For more than a decade, the rate of digital transformation has fuelled a perpetual supply-and-demand issue. The need for people with skills in data analysis, artificial intelligence, programming and machine learning has far outstripped the number of people with such capabilities.
Yet in a government report from May this year, which looked to quantify exactly how large this skills gap is in the UK, it made an interesting discovery. It’s not that there aren’t people interested in entering this field, nor is it that people are lacking the aptitude for such technologies and careers, it’s that they just don’t know where to start.
It found that 39% of people thought the path to AI careers was unclear, and one in eight (12%) had no idea about it. Women were more likely than men to lack clarity, as were younger people and undergraduate students versus those studying for master’s and PhDs. And one of the biggest issues surrounding this confusion? The myriad terms used to describe the different roles and skills, and which of these skills employers actually want.
The changing landscape of AI
There isn’t any part of our lives that isn’t touched in some way by artificial intelligence. It manages to be both the most ubiquitous, talked about technology around while also being one that few of us truly understand. Although there isn’t a simple definition for the term, it generally refers to the use of data and creation of machines designed to perform and automate tasks associated with human intelligence – such as learning, analysis, and problem solving.
This can be as high end as the algorithms that power self-driving cars or put NASA Rovers on Mars, down to Amazon and Netflix’s ability to recommend shopping and shows based on your habits. This vast range of uses and ubiquity means there are actually many paths into a career in AI. And not all of them are what you might expect.
Since 2015, demand for AI roles such as machine learning engineers has risen by 344%, while demand for data scientists has jumped 78%. UI and UX designers fall under the category of AI careers, as do full-stack developers, software engineers, big data engineers, research scientists and more. More recently, roles including process automation specialist and data ethics officer have emerged and are growing in demand.
As you’d expect, these roles require a set of hard skills in order to be able to handle and analyse data and program the software, yet the level of technical knowledge varies depending on the role. Plus, these roles are increasingly relying on a mix of other softer or more general skills that many people looking to get into a career in AI may not realise they already have.
In the government report in May, employers were asked which were the most important skills that their companies needed to improve on. Their top 10 wishlist was as follows:
- Information management
- Knowledge of emerging technologies and solutions
- Data communication skills
- Communication
- Database management
- Data literacy
- Data ethics
- Analysis skills
- Analytical mindset
- Adaptability
What’s more, the top five priority skills businesses want to improve in the context of working with data include communication, professionalism, problem-solving, data ethics and then basic IT skills.
As both lists attest, knowledge and communication skills are now up there with the technical, day-to-day know-how. Getting into a career in AI in the early days largely relied on knowledge of very precise programs. Getting into a career in AI in the years and decades to come is going to more heavily lean on the interplay between humanities and science.
A shift that Professor Brian Ball, Head of Faculty in Philosophy at New College of the Humanities (NCH) and Senior Lecturer on the college’s MA Philosophy and Artificial Intelligence course isn’t surprised by.
“When speaking to people in industry, there is a very real and growing need for people with technical AI knowledge to also have the types of skills humanists are renowned for – the ability to think about the wider impact of technology, to be able to communicate with teams of engineers and managers about these impacts effectively,” he tells us. “It’s common to find people in organisations who have the education and technical skills to be an AI engineer, for example, yet can’t communicate with people in their own team, company or their clients.”
He also adds that the concept isn’t exactly new. “There has been a longstanding connection between philosophy and computing – the two have been intertwined since the early days. Formal logic in philosophy about machines being able to make rational calculations some 250 years ago ultimately led to the advancements made in Alan Turing’s work.”
Taking an interdisciplinary approach
Reports suggest there are between 178,000-234,000 data-related roles in UK companies to be filled yet internal analysis of Higher Education Statistics Agency data estimated that the potential supply of data scientists from UK universities is unlikely to be more than 10,000 per year.
This skills gap, and need for an interdisciplinary outlook was a driving force behind NCH’s MA Philosophy and Artificial Intelligence course, and has more recently been the influence behind the college’s latest course in this space, MSc Artificial Intelligence with a Human Face.
Both courses combine exploring philosophical issues, including ethical issues related to computer data, with opportunities to develop the skills and techniques of data science. For the MA, the focus is weighted more on the former, teaching students with no background in either philosophy or programming how to code, while also equipping them with the combination of thinking, communication, and technical skills currently sought by employers. For the new MSc, more weight is given to developing the coding skills, while students are introduced to the ethical and cultural impact of the technology and philosophical thinking surrounding it.
“We’ve already seen some of the potential consequences on society as we increase our use of data,” says Ball. “In 2011, the general consensus was the internet was amazing and will help democratise opportunity. Fast-forward to the 2016 referendum and the questions raised by the Cambridge Analytica scandal, or the 2016 presidential election, or the way social media was said to play a vital role in the Arab Spring, and official bodies are very concerned about the threat this once-heralded democractic tool could have on democracy as a whole.”
There have also been a number of issues concerning bias raised by our increasing use of data and AI. For all the benefits of AI, the programs are only ever as good and fair as the data they are fed. An Amazon program, for instance, that used AI to filter job applicants discriminated against women because it was trained on resumes from a predominantly male workforce. An algorithm used in US court systems to predict the likelihood a person would reoffend was found to predict twice as many false positives in black offenders versus white.
“All technical decisions have consequences. Twitter’s recent decision to increase characters, a change that was seemingly innocuous, had a huge impact on public discourse. This, in turn, has had a huge knock on effect with ethical political consequences.” All because a figure in a line of code was changed from 140 to 280.
How to kickstart your career in AI
Given the limited supply of graduates likely to fill data roles, the government proposes that upskilling the workforce is vital to bridging the data skills gap and that there is a need to develop a suite of skills across both hard and soft skills – a mix of skills that may make the path into a career in AI more appealing, or seemingly more accessible to a wider number of people.
“These soft skills are hugely important – without them there is the potential for data to be misread or miscommunicated, which can have significant implications for businesses and the decisions they make,” said John Whittingdale OBE, Minister of State for Media and Data.
If you’re interested in a career in AI, instead of looking at which role(s) in particular you think would best suit you, or trying to navigate what each of them do, look at your skillset and aptitude. Identify which of these hard and soft skills you already have, and which you feel you’d need to develop to take this step. This will make it easier to determine where to start and which path you’ll need to take to get there.
