Artificial Intelligence: Could robots take over your job?

If you have used a computer, smart phone or browsed the web before, you have probably interfaced with Artificial Intelligence (AI) more than you think you have. AI is simply the term used to describe technology that ‘thinks’ and ‘learns’; also termed as acting humanly.
From predictive texts in our smart phones, to intelligent home systems like Google Home, to personalised advertisements, AI has the potential to transform life, and revolutionise customer experience. For businesses, the opportunities scale further beyond, as solutions like Robotic Process Automation are enabling innovation and increasing productivity.

How AI will shape tomorrow’s business
With the rapid evolution of AI and the enthusiasm demonstrated by both governments and large corporations, it is nearly impossible to give a precise vision of what the AI space will look like in the coming years. PwC’s “2018 AI Predictions: 8 insights to shape business strategy,” gives us an idea of what the near future will look like.
Businesses today are experiencing what would best be termed as the “data dilemma”; with tremendous amounts of data sitting in data warehouses and data centres and yet no value is being gained from this data. PwC’s Global Digital IQ Survey found that few businesses get value for their data, with only 26 per cent in the financial sector agreeing that they utilise all the data they capture to drive business value. The public sector and power utilities ranked highest at only 43 per cent and 37 per cent, respectively. AI will open new opportunities for use of this data. Use cases include building and training predictive models, analysing and benefiting from customer habits and management decision making.
It will interest you to know that 72 per cent of executives are already looking to AI as the business advantage of the future. The future will be shaped, not by having the right information but being able to turn that information into actionable plans.
However, time is just not enough for such value-adding analyses as staff are bogged down with routine tasks and time-consuming information gathering activities. The value of AI lies not in creating entire new industries but rather in empowering current employees to add more value to existing enterprises; by removing tasks which do not genuinely require human intervention from the trays of employees to software robots programmed to constantly apply a level of human-like reasoning to get the tasks done. AI will not only take over such tasks, but will also perform them with precisely, making information more accurate.
AI will inevitably become a necessity to business just as computers are today. Also, inevitably, some human positions will become redundant. However, new positions will in turn be created.
The 2017 research by Gartner predicts that AI will create more jobs than it will eliminate. By 2020, AI will generate 2.3 million jobs exceeding the 1.8 million it will wipe out and in the following five years, the net new jobs created in relation to AI will reach 2 million. There is no doubt that AI will encourage a gradual evolution in the job market and with the right preparation, this will be a positive transition.
Malware and ransomware will soon become more intelligent: learning as it spreads, using machine intelligence to coordinate global attacks and data analytics to customise such attacks. The tools that will be used in fighting future cyberattacks will need to be equally intelligent. In our 2018 Global Economic Crime Survey, almost 30 per cent of Ugandan respondents who had suffered crime from external fraud cited hackers as the perpetrators with malware and phishing as the most prevalent at 39 per cent and 31 per cent, respectively. Organisations will consequently turn to AI powered solutions to fend off cyberattacks. Cybersecurity will no doubt be many enterprises’ first step towards embracing AI.
AI itself, if not well-protected, opens organisations to a completely new level of vulnerability. A quick example but sensitive area is in algorithm training sets. AI algorithms and models work best when exposed to various datasets in a process known as ‘training’. Any bias in such data will lead to inaccurate solutions or predictions. Imagine malicious actors injecting biased data into AI models such that results from such models are inaccurate or skewed to favour a certain group. A good example would be a flawed predictive model for approving loans in a bank, which approves some customers who do not qualify and disapproves those who qualify. Such attacks can lead to reputational damage, loss of business and erroneous investments. It is, therefore, of utmost importance for any organisation seeking to adopt AI to establish a sound AI driven cybersecurity agenda.

AI in businesses
We are in the information age where businesses thrive on their ability to generate, analyse and make good use of the right information. AI comes at the helm of such endeavours. AI will be the key technology in building informed customer attraction and retention strategies, in deciding best candidates to recruit for a position, and in making investment and management decisions; among others.
The agriculture industry is one of the sectors that will be impacted by AI in a significant way. AI powered solutions are coming up, such as the South African hailing Aerobotics drone system, which assists farmers to analyse processed maps and extract actionable information to identify problem areas in crops.
Similar pilot schemes are being tested in Uganda in the coffee industry. AI powered Farmer to consumer technologies will make it easier for farmers to access diverse markets for their yields.
The health industry is also changing considerably. The 2016 research by YouGov established that 54 per cent of individuals were willing to engage AI and robotics for their healthcare needs. Solutions such as intelligent healthcare assistants and robotic surgeons will shape tomorrow’s healthcare and AI will be the enabling technology.
African businesses stand to reap from AI as the majority of the African population is young and comprise the early adopters of AI.

When is the right time?

Leaders do not need to adopt AI for AI’s sake. Instead, when they look for the best solution to a business need, AI will increasingly play a role. The best time to start thinking about AI is now.
Organisations looking to incorporate AI into their operations should start by buying (instead of developing) proven AI solutions. This gives the advantage of leveraging on years of research done by the vendor and reduces expensive mistakes. Starting with small projects that would benefit from cognitive technology and learning from them will make the AI agenda more affordable and effective. Organisations can start by making existing applications and offerings more intelligent and more autonomous.
Experts have warned that AI requires a long-term mindset and businesses may not expect to reap immediate direct results from AI. Short-termism is therefore discouraged and a long-term mindset should be incorporated in the AI strategy.

The writer is an associate in risk assurance at PwC Uganda