Gartner tells us in its Top 10 Strategic Technology Trends for 2019(1) that autonomous things are number one in the list. Autonomous things use artificial intelligence (AI) to perform tasks traditionally done by humans. We are talking robotics on a new level of sophistication and with greater levels of autonomy than seen in previous generations of robot technology. The types of autonomous things include vehicles, drones, appliances, embedded software agents and specialised industrial robots such as agricultural harvesting machinery. The spaces in which these devices operate include the physical land, sea and air environments and the logical digital environment often referred to as cyberspace. The levels of capability, coordination and intelligence vary widely according to the application. Some require human assistance (or perhaps we should say some are designed to assist human activity – after all, who is in charge here?) whereas some, such as farming machinery, can work completely autonomously without a human operator.
What does all this mean for business risk and the need for security to manage risk?
Digital disruption is the term that has been coined to describe how the introduction of digital technology is completely changing some business models in some business sectors. Some examples from the past include digital photography, digital entertainment, digital music, electronic banking, and digital publishing of written materials (books, papers and the like). In each of those cases the entire industry has been remodelled. The changes span the whole value chain (and contributing processes) of those businesses affected, from product innovation and design, product marketing and sales, goods and services production, delivery to customers and after-sales service. Some traditional functional service providers have been disintermediated and found themselves with no business. Some traditional manufacturers and suppliers have met the same demise.
Those that succeed as their business sector undergoes digital disruption are those that see the changes coming and adapt their business processes early. New entrants often have an advantage of no legacy. Agility is the key to success in such a volatile world. One thing is certain – autonomous things will bring on another round of digital disruptions that might threaten your business model. As with all risks, there is the huge opportunity but an equally large threat. Are you even looking to see what’s coming over the horizon for you?
That’s the strategic risk, but what about the operational risks? It is clear from all human attempts to automate and mechanise processes that there are some serious health and safety issues to be addressed. Whilst AI has certainly gained some ground in recent years, can it really replace human intelligence? The Attributer believes that right now (and maybe forever) it cannot. So, to what degree are we prepared to trust the operation of autonomous machines?
We drew attention to this issue in the previous Attributer blog column with regard to verification of fake or flawed information. AI can only look out for the input signals for which it is programmed. Human intelligence has a huge capability to spot very subtle changes in the input that signal some potential threat – not just single changes but multiple simultaneous changes of minute significance individually, but of major significance collectively. The human can ‘sense’ a situation that doesn’t ‘feel’ right without necessarily being able to analyse immediately all the reasons why. It seems doubtful that AI will ever achieve this level of intelligence. What AI can do is analyse huge numbers of possible actions very quickly and choose the optimum one for the given situation. It can also follow a programmed process without deviation, something at which humans are not so good. We must be clear – humans do some things better than AI and AI has its advantages too.
SABSA views risk as the effect of uncertainty on objectives. In any business system we can express the objectives as a Business Attributes Profile, with measurement and performance-monitoring built in. We can analyse our potential autonomous operation targets and get a clear differentiation between those that are viable and those that are not, within the performance criteria for efficiency, effectiveness and safety that we deem appropriate. As always, if it’s a complex business system, SABSA has the intellectual tools to analyse the risks, both opportunities and threats.