We wouldn’t deny a farmer’s right to use a plough on the grounds that the tool is replacing workers. Artificial intelligence (AI) is simply a more sophisticated plough.
In order to stay competitive, product manufacturers must find ways to create efficiencies. The biggest cost for any business – however small or large – is labour. Cutting jobs should be a last resort. But improvements in AI and machine learning have made them very attractive alternatives.
Perhaps the best-known modern day example of machines replacing the jobs of workers is the rise of self-service checkout machines. Given how frustrating they can be to operate, they aren’t the ideal poster child for AI and machine learning – two technologies that are becoming commonplace on the factory floor.
"More and more, technologies such as these are stepping to the forefront of our strategies." That's John Condon, country sales director for Rockwell Automation, one of the world's biggest manufacturers of industrial automation technology. They build the factories needed by manufacturers to put Coke into bottles, the fig in the fig roll as well as your prescription into its correct dosage in pill form.
(Quick clarification: machine learning and AI are not the same thing. The former refers to machines capable of learning and adapting through experience, whereas AI is a step up in the sense that machines execute tasks “smartly”. Put another way, AI applies machine learning to solve actual problems.)
But they do have things in common. Both machine learning and AI share a common ancestry dating back thousands of years. They are the latest advancement in the tools our ancestors crafted from stone and wood thousands of years ago as a means of lifting another laborious task from their to-do list.
Technologies that make labour less laborious have always suffered from an identity crisis. They are demonised for replacing workers with machines. Yet everyone secretly loves doing less work by using tools that maximise one’s downtime.
“There is no reason to believe that the factory floor of the future will lose the human aspect, because these types of advanced technologies simply need humans who understand how they work in order to develop, adapt and work with them.” says Condon.
Technological unemployment
Why is this reporter providing a platform for a global corporation to air its views? Well, if anyone understands how unskilled workers will be affected by “advances in the means of production”, it’s the people providing those advances.
Artificial intelligence (AI) technology is increasing industrial productivity exponentially. It can ease decision-making for factory workers by providing informed predictive analytics. With AI, industrial workers can now more easily use the data from their equipment to predict production issues and improve processes with their existing automation and control skill set. “AI can also detect production anomalies and alert workers so they can investigate or intervene, as necessary, thus improving product quality and reducing downtime,” says Condon.
"Technological unemployment" as it was first described by John Maynard Keynes, leads to a "temporary phase of maladjustment". Short-term job prospects are likely to be affected.
But it’s important to remember that when the doors close on executive board meetings, the people inside don’t immediately grow hooves and begin scheming new ways to put their employees out of work. AI and machine learning are responses to calls for greater efficiencies from manufacturers.
New technologies pose a threat to old jobs. They always have. But do AI and machine learning represent a tectonic shift compared with every other innovation that has come before them? The “end is nigh” brigade make themselves heard every time new technologies lead to a change in working conditions (there is mention of humans being displaced by machines in the works of Aristotle). Rarely do they rally on a platform of “credit where credit is due”, in reference to how extraordinarily adaptable a species we are.
Telephone skills
You don’t hear about anyone being out of work because they lack telephony skills. Telephony skills were, at one point in the 20th century, a specialised skill. Making a phone call is now a ubiquitous skill like the mechanics of AI and machine learning are likely to become. We adapt to changing circumstances.
Training today’s workforce with skills required to adapt, adopt and implement automation and other forms of technology such as AI will increase productivity growth across all industries. “The skill shift we are seeing now, which has accompanied automation and digital transformation in the workplace, is resulting in the need for increased technological skills,” says Condon.
These technologies can provide workers with real-time insight into machinery compliance, causes of safety shutdowns or stoppages, and safety anomalies and trends. Safety and productivity are not mutually exclusive. They are complementary.
This isn’t some capitalist rant in support of the bourgeoisie and their continued monopoly over the means of production. It’s just an attempt to fuel a more realistic debate about our attitudes to technological advances of capital goods. AI and machine learning are evolved versions of earlier tools of production. Tools like the plough.
