Once upon a time, artificial intelligence (AI) was all science fiction and no fact. Now that we are experiencing products and services that are enabled and enhanced by AI, we are starting to appreciate its capacity to significantly change how we live and work – and to consider some of the actual and conceivable implications of this.
Public perception changes as we become more familiar with concepts such as AI and machine learning (see ‘AI concepts and categories’ below); but predictions about their potential pros and cons in the future span a bafflingly broad range.
At one end of the spectrum is tech entrepreneur Elon Musk. ‘Mark my words, AI is more dangerous than nukes,’ is just one of his dire warnings about all-singing, all-dancing ‘general AI’, as opposed to the kind of functional, narrow AI that’s used in his Tesla cars (and other current AI applications).
AI concepts and categories
Understanding the potential of AI means also understanding some of its concepts, sub-categories and techniques, such as machine learning, natural language processing (NLP) and predictive analytics.
AI is the theory and development of computer systems that can perform tasks that normally require human intelligence, such as decision-making, language translation and speech recognition.
An algorithm is a set of rules or a sequence of instructions that are followed to complete a task.
Machine learning is an application of AI that uses an algorithm or model to process data, identify and learn from patterns in it, predict similar patterns in new data and use this to improve its performance.
Examples of this include Pinterest (content discovery) and Twitter’s curated timelines.
Deep learning is a subset of machine learning, where artificial neural networks (algorithms inspired by the human brain), learn from large amounts of data. Examples include PayPal, which is using deep-learning fraud-detection algorithms to monitor transactions and identify suspicious behaviours.
NLP facilitates human and computer communication by recognising and responding to nuances in human language. Examples include Baidu (search engine), IBM Watson and Amazon Alexa.
Predictive analytics are used by programs to analyse historical data in order to predict future outcomes. They are often combined with AI techniques. Examples include: American Express (fraud detection) and Einstein Analytics from Salesforce.
Many products and services utilise a combination of multiple AI techniques and tools.
At the other end of the spectrum is Steve Wozniak, co-founder of Apple. He used to share Musk’s forebodings, but in 2018 he declared: ‘AI doesn’t scare me at all,’ because a two-year-old child only needs to see a dog once to always recognise one and a computer can’t get near that until it’s seen a dog over and over again.
Dame Wendy Hall, computer science professor at the University of Southampton, and an expert on AI, has a more balanced perspective on its future. ‘There will be lots of positive benefits. But we need to get a grip of the downsides,’ she says, because change is happening very fast. AI technologies such as natural language processing (NLP), machine learning and machine processing are already being used to improve processes, enhance interactions, solve problems, perform functions and make decisions that used to be the preserve of humans, and Hall says we can expect ‘escalation and acceleration’.
AI will bring lots of positive benefits. But we need to get a grip of the downsides, because change is happening very fast
Here and now
Only time will tell what AI is capable of. Meanwhile public and private sector organisations across all industries are jumping on the AI bandwagon (see ‘The AI gold rush’, below); implementing solutions that AI makes possible today and exploring what it could make possible tomorrow. The really big AI successes, however, may be concentrated among the biggest online service and storage companies such as Alibaba, Amazon, Google and WeChat, because they have a head start and an inherent advantage – the vast amounts of data they are collecting.
Although AI raises potential ethical concerns for many professions, accountants consider ethics through the prism of the code of ethics of the International Ethics Standards Board for Accountants (IESBA). ACCA’s report Machine learning: more science than fiction considers how ethical challenges around AI may challenge or compromise the profession’s core principles of integrity, objectivity, professional competence and due care, confidentiality and professional behaviour.
Accountants can bring this perspective on ethics to some of the wider debates taking place on issues around algorithms, machine learning and so-called ‘black box’ systems. The profession may be well placed to develop some of the assurance frameworks that may eventually be needed to demonstrate that organisations developing and using AI are doing so in accordance with the necessary ethical principles – if consensus can be reached on these.
Dame Wendy Hall, computer science professor at the University of Southampton, who is a leading figure in the UK on the development of AI technologies and the associated ethics, says: ‘It is very hard, once these algorithms are let loose, to unpack exactly what they are doing.’ Perhaps those developing and using AI should be in some way accountable. Hall says: ‘All companies should be aware of their responsibilities in this area and of the ethical issues in what they are doing.’
‘Data is the key raw material that feeds machine learning algorithms,’ says Narayanan Vaidyanathan, head of technology insight at ACCA. Massive growth in the volume of that raw material is one of the keys to recent and coming AI advances. ‘As a civilisation we are producing lots more data than we have in the past and our processing and computing capabilities are also expanding like never before. These things combined mean that use of tools like machine learning is poised for significant take-up in the future, because we have raw material and the ability to process it,’ he explains.
Access to that all important raw material can be somewhat uneven. AI pioneers such as Amazon and Google have always valued the vast amounts of data we have willingly ceded to them and they’ve spent years preparing for the transition from micro to macro-level applications of AI and machine learning. ‘The last 10 years have been about building a world that is mobile-first, turning our phones into remote controls for our lives. But in the next 10 years, we will shift to a world that is AI-first,’ wrote Google CEO, Sundar Pichai, in a 2017 blog.
Pichai predicted a world where ‘computing becomes universally available – at home, at work, in the car, or on the go – and interacting with all of these surfaces becomes much more natural and intuitive, and above all, more intelligent’. This shift appears to be well under way, as AI is trickling into more and more areas of our personal and professional lives. There are chatbot educators, while lawyers, therapists and finance professionals are interacting with AI applications in specialist areas as diverse as audit, the delivery of financial services, financial close processes and fraud detection.
ACCA explores some early stage AI applications in its new report Machine learning: more science than fiction. ‘It offers an introduction to machine learning for professional accountants,’ says Vaidyanathan. The report outlines what machine learning is, shares current thinking on its use, considers ethical implications from the professional accountant’s perspective (see ‘Ethical dilemmas’, above) and explores how machine learning developments will influence future skills for the professional accountant; and all of this is underpinned by primary research with around 2,000 ACCA members across the world.
‘This research will provide some reality on what accountants are currently doing and adoption they are seeing in their organisations. There are lots of insights from the profession,’ says Vaidyanathan. The report explores how finance professionals feel about machine learning, its influence on interactions between accountants and technology, and emerging issues such as explaining how machine learning algorithms make judgments, avoiding bias in data sets or algorithms, algorithmic accountability (see ‘Challenges and opportunities’ section, below), and ensuring the provenance and veracity of data.
Roles for the profession
‘Data can only be used to create insight if you have clean data that has been validated and properly managed,’ says Vaidyanathan. This presents an opportunity for the profession. In many organisations, CFOs, FDs and other senior finance people have a control responsibility in terms of managing the governance of the organisation and its structure and processes, there is already an overlap with existing technology resources and the associated data, and as the amount and value of data increases, so will the involvement of finance professionals.
As data becomes ubiquitous, so may the accountancy profession. ‘We know that accountants are constantly expanding their field of vision,’ says Vaidyanathan. They have access to data from across the business and in many organisations this is not just finance-related data. Finance professionals are also working with operational data and getting more involved in processes that reflect the changing nature of strategic and corporate reporting, for example, by broadening their scope to encompass processes that feed into environmental, social and governance (ESG) reporting and integrated reporting more broadly.
Professional accountants can add value by bringing their professional scepticism and ability to interrogate, and having oversight of what the algorithm is doing
In the brave new world of data-enabled AI, members of the profession bring some specific and very valuable skills to the table. ‘If you want to get insights from the data that add value, you need to understand where you are going as a business and you need to link what you are doing with the data with where you are trying to go as a business,’ says Vaidyanathan. ‘This is something many accountants already excel at and a new final case study module “Strategic Business Leader” was recently added to the ACCA Qualification to emphasis the importance of this area.’
As AI changes the business environment, professional accountants will play an increasingly important role. ‘There will be a much higher premium on the ability to get out there and really understand the priorities and risks within the business,’ he says. Accountants will understand and communicate how an organisation’s strategy, financial and non-financial information interact with each other to create a picture of value creation and direction for the company. ‘All of this is essential when you are dealing with machine learning algorithms, because they don’t have any of that wider context,’ explains Vaidyanathan.
The AI goldrush
All sorts of organisations are investing in all sorts of AI: ranging from systems that can converse with a human to those that can perform better than a human – and it’s not just AI pioneers such as Amazon and Google that are mining for gold.
Billions of dollars, euros and yuan are being invested by private equity firms and AI startups, corporates that want to beat them and their competitors to the benefits, and local and national governments jockeying for position as the leading country or region for AI.
Big tech has a head start. AI already powers many Google products and services including Google Assistant, Duplex, Maps and Search. This provides it with massive amounts of data to feed the machine learning algorithms that are key to its success today and its strategy for tomorrow. Google’s decision to share its machine learning platform TensorFlow with an Open Source community is not an act of altruism. If you are not paying for the product then you are the product.
Entire countries are now trying to play catch-up. China has declared its intention to oust the US as the world leader in AI by 2030 – and has committed billions to the pursuit of this aim. China’s giants Baidu, Alibaba and Tencent (often abbreviated to BAT) are investing heavily in research and development; and the country has a massive advantage over many others in the race for AI gold, because it has very few obstacles to data collection and regulations on its usage.
An algorithm can do lots of clever computation but you need business knowledge to ask the right questions and interpret the answers. Accountants can also play key roles in addressing the control and governance issues that are emerging around machine learning.
‘Professional accountants have the potential to add value in terms of bringing their professional scepticism and the ability to interrogate, and having oversight of what the algorithm is doing,’ says Vaidyanathan.
ACCA and Alibaba
ACCA and Alibaba Cloud Computing have signed an exclusive agreement to see a closer working relationship for the two organisations to focus on course development, research and professional insights. The agreement sees both organisations working together to shape the future of the profession during a time of digital transformation. ACCA chief executive Helen Brand said: ‘Our collaboration will be broad – from producing joint research to looking at course development for ACCA members about digital innovations.’
Challenges and opportunities
The spread of machine learning algorithms raises a host of thorny questions on accountability. Professor Karen Yeung, interdisciplinary professorial fellow in law, ethics and informatics at Birmingham Law School, in the UK, says there are difficult questions to be asked about the distribution of authority, responsibility and liability, and who is held accountable if there is harm. ‘The fake news debate is a great example of how there are real tangible consequences from using these systems, yet we have no real effective way of governing those,’ she says.
Although many regions are strengthening their data protection legislation, AI and its utilisation of data is creating new issues. Yeung says: ‘The European Union has been the world leader on the regulation of automated decision-making.’ There are some mechanisms in its General Data Protection Regulation. However, meaningful ethical regulation of AI systems will be more difficult to mechanise, not least because AI components and data from multiple jurisdictions are being built into products and services. ‘Grappling with these questions is the Wild West; nobody really knows what data ethics is.’
The power of digital
ACCA has been focusing on how digital developments are changing the accountancy profession, highlighting the key impacts, offering learning and development opportunities, and evolving the ACCA Qualification to ensure it remains cutting edge. You can explore our professional insights research, content and opportunities on the ACCA website. During May 2019 we’ll be introducing a new CPD course on robotics based on our joint ACCA/CA ANZ research on the topic, along with a new microsite which will bring together all of our learning and resources on digital.
More data seems likely to bring more problems. ‘Every reliable estimate suggests that the amount of data we create is going to increase exponentially not linearly,’ says Vaidyanathan, and the abilities of machines are growing along with data volumes. At some stage there may be an argument for breaking up some tech giants or legislating to curb their emerging monopoly on data – before their AI-enabled automation becomes ubiquitous in every walk of life. If we want to enjoy the benefits of AI we need to deal with some of the burdens, and fast. Because as Hall observes: ‘The genie is out of the bottle.’