Artificial intelligence (AI) is gaining a lot of media attention in recent times. Stories were written about how computers will soon be able to mimic human intelligence.
Google demonstrated how its AI can carry out a phone conversation without the human listener being aware that it was a machine. Apps today can create realistic photos of humans that do not exist – a promising way to save paying real-life models in advertising! IBM’s DeepMind has beaten humans in complex games like chess and Go.
Last year, 15 top doctors in China competed – and lost – against the BioMind AI system in a competition to diagnose brain tumours and predict the expansion of brain haematoma.
Recently, IBM’s AI competed in a debating competition. Although it lost, it is significant how the system can comprehend the context of the debate question, scour through various databases to extract relevant information and put together a coherent argument.
The AI Controversy
Stories of machines turning against humans are plentiful in movies. Even prominent people such as Elon Musk and the late Dr Stephen Hawking have voiced their concerns that AI could lead to the end of humankind.
Commercial-level AI is far less sophisticated as technology is not that advanced yet. The AI breakthrough stories are mostly proof-of-concept or pioneering research work. Even large corporations are not developing in-house AI for the sake of having AI: technology is developed to meet business objectives. These corporations may also engage third-party developers who specialise in certain aspects or components of AI, so that the end-result is a combination of smaller pieces working together.
The pace of AI development accelerated rapidly in response to the massive volumes of data generated today by mobile phones, websites, social media postings, sensors and various Internet-of-Things devices. Businesses recognise that such data hold the key to improving efficiency, reducing cost and increasing revenue.
AI at Work
One of the first benefits of using AI is obviously the speed at which it processes information. It can be used to process office paperwork – files, minutes of meetings, emails, memos, archived data – and produce comprehensive reports to help users make better-informed decisions.
In customer service, AI can be used in chatbots – automated response systems – that is available 24/7 and can offer personalised interaction. In their simplest form, they can answer frequently asked questions automatically. These could be “What is my credit balance?” for credit card companies, “Do you have this item in stock?” or “What time do you open/close?” Chatbots can also sort out request tickets and forward them to the relevant departments.
AI development requires expensive and complex infrastructure and very specialised talent. Companies like Alibaba, Amazon, Facebook, Google are at the forefront of the technology because they can afford it.
It is better to “buy not build” applications that are based on AI. For example, SMEs can engage Facebook and Google to target advertisements at specific groups based on their likes, location, gender and other interests. The ads go to the people who are most likely to respond.
Hackers are deploying increasingly sophisticated methods to attack their victims. Obviously, SMEs are easy targets as they lack the tech resources for strong security. They are especially at risk if they accept online credit and debit card payments.
AI-driven software address specific business requirements. Stripe uses adaptive machine learning to detect fraud in credit and debit card transactions. FirstHive can track customers behaviour. Salesforce uses Einstein AI on its customer relationship management application to analyse phone conversations, emails, social media posts and customer reviews and report on consumer sentiments. It gives feedback on how to adjust marketing campaigns and lead generation.
Acquisio can manage and track marketing operations across various channels such Adwords and Facebook. It can analyse how the ads perform and offer suggestions for ad distribution to get the best pay-per-click results.
Crayon can track a company’s activities across online channels, to detect price changes, shifts in messaging and PR activities. Market intelligence apps can be invaluable to an SME seeking a clearer view of its competitor’s strategies, and to identify product gaps, strengths and weaknesses.
Where do with Stand?
Despite the benefits, AI is still in the early stages of development. They can optimise processes but not solve every problem. A company does not become more profitable simply by adopting AI.
The main consideration in determining whether AI is needed, is ROI. Choose software that can meet your business goals – if it uses AI, consider it a bonus. Do not choose AI for AI’s sake.
AI may become more mainstream, more intelligent and easier to use in future. Meanwhile, SMEs should focus on their core business and improve how they collect and store data. Invest the effort and money to set up a proper data infrastructure, to be AI-ready.
This may involve working with a managed services provider, paying them to set up and manage the hardware or putting it in the cloud. By using a “pay-as-you-go” model, SMEs may save costs of buying, maintaining and upgrading hardware, securing the environment from hackers and hiring technical staff.
AI may not be sophisticated enough to mimic humans or threaten its existence, but there is plenty enough of it that can help SMEs gain better business intelligence.