Artificial Intelligence (AI) and Machine Learning (ML) have come to such a level these days, that it has the potential of changing many aspects of modern life, including manufacturing technologies. However, with such rapid and radical change, comes the fear that it may make things worse in the future: Will machines take away all of our jobs? Will we live in a dystopia like Skynet in the Terminator movies, like Stephen Hawking and Elon Musk suggest? Is there any justification for all these claims of benefits, as well as fears or is it just hyperbole?
We will discuss some of these topics and try to find some answers in this article. Let’s start with the proven benefits of AI and ML in manufacturing.
Benefits of Artificial Intelligence and Machine Learning in Manufacturing
The never-ending challenge of manufacturing is to produce higher-quality goods at a lower cost, faster. The developments in AI and ML offer remarkable new ways to achieve this goal. Many people in the industry call this another revolution in manufacturing, and naming it “Manufacturing 4.0” or “Industry 4.0”. But what does this mean exactly? How do AI and ML reduce costs, improve quality and quantity?
Here are some ways:
Even the most experienced human quality inspector would have a difficult time detecting a flaw in a product, less than the thickness of a human hair, by the naked eye. Robots with ultra-high resolution cameras can detect such small defects much better. A human inspector then looks at the pictures, and tags them as flawed or not, and the machine learns to detect flaws better.
One of the highest costs of manufacturing is the downtime of the whole production line, because of one critical piece of machinery is broken. AI is now used to predict defects in machinery before problems happen. This is called predictive maintenance, and it reduces or eliminates costly downtime.
Design is an iterative process that requires lots of trial error. Now thanks to AI, design goals and engineering constraints can be entered to a computer, AI makes thousands of permutations and combinations of all the possible design choices, and then comes up with a handful of some of the best solutions. This is called generative design. The design team then evaluates a handful of preselected designs.
The steel industry uses AI to reduce mill scaling, which wastes 3% of the steel produced. AI was able to reduce this waste by 15%, saving millions.
Supply Chain Management
Just like AI and ML predicting maintenance problems, they also predict potential problems and delays in the supply chain. Therefore, it now becomes possible to avoid and minimize delays, as real-time updates are available through the whole supply chain.
These are only 5 of the many ways that AI can help with reducing costs, while also improving quality.
Will robots and computers take all the jobs?
First, not all jobs can be automated. An OECD report released in March 2019, estimates only 14% of jobs can be lost to automation, which is much less than the 2013 Oxford report, which estimated 47% of jobs in the USA can be lost to automation.
Second, it is necessary to realize that, AI is not a replacement to, but rather, an extension of human intelligence. All those machines still need humans to program, maintain, control, and improve upon. Plus, the fear “machines will take our jobs” is largely based on a misunderstanding of how the economics of technology and automation work.
Say, because of new technology, 10% of jobs become obsolete, and say a car factory of 1,000 workers fires 100 employees. What happens next? Costs of the factory are reduced, and therefore, cars become cheaper. What happens when cars become cheaper? People buy more cars. What happens when people buy more cars? Demand increases. What happens when the demand for cars increases? The factory expands and hires more people.
Over the long term, automation results in lower prices and more jobs for everyone. Those who were originally fired may require retraining or may start working in some other jobs in some other field, but overall, automation benefits everyone over the long run.
What about the low IQ, the poor, the disabled?
One concern about automation taking jobs is, even though the overall demand for employees get higher with technology, automation, and artificial intelligence, that demand is for highly intelligent and well-educated employees. According to US Army IQ tests, about 10% of the US population has below 85 IQ points and is “unemployable” by the military standards. In a highly complex society, there aren’t many employment prospects for such people. What will happen to them?
Well, the thing is that even the low IQ and disadvantaged populations are better off through automation in the long run. Due to automation, prices drop. Because prices drop, now it is cheaper and easier to take care of the basic needs of the disadvantaged. Also, because of technology and AI, people become more effective and have more time in their hands. Therefore more people can devote their time to charitable activities. The friends, family members, and people in charitable organizations can take care of such people.
Just like it happened with the agricultural revolution first, and then, the industrial revolution.
Will we live in a “Terminator” world?
While there are many advances in technology and AI in particular that make our lives better, all those dystopian science fiction relies on something more than artificial intelligence: artificial self-consciousness. Which is, a machine that can identify itself as an independent sentient being.
Such dystopian science fiction makes good clickbait and blockbuster movies. However, we are far from creating self-consciousness. In fact, all the scientists and engineers in the world are far from even defining consciousness, much less replicating it in machines.
As with any other rapid and radical change in society, it is normal to expect some fear of the unknown and unfamiliar, and therefore resistance against any technological advancement. Same thing with AI and MI. However, AI and MI have many exciting benefits opportunities, with little downside. They are here to stay, and worries about AI causing massive job loss, or a Terminator-like dystopian future are largely unfounded.