Several leading digital specialists in Estonia share their predictions of what the year 2022 will bring in the realm of artificial intelligence. Although the interviewees come from different walks of life, the visions they put forward are surprisingly similar.
In addition to technological trends and important questions they verbalised, they all agreed that the Estonian public sector will continue its leading role in the digital strategy. The outgoing Chief Information Officer of Estonia, Siim Sikkut, most clearly declared that his job has been not to predict the future, “but to make the future”. In this article, I will give an overview of what kind of future is in the making.
Communication with AI becomes very natural
Overwhelming agreement predicts AI taking the next step closer to humans in terms of its ability to interact with us. It will be able to understand – and speak – our language, it will be able to see as humans see – and even look like humans. Harnessing of AI in different applications will therefore become so prevailing that the distinction between “normal” and “AI-based” digital systems will lose its significance. As Indrek Velthut, Head of Data, Analytics & AI at Nortal, put it: “From an individual’s point of view, communication with AI becomes so natural that in the very near future we will not imagine them otherwise.”
An example that many mentioned is Estonia’s Bürokratt – a seamless, life-event-based network of AI applications that enable citizens to use public services through voice-based interaction. While prototypes have been already tested, Siim Sikkut reported that it is reaching the first tangible milestones and applications in 2022.
Other countries are bound to follow, because „cloud-by-default is becoming the norm even in conservative digital governments”, as Kristo Vaher, Chief Technology Officer, suggests. Indrek Vainu, Head of Conversational AI at Zurich Insurance Company sees parallel developments in the private sector that will enable, for example, just giving command of “prepare a contract on condition X” or “buy product Y” and the machine will complete the task.
Behind this development lies the adaptation of 5G networks and maturing of modern cloud platform technologies leading to low maintenance costs, ease of scalability, and security. This combination will open up the possibilities of harnessing AI power like never before. One example that Karen Burns, CEO, and Co-Founder at Fyma, pointed out is retail shopping without cash registers like Amazon Go is already doing, and competitors stepping up to.
Citizens will have more direct control over their data
Increased application of AI and its integration into everyday lives bring the question of data governance to the foreground. On one hand, democratic countries like Estonia will increase citizens’ control over their data. Ott Velsberg, Estonian Government Chief Data Officer predicts: “Citizens will have more direct control over their data than ever before. It will be possible to easily, reliably, and securely give permissions to share data between government and private sector and vice versa – while also limiting said access, if necessary.”
However, this trend is definitely not unidirectional. “Because of shortages of electricity, processors, people and – worse — capital, countries, and companies need to start asking which systems based on artificial intelligence or blockchain technology are worth the investment,” warns Dan Bogdanov, Director of the Information Security Institute at Cybernetica. This will increase the divergence we are already seeing. Karen Burns titles this a real marmite-moment as „companies either really want face detection or they’re vehemently against it. There is a deepening divide between how personal data is seen and owned by the US and EU, with this unlikely to be settled anytime soon.” The divergence between the EU and China in data governance and use are also apparent.
AI in the hands of the bad guys
Several people also pointed out that AI use is not always benevolent. From a security standpoint, AI can also destabilise the situation. Lauri Almann, Co-founder of CybExer Technologies, points out that „recent cases have shown that cyber-attacks have been enhanced by machine learning algorithms that automatically detect vulnerabilities and launch attacks. This is one of the most dangerous trends in cyber security today, which means that we must also be prepared to develop our defence methods and tools accordingly.” In addition to evil parties, it is also apparent that large technology companies are trying hard to take control of infrastructure, manage people’s identities, cash flow, and entertainment.
Therefore, both Ms. Burns and Mr. Velthut agree that lawmakers need to react to regain democratic control over technology and make investments before it is too late. Mr. Velthut is confident that Estonia has several good examples of distributed trust technologies. The X-Road and Smart-ID share data and keys between parties while avoiding central control. However, he adds: “Our job in the new year is to show that both data and keys can be better protected from all parties in cyberspace, and this game is not lost.”
Problems and possibilities in the internal functioning of AI
While a lot is happening around the application of AI, its internal functioning is also turbulent. Ms. Burns points out new possibilities. will open up by synthetic, computer-generated data that will speed up the development of intelligent algorithms. According to Ms. Burns, “A system can only be as intelligent as the data you feed it with. As demand is picking up and pressure to develop solutions faster is on, synthetic data is one solution to speed things up.”
On the other hand, we are releasing more of the internal complexity of the neural networks of AI. Therefore, Mr. Vainu predicts that “Much attention is paid to making the AI decision-making process clearer so that people can understand why AI offered a particular answer or solution.”
In sum, whether we like it or not, 2022 is going to be an interesting year, not least because of the developments in the AI realm. Be sure to glance at the direction of Estonia for making sense of the trends also this year.
social scientist at the university of helsinki and the estonian university of life sciences