Artificial Intelligence - the technology of the future
29 October 2018
Artificial intelligence, also known as AI, is the phrase currently on everyone’slips and minds. Having moved beyond the realms of science fiction into mainstream consciousness in recent years, there is no doubt that AI is transforming the world in which we live and work. The LINK spoke to a range of experts within the sector to get their views on AI, and how they believe it can transform our future.
Mention AI, and most people will think of robots. But what is AI really? Simply put, AI
is intelligence demonstrated by machines. Luka Crnkovic-Friis, CEO and Co-Founder of Peltarion, a provider of operational platforms for producing real-world AI applications at scale and speed, and also members of the Swedish government’s recently formed AI council, explained that AI is a field of technology that uses large datasets to train computers to learn from experience, adapt to new inputs and complete tasks that resemble human intelligence.
Tommy Högström, Chief Smartification Officer at Applai, a company delivering AI for business solutions, said that although it has many definitions, he would explain AI as an umbrella term for technologies that enable machines to mimic human intelligence, such as seeing, comprehending and learning. An important part of AI is Machine Learning (ML). Högström said: “ML is a subtechnology of AI that learns from data,often to predict the future of different kinds of human and machine behaviour.” Crnkovic-Friis, also noted that the most efficient and popular AI technology today is called Deep Learning (DL) which is a subfield of ML aimed at finding hidden patterns in large amounts of extremely complex data. DL works as a process that is loosely modelled after the human brain and can be trained to do things like detect leukemia earlier than human experts, drive a car, help a restaurant better predict food demand or enable a global retail company to optimise the logistics processes. In fact,Crnkovic-Friis said that the current AI boom is directly related to the advances that have recently been made in DL.
The prospect of the widespread application of AI is now spreading beyond the technology sector, resulting in large transformation for companies, workers and consumers. AI is already all around us, enabling Apple’s Siri to understand us, emailing clients to detect spam and retailers to recommend products based on internet behavior, Högström said. Another example is self-driving cars which are on the rise as well as digital assistants, assumingly becoming harder to distinguish from humans.
For Högström it is obvious that AI can help improve several areas and he highlighted that what AI really brings to the table in our everyday lives is convenience.
Crnkovic-Friis said that there is really no limit to the areas where AI can be used. The sectors or industries which already have access to lots of data, have in-house AI expertise and can afford expensive computing power, are naturally going to see the quickest return on investments. However, even 50 years after the adventof AI research, we’re still in the early days of applying AI to business, Crnkovic-Friis highlighted. Yet, important to note is that we are on the brink of a rapid upward trajectory, he added.
The application areas where companies are using AI today are many and quickly increasing. Högström mentioned customer intelligence, financial analytics, manufacturing, supply chain and intelligent automation as common implementation areas. Within customer intelligence, it enables companies to know what consumers want, when they want it and how they want it, but also exactly what customers are monetarily worth. Högström said: “Most large retailers have already started to analyse their customers and to personalise communication However, increased customer demands for personalised communication will ramp up this development during the upcoming years.”
For financial analytics, various types of transactional data in combination with customer profiles, industry trends and text data are of tremendous value to banks and insurance companies who use it to perform advanced risk calculations and fraud predictions.
Within manufacturing, production companies collect vast amounts of data such as equipment usage, vibrations and temperature which can be used to maximise production up-time and optimise quality but also to predict maintenance needs. AI is also important for the supply chain where “smart algorithm companies” can perform precise availability forecasts of products in stock and delivery times to customers. Further, intelligent automation is an area which is extensively used. Commonly applied areas for automation are chatbots interacting with customers and various manual data entry tasks such as order processing or internal IT ticketing. It is clear that the implementation areas of AI are vast and diverse.
Margareta Groth, Head of Industrial Technologies Department and Mikael Hedelind, Programme Manager and National Expert at Vinnova, Sweden’s innovation agency, stated that AI has the potential to be applied in most industries as well as in the public sector. They highlighted that the effects of AI can be beneficial in both the internal processes of organisations and as part of products and services provided to customers. In addition, they highlighted that we can find AI being used in both the law and financial sectors already and examples of current research areas include autonomous vehicles and health applications. Vinnova also stated that: “Some of the ‘hottest’ areas of development include image processing such as facial recognition (which can be incorporated in a wide range of applications) and hardware (chip industry).”
Even though AI has been praised a lot, the technology has also frequently been a target for criticism. For example, Högström, mentioned that the implementation of AI is not trouble free and there have been vivid discussions of what data that companies should be allowed to use. Questions such as “should companies be allowed to base individual insurance pricing on criminal history of geographic origin?” have arisen. Others question whether or not we should encourage a world in which robots and machines are given more and more space over human elements.
Groth and Hedelind highlighted that as a governmental agency they have to consider other implications of various technologies such as dynamics related to labour and employment. The impact AI might have on jobs was recently explored in a report by PwC on the UK economic outlook. In the report, PwC concluded that AI and related technologies should create as many jobs as they displace. “AI and related technologies such as robotics, drones and driverless vehicles could displace many jobs formerly done by humans, but will also create many additional jobs as productivity and real incomes rise and new and better products are developed,” the report stated.
Although opinions on the development and implementation of AI may vary, countries and companies all over the world are putting a lot of resources into its implementation and development. Groth and Hedelind said it is a disruptive technology that can change businesses and value chains completely. For Vinnova, supporting the research of AI is important since the agency wants to support Swedish industries to be as productive and competitive as possible, as well as making the public sector efficient and easy to access for citizens.
One example of how Sweden works with the implementation and development of AI is through initiatives such as the Wallenberg Artificial Intelligence, Autonomous Systems and Software Program (WASP). SCC Members including Atlas Copco, Electrolux, Ericsson, Investor AB, Saab Technologies and SEB have all contributed to the programme. As a major national initiative for strategically motivated basic research, education and recruitment in artificial intelligence, autonomous systems and software development, WASP’s ambition is to advance Sweden into an internationally recognised and leading position in these areas.
Fredrik Heintz, Associate Professor of Computer Science at Linköping University and highly active in WASP, stated that AI has the potential to be the next major transformation of our society. Understanding, developing and applying these techniques are therefore crucial for Sweden, he continued.
The top four countries for prominent research within the fields of AI, according to Heintz, are the US, Canada, the UK and China. He further explained that AI research in China has seen a truly impressive progress over the last five years, which shows that if a country really puts in the resources and makes AI research an important goal then it is possible to rapidly make significant improvement.
Heintz highlighted the importance of Sweden staying competitive in the field as AI is expected to be a significant part of basically every product and service of national importance to Sweden. Heintz said: “I believe that Sweden needs to be self-sufficient in knowledge and competence in AI as the rest of the world are also pursuing it and there is a huge lack of competence and knowledge.”
As a country, Sweden has a long tradition of AI research, he added. “We have been doing research since the 1960s thanks to the efforts of Erik Sandewall who brought AI to Sweden after spending time with John McCarthy, Marvin Minsky and the other founders of AI in the United States.” Nevertheless, Heintz emphasised that research has not been prioritised for the past 20 years and as a consequence, Sweden currently only has a small number of AI researchers and needs significant and sustained funding to expand. Groth and Hedelind also noted that Sweden as a country are “doing okay”, but that there is a potential to improve.
Vinnova and WASP both agreed that Sweden needs to invest in the development of AI in order to stay competitive. Heintz stated: “Even though there has been impressive progress in AI research the last 10 years, we are still only in the beginning of the development. There are many fundamental research challenges that haven’t been addressed.”
Examples of research challenges, he said are how to learn from significantly smaller data sets, how to transfer what has been learned about one task to other tasks, how to verify and validate ML models to provide guarantees, learn causal models and make the machines able to explain their decisions.
Crnkovic-Friis highlighted that even though more and more people, companies and organisations are hoping to use and implement AI, the technology is still inaccessible to most companies, because they do not have the right tools and resources to support it. What the future will hold for AI is still unknown but it is evident that we will have to continue to explore the technology. Groth and Hedelind believed it will continue to grow and be incorporated somehow in most (if not all) parts of different industries and the society as a whole. In general, they said, they think of AI as a tool that can bring a lot of good to society and as long as it is incorporated in a sustainable and responsible way they see a bright future based on new technologies.
AI is changing the world we live in, transforming the way we live and opening up for opportunities we could not have dreamt of a hundred years ago. Regardless of how AI develops in the years ahead, there is no doubt that the world will forever change as a result of advances in AI.