The future workforce – work smarter in collaboration with machines
According to Gartner (2018), spending on Artificial Intelligence (AI) has reached $1.2 trillion in 2018, that’s 70% more compared to 2017. The same source forecasts that AI spending will be as high as $3.9 trillion in 2022. To put that into perspective, that’s equivalent to the GDP of Mexico ($1.1 trillion).
Many say that AI will take our jobs and leave us with pretty much nothing to do. Recently, NOS (2018), stated that in seven years 52% of our jobs would be in the hands of robots. This might be a true statement, but news articles like this often don’t consider the new positions that might arise from these advancements. This article aims to provide a favorable view of how the future of work may look like in combination with developments in AI.
A very brief history of work
Major historical ages in the past have always defined the way how humans work. Firstly, the hunter-gatherer age (which lasted millions of years), followed by the agricultural age (thousands of years), the industrial age (a couple of centuries) and finally the information age (a few decades). Currently, we’ve arrived at the augmented age (Ted Talks, 2016).
In the augmented age, many human limitations will belong to the past, since we will be augmented by computational systems that help us think, construct things or support us with tasks we don’t necessarily enjoy. You probably think this sounds more like a Science Fiction movie, but in fact, we’re already quite augmented. Due to the fact we all have smartphones, we can find the answer to almost any question within seconds. However, this is just the beginning (Ted Talks, 2016).
Smart manufacturing with cobots (collaborative robots)
Within the manufacturing industry, there have been two significant waves of business transformations. In the early 1900s, Henry Ford introduced standardized processes with the assembly line. Afterwards, in the period between 1970 and the 1990s, Information Technology made automated processes possible, which caused many people to lose their jobs. Nowadays we’re at the brink of stepping into the third wave, in which adaptive processes will become a reality.
Adaptive processes mean that processes are customizable without too many interventions, something which is now unimaginable. This adaptive nature of the new work floor is made possible by real-time data. For example; traditional robots were pre-programmed, which meant that they had to be re-programmed whenever the process changed. Robots with Deep Reinforcement Learning software are given a picture of the desired outcome and then uses trial and error to figure out its own solution. After 8 hours of training, the robots have an accuracy of 90%, which means that the previously occupied programmer can now focus on more complex issues. These robots can work together very well with the human workforce on the floor, who will still be needed for positioning small objects like wires or handling dynamic objects (Daugherty & Wilson, 2018).
Optimize back-office operations with intelligent systems
Have you heard the story about a Dutch bank that has to pay €775.000.000 in addition to the stepping down of their CFO after a massive money-laundering scandal? (NOS, 2018) Well, perhaps they should take a look at AI-based analytical tools, developed for anti-money-laundering. A major US bank had much trouble with complying with the extreme expectations of the US Department of Justice. Thus, they decided to go all in. The new system they introduced is better able to segment transactions and accounts thanks to machine-learning algorithms, as well as setting optimal thresholds for alerting investigators because it uses the most recent available data. In addition to that, the system even finds undiscovered patterns between customers of the bank (Daugherty & Wilson, 2018).
The results thus far are astonishing; the ratio of false positive alerts decreased by almost 30%. In the scenario of ING, a false positive would be when there’s an alert of money laundering, which has to be investigated, but in reality it seems to be a false alarm.
Also, it takes the bank’s employees less time to investigate each signal, leading to a cost reduction of 40%. Luckily, the bank will still employ people who work in this field of expertise, since they will always be needed for judgment cases and compliance expertise (Daugherty & Wilson, 2018).
Less creativity needed in R&D with generative design
Generative design might be my personal favorite, since it might provide people who are not as creative as they’d like to be the opportunity to design awesome things. The best example of this concerns AI-powered generative design tools, which use IT (i.e., software, data, and algorithms) to synthesize geometry to come up with designs for all sorts of things by itself. Someone only needs to determine parameters like goals and constraints when using this software (Ted Talks, 2016). The use of this software makes it for example possible to create designs which make aircrafts lighter and buildings stronger, without having to deal with the challenge of designing it by yourself. Generative design calculates the requirements needed and provides several possibilities. Afterwards, a designer decide which model to pursue.
A manufacturer that applied this technique is aerospace company Airbus, who wanted to create the best possible design for the partitions that divide airplane cabins. The final design appeared to be half the weight of the previous design, while still being complient with safety regulations. This invention resulted in a huge decrease in fuel cost for the company (Airbus, 2016).
Another example is a walkway over a canal in the city of Amsterdam. The bridge is generatively designed by a team of people that uses AI-powered software. Now that the design is finished, they’ll test the bridge, after which they will build it. The constructing of the bridge, however, will happen without any human intervention. 3D printing robots will start making the bridge and won’t stop until it’s finished (Ted Talks, 2016). Optimistic estimations even state it might be possible to design entire buildings using only generative design tools, within the next ten years (Starrapid, 2018).
You might be asking: “Why is this so valuable?” The thing is that generative design tools can perform thousands of iterations much faster than a designer could. In addition to that, designers are often biased (whether consciously or unconsciously), because they factor in manufacturing limitations. AI, however, reasons from first principles. The image below shows how the software has generated a design for the swing arm of a motorcycle, which would be highly difficult for a human to come up with.
Source: Autodesk, 2018.
Focus on what’s important; skip the annoying tasks in customer service & marketing/sales
If you’ve ever worked in sales or one of the other fields mentioned in the title of this paragraph, I know you’ve had several annoying tasks you just had to do to get the job done. Examples of these are the endless e-mails you had to write, difficulties regarding scheduling appointments with clients and taking notes while you’re talking to clients. Developments in AI will solve these frustrations. Examples of this are Optimail, which takes care of the timing, content, and personalization of your emails (Optimail, 2018). Then there’s EVA by Voicera. Eva records your meetings, transforms it into text and automatically highlights important content. Afterwards, you’ll get an automated email with your notes (Voicera, 2018). The last example is called Evie, which takes care of scheduling your appointments as an assistant would do. When you’re scheduling an appointment with someone, you add the bot Evie to the “CC” in the e-mail, which automatically sends invites back and forth based on your availability, until it fits (Evie, 2018). How convenient is that?
So, while one could have a pessimistic view regarding robots taking our jobs, it’s also quite likely that AI will enrichen our lives by omitting the annoying activities which are incorporated in our work. Within manufacturing, it’s likely that an actual workforce will still be required, although their jobs might look a bit different. Finally, the new possibilities regarding endless designs in R&D are quite exciting if you’d ask me.
Airbus. (2016). Pioneering bionic 3D printing. Retrieved from Airbus.com: https://www.airbus.com/newsroom/news/en/2016/03/Pioneering-bionic-3D-printing.html
Daugherty, P. R., & Wilson, H. (2018). Human + machine : reimagining work in the age of AI . Boston, Massachusetts: Harvard Business Review Press.
Evie. (2018). Corporate website. Retrieved from https://evie.ai/
Gartner. (2018). Gartner Newsroom. Retrieved from https://www.gartner.com/newsroom/id/3872933
NOS. (2018). Financieel topman ING web om witwas schandaal. Retrieved from https://nos.nl/artikel/2249862-financieel-topman-ing-weg-om-witwasschandaal.html
NOS. (2018). Over zeven jaar 52 procent van de banen in handen van robots. Retrieved from https://nos.nl/artikel/2250766-over-zeven-jaar-52-procent-van-de-banen-in-handen-van-robots.html
Optimail. (2018). Corporate website. Retrieved from https://www.optimail.io/
Starrapid. (2018). The evolution of generative design software. Retrieved from https://www.starrapid.com/blog/the-evolution-of-generative-design-software/
Ted Talks. (2016). The Incredible Inventions of Intuitive AI. Retrieved from https://www.ted.com/talks/maurice_conti_the_incredible_inventions_of_intuitive_ai
Voicera. (2018). Corporate website. Retrieved from https://www.voicera.com/