Artificial Intelligence (AI) is developing fast and its application across industries could, and arguably should, be about to reshape the way we structure our approach to labour and the workforce. Given this proposition, we need to examine exactly what so-called ‘machine intelligence’ in the shape of AI could bring to modern business processes.
The World Government Summit has tabled precisely this kind of question across its futurist and information intelligence tracks this year, so who benefits first and which industries could put AI to work hardest?
The important thing to remember about AI is that it can eradicate ‘donkey’ work. Or to put it in more refined terms, AI can help reduce some of the base-level functions and roles that every business structures itself around. From construction to leisure to food production, every operation has a certain number of underlings, trainees, support staff – jobs that require few skills. This is one of the key realities that AI impacts.
1 – The legal trade
The legal trade has plenty of juniors. Not everyone in the law profession is a high-ranking case lawyer, judge or governmental council superior; and while many in the lower echelons of the business may have fancy job titles like ‘paralegal’, their work can be very tedious. It is these poor souls who spend hours examining hundreds of letters, deeds, case materials and supporting documentation. The job is perfect for AI automation if we can direct software ‘e-discovery’ tools to the data we need to process faster.
With today’s legal information topography also including emails, video and even social media streams, the job is arguably too large for a human brain. Applying AI to the donkey work or grunt work means fewer headaches, more speed and improved consistency of results and who wouldn’t want that?
Even the paralegals quite like the idea as they can start to work higher up the food chain. Some will inevitably be made redundant, but that is the price of progress.
2 – Advertising
Firms including Intel, German data crunching specialist Software AG, IBM and others have been attempting to apply AI to the advertising, marketing and promotions business for some time now.
A central manifestation of this technology has been seen in the production of camera-enabled electronic ‘posters’ used to display advertisements. At its most basic: when the camera detects a male in front of it, it will projects a car advertisement perhaps or when the camera detects a female, it can present an advertisement for perfume.
Inappropriate gender stereotyping aside, this is obviously great news for advertising campaigns if they can be made more ‘live targeted’. Immediately we can envisage privacy concerns here of course as the camera will is able detect more than a person’s sex. Approximate age, ethnic background and even mood is also trackable.
But do we want our computers knowing this much about us? Possibly yes, but we will have to work with guidelines that we lay down now while it is still at the embryonic stage.
3 – Financial markets
As logical as it sounds, financial trading is a perfect place to apply Artificial Intelligence. The volume of world financial trades is increasing and the amount of supplementary and ancillary information attached to each trade is also increasing exponentially. Today we know that trading data is about more than current commodity price and strength of the dirham or dollar. Phone discussions, emails and video material relating to the status of any potential trade all go into the mix i.e. this information (as unstructured as it is) is also trackable.
We will now use techniques including sematic text analysis and natural language processing to analyse these data streams so that we can understand them contextually. Working out what data means, in context, is precisely what AI is all about. From here we start to talk not just about decision making, but about evidence-based reasoning and event-based decisions. The stock market just got hotter for sure.
4 – Healthcare
Healthcare AI is not primarily concerned with creating robots to care for sick patients or emotion sensors to act as early warning systems to detect depression, that’s all still in the ‘what if’ prototyping stage.
Instead, we are applying AI to complex human genomic sequencing analysis to look for mutations and so prevent disease. Closer down to Earth, AiCure is a great example of an app that falls into the area known as directly observed therapy (DOT) through the use of smartphones.
The patient simply videos themselves taking their medication and the app uses imaging technology to confirm ingestion. Educational content, real-time feedback and other such incentives are all tailored to the patient’s needs. Yes, you are allowing the AI computer to help monitor you, but essentially, it is for your own good.
5 – Self-driving cars
The still-nascent world of self-driving cars clearly requires machines that are capable of a reasonable degree of intelligence. From motion sensors to spatial awareness cameras, self-driving car ‘brains’ very much fall into the machine-learning and human machine interface (HMI) category of AI.
To be safe on the road, we need to get to a point where the computers driving automated cars are not only aware of defined and predictable physical factors in the world around them but also the undefined, unquantifiable and unpredictable facets of driving caused by the random actions of human drivers who will, for now at least, also share the roadways.
Quite where we get to in the mix of human controlled vs. self-driving cars and how that balances out has yet to be defined.
6 – Nanobots & biotechnology
Not actually a ‘thing’ as such yet and still very much in the developmental stage, nanobots are so called due to their size (a nano denotes one-billionth (or 10-9) so therefore one nanometre is one-billionth of a metre). Nanobots then are miniature robots that we would introduce into our bloodstream in order to reprogram our genes or act almost like some super-intelligent white blood cell that works to keep us healthy.
Nanobots have also been called nanoids, nanite, nanomachines and even nanomites and they are still very much in the prototyping stage. Scientists predict that we might reasonably see nanobots in existence within the next quarter century. Part of what we call the biotechnology revolution, topping up your nanobots could be as normal a thing as taking an aspirin one day.
7 – Government
Last but not least – government. Much has been written about automated government intelligence and major nations such as the UAE have already driven extensive e-government programmes that make extensive use of machine intelligence to power their operations.
The more conceptual (pure theory, if you will) behind AI government sees the human race lay down an encoded and democratically agreed set of rights and principles that would exist beyond the purview of tampering politicians or dictators.
As theoretically perfect as the notion of Über e-government sounds if we were to start building some kind of intelligent self-protecting and self-regulating constitution to manage the bureaucratic functions of government, we do generally come against a problem. Good governmental practice already predominantly exists to be just that already i.e. a self-regulating function that exists for the good of the people.
It is arguably far more likely that we will adopt certain aspects of AI to automate aspects of government rather than completely replace it by any stretch of the imagination. Industry analysts have asked the question: could an AI be the ultimate benevolent arbitrator of governance? The answer is, in the face of human compassion and love, generally not.
The human factor in AI
The wider responsibility that we all share today as we continue to develop Artificial Intelligence should, arguably, not be focused on the computers. We know that technology innovation is rapid, wide-ranging and comparatively unstoppable. Our focus should therefore be directed at ourselves as humans so that we build AI the way we want it to work and stays well within our realm of control.