Does Artificial Intelligence lead to the data revolution?

In recent years, artificial intelligence (AI) has become more widespread in both the public and private sectors. As data is collected from multiple sources and analysis becomes increasingly complex, it seems that those employed to identify trends and meaning are less able to keep up. Indeed, in its Tech Trends 2017 report, Deloitte identifies an exponential growth in data that is taking place, asserting that the digital universe is doubling in size every 12 months. The available human resource qualified to analyse this ever-increasing volume, then, is falling shorter as each year passes.

Many businesses are therefore looking to AI – and more specifically, Machine Learning (ML) – to help with big data management and analytics. Machines with the ability to reason and act could mean much faster processing and analysis than humans are capable of achieving, which in turn indicates the potential for a significant shift in the way data is handled and business is conducted. Here, we take a look what the benefits and risks of ML are, how it is currently being used and if it will really lead to a data revolution.

The benefits of ML in relation to data

The very nature of Machine Learning is suited to big data: with high volumes and complexity readily available, the algorithms can improve themselves quickly and efficiently. This has exciting implications for businesses in an array of sectors, from retail and telecoms to banking and finance.

1. Faster analysis

As mentioned above, ML enables advanced data storage, retrieval and analysis. Their capacity to improve simply through contact with data means the systems no longer need to follow programmed instructions, and can identify patterns in much the same way as a human might. Although the goal of ML is to mimic the human thought process, its own processes for retrieval and analysis and trading risk management are much faster, allowing patterns to be recognised and acted upon more quickly.

2. Accurate recommendations

With fast analysis comes an ability to make increasingly accurate predictions and recommendations, whether internally or to customers. Upselling and cross selling are a key part of sales, and unsupervised ML enables the discovery of obscurer behavioural or purchase patterns, which can be turned into sales opportunities. This inevitably creates a better customer experience and has the potential to positively impact business revenue. Such ML application is already being seen on sites such as Amazon, in their customer recommendations function.

3. Frees up employee time

Another advantage of ML is that it lifts time-consuming tasks from the shoulders of fintech employees, enabling them to focus on more important jobs. This is already proving beneficial for financial institution JP Morgan, whose programme interpreted commercial loan agreements in a fraction of the time it previously took the firm's lawyers and loan officers to decipher. Not only can ML speed up the mundane work that consumes hours of employee time, it can also reduce the possibility of errors occurring.

This isn't only true for financial services. In retail, using ML to automate processes such as order placement, transaction lookup and answering basic customer queries can free up a huge percentage of staff time (up to 80% according to deepPIXEL), resulting in greater value for both customers and companies.

Risks of ML in relation to data

As with any technology, the use of Machine Learning in the analysis of big data comes with certain risks. Businesses will need to consider if and how they wish to mitigate these when investing in AI technology to enhance their productivity.

1. Data protection

Perhaps the biggest consideration when it comes to big data and the use of ML for analysis is the protection of personal data. This is particularly pertinent for European businesses, who will be subject to the more stringent General Data Protection Regulation (GDPR) from May 2018.

Using ML means the data held by companies can be repurposed and conclusions can be drawn about individuals that may not have been possible previously. This could in turn lead to unwelcome or discriminatory services. In a recent report on big data, for example, the FTC identifies an instance of people's credit limits being reduced because others who shopped at the same stores had poor repayment histories.

Firms collecting and storing large amounts of data should therefore consider the effect of inferences made by ML on the individuals and social groups concerned. Such consideration could be made in the form of a privacy impact assessment or by combining ML with human input to avoid unfairly targeting or impacting people.

2. Accumulation of technical debt

As systems that do not self-replicate or self-optimise, ML systems are prone to the accumulation of technical debt that cannot be solved in the traditional ways. Hidden debt can build up due to the complexities at system level, which can result not only in a series of undesirable outcomes such as entanglement and undeclared consumers. This means that it's surprisingly easy for system maintenance costs to build up.

Businesses looking to invest in ML must therefore understand the implications of technical debt and weigh up the risks of accruing moderate levels of such debt in order to progress. Employing competent software engineers, technical staff and data scientists is necessary to mitigate this risk.

3. Changes in the external world

Machine Learning is intriguing for its ability to interact with the external world. This world, however, is ever-changing and can become a source of technical debt. As an example, if a decision threshold is set manually and an update based on new data occurs, the existing threshold may become invalid. This might be manageable for one model and one threshold, but if there are many, manually revising them can prove time-consuming.

A data revolution?

Many businesses will be reviewing the pros and cons of ML in the coming months: Deloitte's 2016 Global CIO Survey showed that 64% of its respondents were looking to invest significantly in what the report calls 'Machine Intelligence' (MI) during 2017 and 2018. This suggests a more rapid uptake of the technology across a number of sectors as businesses compete to stay ahead.

Equally, the ICO is viewing the increase in ML usage for 'business as usual' as a step change in regards to data protection. In a report on big data, ML and data protection, it discusses the potential difficulties around applying conventional data regulations to big data analytics and the need for new procedures when it comes to protecting personal data. This indicates that regulatory bodies expect something of a transformation in the way data is used by companies on a daily basis, and points to an imminent, large scale change.

Indeed, it seems that a revolution has already begun in some sectors. We mentioned above how JP Morgan dramatically reduced the time in which commercial loan agreements were interpreted by using a programme called Contract Intelligence, or COIN, but there are a number of other examples of industries using ML to improve processes.

In the manufacturing industry, firms are collecting larger amounts of data on their machinery and production processes, and using predictive technologies and analytics organise the service schedule. By being able to predict when equipment might start to fail and why, manufacturers can ensure it is serviced and potential issues rectified before losses are incurred. Of course, this results in greater efficiency and bigger profits.

Meanwhile, ML is helping biological research by finding patterns in the huge amounts of data produced by experiments, and facilitating the formulation of scientific hypotheses. One of the advantages of the technology is that it is free from a bias towards the familiar, which opens up the possibility to find answers that humans may not look for. It is with the help of ML systems that some biological scientists are looking for new therapies to treat cancer. Data on normal and abnormal protein codes that would take humans months to go through can be analysed in far less time by an algorithm, thereby speeding up the process of finding correlations.

These examples suggest an already burgeoning use of ML in relation to large data sets. With the abilities of systems and technologies evolving all the time, the data that can be stored and used by businesses has the capacity to increase, with more complex findings and predictions coming from varying sources.

Conclusion

Although Machine Learning has its risks – not least those rooted in personal data protection and the accumulation of technical debt – the systems and technology that it uses are evolving all the time. As more businesses invest in it, ML systems will become more effective and 'smarter', enabling organisations to make progress in ways that haven't previously been possible.

Current examples of ML usage show that it is already beginning to revolutionise the way businesses handle and analyse data, and the kind of actions they can take as a result. For European companies, the possibilities may be tempered – or at least, regulated – by the introduction of GDPR in 2018, but it wouldn't be unreasonable to suggest that a large scale change in data management will occur across a number of industries as ML is adopted.