For years, AI has been slated to disrupt, and ultimately revolutionise, just about every industry. There are regular stories about which professions will cease to exist completely, as AI will be able to do their jobs more effectively for less, and entire genres of media exist describing AI-dominated futures, both as fiction and real-world predictions.
What we haven’t seen yet is disruption that even comes close to this scale in any market sector. AI is increasingly being integrated into consumer goods such as digital assistants, and self-driving vehicles are clearly on the horizon, but this is far from the all-permeating disruption that has been promised.
Some real-world business applications of AI are starting to emerge, though. The question is whether the technology is mature enough to be worthwhile, and what the potential risks to investing in AI amount to. In short, is 2018 the year AI really comes of age for business?
While there hasn’t yet been wide-scale worker replacement by AI, there are some potentially business ready systems that can replace humans, cutting down costs for smaller businesses, or potentially creating efficiencies for any business.
On a cost-cutting level, there exist AI ‘bots’ in two real forms, chatbots and assistants, that can provide some degree of business benefit. Chatbots can be employed quite reasonably in a first line support role, directing employees or customers to the appropriate knowledgebase pages without needing human intervention, or passing more complicated matters on to human support workers. Second are AI assistants, which build on home automation technology to carry out simple secretary-like functions, such as providing reminders, arranging meetings, and carrying out some simple communication.
Also available, and probably far more potentially useful, are more fully-formed neural networks which can be used to sift through vast quantities of data in a way that a human simply could not. These AIs are not yet available in the ‘off the shelf’ variety that bots are, but they have far more potential to increase efficiency and create disruption.
There are risks involved in adopting any technology early, and AI is no exception. For instance, a bot may provide cost savings in some support situations, but you run the risk of putting customers in a position akin to trying to navigate a 1990s-era telephone voice recognition system – something that no-one should be put through if it can be avoided.
Further risks involve the ‘learning’ aspect of machine learning systems. These systems will not be ‘ready to go’ instantly – they will need careful monitoring and double checking for some time, before they can be let loose on a task with predictable and usable results, meaning the investment may take some time to ‘mature.’ Following on from this, an improperly ‘trained’ AI, or one that hasn’t been checked carefully, may give back largely incorrect information while believing it to be correct, potentially leading to substantial embarrassment or loss if unnoticed.
So whether or not AI is mature enough is still a bit of an open question, but it may even be worth simply getting used to working with AI systems, as there is no doubt that they are the future.
Sources incl: Gartner, Forbes, Beebom.com, Nasdaq.com, Coindesk.com
