network hardware solutions for video and vision applications, computer architecture & micro-architecture, algorithm optimisation for hardware implementation, digital ASIC & FPGA design.
We talked to Alexsandr for a quickfire Q&A of some Artificial Intelligence (AI) related issues.
Do you see a place for AI in all businesses?
I think that AI technologies may be incorporated into various processes within a wide scope of businesses. Machine learning may be quite successful and well surpass human performance in cases when there is a huge amount of data, it’s well structured, organised in database and the data cannot easily interpreted by humans. I believe there are always functions which require human being and may be irreplaceable by AI but even such roles may be enhanced by AI via advisors, for example.
Do you think people should have doubts about the power of AI?
We definitely should not overestimate its power. We saw Artificial Neural Networks' "winter" in the past, and we may be stuck in making further progress in AI again. Current AI technologies will not immediately bring us to the technological singularity or some futuristic world, so society's expectations should not be too high.
Having said that, we should not miss opportunities for AI technologies adoption to various applications so our efforts can be applied to the right problems. Also, ethics and safety is an important area where we cannot blindly count on any technology. Human roles and responsibilities are still very high here. We have already heard a lot about the bias problem in machine learning and seen many examples of chatbots learning inappropriate behaviour from real human conversations. AI is just a sophisticated tool, and a human is behind it.
Some people think AI won’t last, whilst others believe this about to be the biggest shift we have seen in tech. What are your thoughts?
I would consider the current AI boom as the next iteration of computer science and engineering advances. Indeed, whilst we had some kind of AI in the past, we have it in more advanced forms now, and it will evolve in the future. I think that, despite further AI technologies development, we are still at the beginning of applying the most recent AI algorithms to real-world problems and we will see more successful AI applications in the nearest future. Some of them may be not very obvious at first glance. For example, you may not even know that your salad is made from vegetables sorted by AI at a farm.
Some people doubt that we will get self-driving cars very soon. Even if they were correct, I would not consider it as AI total failure, because we may still get something instead of self-driving cars, like AI based driver assistance systems or autonomous vehicles of other types such as ships etc. So, I expect to see a lot of exciting things happening in the foreseeable future.