For the majority of us in developed economies, our lives run on software. Computing power is embedded everywhere, in everything and it’s invariably collecting data. The data may reveal our consumer preferences; or it could provide analytics on a factory-floor machine part or the geology through which a drill bit passes, but every interaction of man and machine is producing a stream of data, which is becoming a torrent and creating value where there was previously none.
Data has been cited as the “new oil”, the “new gold”…you will doubtless have heard this. In fact, it’s been likened to any substance that historically created industries, markets and transformed economies. Whilst there’s an element of cliché about data as the new anything, the fact is that data is powering the fourth industrial revolution we find ourselves living through. And its impact will arguably be bigger than that of oil in the long run.
It’s this torrent of available data plus the power and increasing ubiquity of cloud computing, that is powering the adoption of Artificial Intelligence. And AI is the game-changer for every single industry.
So what is AI? Artificial Intelligence can be classified into three different types of systems: analytical, human-inspired and humanized artificial intelligence.
Analytical AI generates a cognitive representation of the world using learning based on past experiences to inform future decisions. Human-inspired AI has elements of cognitive and emotional intelligence – understanding human emotions and considering them in decision-making. Whilst humanized AI is able to be self-conscious and self-aware in interactions.
Much of this sounds like the realms of science fiction. So let’s simplify: AI is a system that can correctly interpret data, learn from the data and use those learnings to achieve specific goals and tasks through – and here’s the interesting part – flexible adaption. So take as an example a chatbot that triages retail investor enquiries, learning from each interaction (think Frequently Asked Questions) freeing up humans to spend time on more complex requests. Or an HR department that assesses a first wave of job candidates via AI – reviewing CVs and in, in some cases, conducting ‘interviews’. Or a system that monitors elevators, medical devices, mine shafts or oil rigs, sending for engineering before a dangerous break-down. The chances are, you’ve already interacted with an AI system and helped it become smarter and you may not have even noticed.
The power of AI and its rapid global adoption requires all of us: commercial organizations; software developers; Governments and citizens to have a point of view on what AI should and should not do. Many of us will be familiar with discussions around data privacy – organizations have wrestled with the implications of GDPR since its roll-out. However, how many of us have considered who is developing AI, how machines are being trained and what data sets are being used to create models?
If only a small group of individuals develop AI systems, it will be set up to be reflective of their world view and experience. If machines learn from trolls, then they’ll behave like trolls. If machines learn and flexibly change, how can we as humans understand the ‘black box’ data that has resulted in an individual or corporation being denied a loan? Ethics and principles must be part of our conversations around AI.
AI’s potential is, truly, awesome. It’s not the new oil. It’s bigger than that.