As artificial intelligence puts our jobs at risk, should we be working with our creations or against them?.

Humans have always feared the power of their own creations. From Frankenstein’s monster to Oedipus, we tend to believe we have the creative potential to destroy ourselves. This seems especially true of mechanical automation. The first great industrial revolution of the late 18th century brought such social upheaval that the philosopher Mr Thomas Carlyle lamented the “demon of mechanisation”, and 200 years later, the digital revolution of the late 20th century was met with similar hand-wringing.

We now look to be on the verge of the fourth great industrial revolution – the true dawn of artificial intelligence (AI) – and once again commentators are warning that we’re “summoning the demon”. This quote doesn’t come from a philosopher, though, but from Mr Elon Musk. When the man behind PayPal, Tesla Motors and SpaceX is warning us off AI, should we take him more seriously?

One way that we’re likely to feel the full disruptive power of AI is in the workplace. The Boston Consulting Group forecasts that a quarter of all jobs will be replaced by automation by 2025, and the Bank of England recently announced that 48 per cent of all human workers will eventually be replaced by robotics. What, if anything, can we do to future-proof our careers?

The term AI may still have the whiff of a Mr Steven Spielberg movie about it, but the truth is robots are already deeply embedded into our daily lives. Whether you’re buying a second-hand concert ticket on eBay, following a scandal on Twitter or asking Siri where the nearest Nando’s is, chances are you interact with a bot on a daily basis.

Investment in AI reached a record £6.8bn ($8.5bn) in 2015 – nearly four times that spent in 2010 – and everyone from Google to Facebook is establishing an AI-ecosystem within their existing services. The McKinsey Global Institute predicts that the transformative power of AI will be 10 times as fast, 300 times the scale and 3,000 times the impact of the first great Industrial Revolution.

These are scary figures, but what they don’t take into account is the shift in employment that new technologies tend to bring about. Certainly AI will replace many existing jobs (more on this below), but it will also create new ones in science, technology, engineering and maths (STEM) that we didn’t need before. It’s worth remembering that in the 19th century, 80 per cent of US jobs were in agriculture. Today, only 2 per cent are. This hasn’t led to the apocalypse (at time of writing).

The BBC recently released a handy online calculator that allows you to assess the risk of automation by job title. As a journalist, I apparently have an 8.4 per cent risk of being replaced by a robot, although I note with some concern that the LA Times already uses a robot for its live earthquake coverage. Unsurprisingly, jobs with a mechanical element, such as train drivers (67.8 per cent at risk) or taxi drivers (57 per cent), are quite exposed, while those requiring emotional intelligence such as nurses (0.9 per cent) and psychologists (0.7 per cent) are the most secure. Creativity also looks like a safe bet, so artists (3.8 per cent) and musicians (4.5 per cent) can relax – for now.

But it’s not just blue-collar jobs that are set to be disrupted. Many roles within law, medicine and finance involve highly repetitive tasks that may be better done by algorithms. Legal secretaries, for example, are 98.5 per cent at risk of losing their job to a robot, and already eBay settles some 60 million disputes through online automation.

DoNotPay was an AI-chatbot developed by 19-year-old Mr Joshua Browder in 2015 and has successfully challenged 160,000 parking tickets in London and New York, appealing ($4m) £3.2m worth of fines and winning 64 per cent of cases. The harsh conclusion to take from this is that if you find your job highly repetitive, then you may well be at risk of automation.

At their best, robots help us to work smarter. The team behind workplace messaging tool Slack claim that 20 per cent of our working lives is spent looking for information, or for the person who has the information we need. Launching their new AI-assistant earlier this year, the company aims to eliminate this dead-time by centralising all basic information within a business.

Mr Shivon Zilis of Bloomberg Beta recently wrote an article for Harvard Business Review arguing that by cutting out repetitive tasks, AI will allow us all to “work like CEOs”, freeing us up to make the creative leaps that machines still cannot handle. Freed from the shackles of diary organising and expense forms, we should have more time to dream, discuss and develop.

The truth is, we’re at an AI tipping point and it’s impossible to know how it will impact our lives in the coming decades. Based on current technology, it looks most likely that AI will remove much of the grunt work from our jobs, but it’s also perfectly possible that we’ll create a superintelligence that evolves fundamentally different goals to mankind, leading to our eventual extinction. You think your current boss is bad? Try working for HAL 9000 or Skynet from The Terminator.

While the 20th century forced humans to work like robots, in the 21st century, robots may help us work like humans again. The optimistic view to take of AI is that it will handle many of the jobs that humans didn’t much enjoy doing anyway, allowing us to focus on becoming specialists in new fields of science and technology or to focus on creative disciplines. This is not to say that creative industries will be immune from AI-disruption.

Advertising firm McCann Japan made much fanfare of hiring its first AI creative director earlier this year, and M&C Saatchi released its first intelligent billboard in 2015, which learns and adapts its messaging depending on the viewer’s emotional response. Similarly, recent horror film Morgan was accompanied by a trailer directed by IBM’s Watson supercomputer, which had trawled hundreds of hours of horror trailers to identify the optimal horror formula.

True creativity is a messy and complex thing, though, requiring conceptual leaps and novel connections that cannot be laid out in code. Until robots become as irrational as humans, there will probably be a job out there for us, it just might not be one we currently imagine.