A long time ago, Mark Twain wrote a note to a friend in which he apologized for not having time to write a shorter letter. In other words, Twain wished he’d had more time to think deeply and craft his message more succinctly. Life was slower back then – automation was just a glint in a futurist’s eye.
In the digital age, my sentiment is along the same vein, but different: I’m sorry I didn’t have time to write you an even more thoughtful and insightful blog that adds more value to you as a reader. Time is always short, so if I’d had a digital assistant to do my research for me, I could have written a better, more insightful blog in less time. More of my time and effort could have been spent thinking more deeply and providing an even better answer to pressing questions like, “Are humans at war with the machines? Will automation and artificial intelligence (AI) take away our jobs – and the future of our children?”
An interesting perspective
Ten-time world chess master Garry Kasparov has an interesting perspective on this because he became the proverbial man in the first real “man-versus-machine” competition. In 1996, he played two matches against the IBM supercomputer Deep Blue, and he lost the second match. It was a crushing loss, and even he couldn’t help wondering, might this machine be invincible? Was he irrelevant now? And would people stop playing his beloved game because it had been conquered by a machine?
The good news is none of these dreaded things happened. Chess has actually become more popular than ever, evolving to include advanced and freestyle forms where man + machine – or “augmented me” – has proven to be the winning combination. In a 1998 advanced chess competition, two amateur American chess players using three PCs coached their machines so successfully that they counteracted the superior chess knowledge of grandmasters and won the match.
But the PCs couldn’t have won without the people. As Kasparov noted in his TED Talk: “Machines have calculations. We have understanding. Machines have instructions. We have purpose. Machines have objectivity. We have passion. We should not worry about what our machines can do today. Instead, we should worry about what they still cannot do today, because we will need the help of the new, intelligent machines to turn our grandest dreams into reality.”
The big takeaway
This story captures what most people who are afraid of AI don’t understand: even AI-enabled machines can’t actually replace people. That’s because machines can’t understand context and coordinate people and processes based on their understanding of that context. Only humans can understand and take into account things like interrelational dynamics, personality, organizational culture, attitudes and preferences, politics, moods, strategic visions, body language, and so much more. For instance, an understanding of context is needed for complex welding repairs (as opposed to repetitive welding in a manufacturing environment), such as those done by plumbers and oil rig workers; these workers need to take into account the age of assets, the weather, the type of pipes involved, and other factors – more than AI can handle.
This is not to say that repetitive jobs will not be going away. They will – and sooner rather than later. Even highly paid people like accountants and actuaries should expect the number-crunching parts of their jobs to disappear. But humans will always be needed to provide insight into what those numbers mean, to weigh priorities and make optimal, context-informed decisions. We will increasingly rely on AI to augment us and do the busywork faster and more effectively (in my case, the grunt research for a blog) so we can do the higher-level work only humans can do. With their help, we’ll just do it faster, and with more and better data-driven insights to support those decisions.
Implications for the enterprise
Figure 1 nicely summarizes how I envision this trajectory of AI-augmented humans from an enterprise perspective. We’ll all do less repetitive work over time, freeing people to solve more problems, be more productive, make better decisions, and focus on more interesting and challenging activities like exploration, ideation, and innovation. These are things people do extraordinarily well – and deliver higher value to the enterprise.
Examples of positive impact
So what can we look forward to in practical terms?
Well, we’re actually already seeing examples of positive impact from AI assistance. Let’s take a look at a few examples.
Enabling systemic change using AI to address bias
Consider Amazon’s recent attempt at automating the recruitment process. It wanted an intelligent system that could look at a collection of resumes and name the top candidates by analyzing the historical data of successful past hires. They eventually detected a problem – the data they had used to train the system reflected the tech industry’s dominance by men. So, AI was downgrading the resumes of women and people who attended women-only colleges, for instance, which perpetuated a hidden hiring bias.
This may sound like a failure, but there’s more to it than that – it’s an incredible learning experience. AI revealed hidden bias that had plagued hiring at Amazon for years. And the project team has since modified the data used to train the AI to prevent such bias, thus ensuring HR and hiring managers don’t perpetuate unfair hiring practices going forward.
Creating new kinds of job opportunities
AI will create exciting new job opportunities – for example, where people act as trainers and explainers. Trainers will be used to train chatbots and digital assistants to offer more “human,” empathetic responses, recognize and respond to sarcasm, and more. Explainers will bridge the gap between humans and the systems augmenting them so people can interact with them more effectively and trust the data-driven recommendations provided by AI. They can also diagnose and fix errors made by AI so they aren’t perpetuated.
Making workers more effective, efficient, and customer-centered
AI can do more than just automate repetitive work. It can also anticipate needs and aggregate information instantly to make people more productive. For example, it can advise field technicians about which tools and replacement parts will be needed before heading to customer sites. It can provide intelligence about customers calling into a call center so agents are ready, have insights into what they may be calling about, and can provide faster, better service.
Providing personalized, satisfying work experiences that boost engagement and retention
The emerging workforce is growing up in a post-digital world that has raised its expectations in terms of “experience.” Employees expect tactile consumer experiences from enterprise software, personalized experiences, and work that is valued, engaging, and enriching – not paper pushing or similar grunt work. This is especially true for employees managing their HR tasks. AI is helping companies create these work experiences by providing self-service options that empower employees to take problem-solving into their own hands with solutions that are just as intuitive as the mobile apps they use in their personal lives.
I’m looking forward to having a smart digital assistant join me so I can work smarter, faster, and more effectively. Are you?
Learn more about how AI is enhancing business; download “Making the Most of Machine Learning.”