It’s safe to say that in 1997, Garry Kasparov wasn’t a fan of AI. His world-renown match with IBM’s chess computer, Deep Blue, marked the first time a chess champion lost to a machine. Though Kasparov initially questioned whether Deep Blue was a legitimate competitor, he’s since changed his perspective on AI. In his 2017 book Deep Thinking, he argues that humans have nothing to fear from machine learning. Instead, humans can augment their creative capabilities with machine learning’s data processing achieve better results. Kasparov suggests that he’s learnt a great deal from Deep Blue’s strategy.
Kasparov’s change of heart parallels business’ adoption of machine learning software. Popular culture’s AI-based anxiety stems from fears that machine learning will render human workers obsolete. Though in practice, employees often benefit from machine learning solutions.
When Automation Anywhere surveyed 4,000 employees from AI-augmented businesses, they found the following results:
- 72% of respondents view AI as a workplace tool, not a technology that replaces them
- 57% of employees believe productivity would accelerate if they could experiment with other AI tools
- Businesses that adopt AI see a 28% performance increase and score 33% higher on factors that make the workplace feel “more human”
This isn’t to say that AI doesn’t create challenges for industries—job retraining is essential for adjusting to automation trends—but early research suggests both employees and businesses can use AI as an advantage. Like Kasparov, individuals who once feared that machines would make us obsolete are now discovering their value.
Machine learning makes us stronger
There’s a popular misconception that once AI advances to a certain degree, it’ll outmatch human capabilities. This couldn’t be farther from the truth. Humans play a major role in how we use and implement machine learning systems. What’s more, humans are essential to ensuring AI solutions function as intended.
Let’s use automated language translators as an example. Even if these systems are perfect—which they certainly aren’t—language constantly changes to accommodate new information and cultural nuances. Microsoft’s Jason Lanier noted that Google’s language software still requires human translators who account for new words, phrases, and contextual meanings. Without it, even a perfect AI-based language solution will become outdated.
After the Deep Blue match, Kasparov observed this phenomenon first-hand. While the AI system dethroned a human chess champion, human players eventually made a comeback by leveraging their own AI. Computer algorithms generated effective chess strategies that human players optimized based on their own experience. This scenario repeats itself constantly. When AI surpasses human experts at a given task, the experts learn new skills from the AI.
Automated systems are powerful, yet limited when it comes to creativity and innovation. Machine learning algorithms can calculate millions of trial-and-error scenarios to find an effective strategy, but when an AI encounters a scenario it’s never seen before, it struggles to find a new course of action. That’s where humans come in, lending their intuition and creativity to improve AI models.
What businesses can learn from AI
The principles of AI-augmented chess also apply to automated workplaces. Effective AI solutions don’t automate in a vacuum; they require human directives to refine their models and processes. Ironically, businesses can’t automate without people. What’s more, automation frees up resources that can be reinvested in other fields, such as customer interactions.
The exact implementation model will vary from business to business, but here are some best practices to consider when integrating solutions with your workforce:
Create an internal AI “startup”
There’s a reason tech startups saw impressive gains with machine learning; these businesses are agile by design, which lets them take risks with experimental initiatives. By contrast, established businesses are reluctant to take the leap with AI despite the noted benefits.
One excellent compromise is to build an internal team dedicated to AI research and initiatives. These “startups” will have more freedom to test machine learning platforms and make recommendations to other departments. In turn, the team can take feedback from departments to refine algorithms or test alternative machine learning solutions.
Support employees, not jobs
While employees can adapt to AI, that doesn’t mean their job descriptions will remain static. Over time, oversight roles and non-automated conceptual tasks will replace monotonous work. Businesses can ease employees through this transition in a variety of ways:
- Offer skill retraining that helps employees make the best use of new machine learning tools
- Work with employees to create job descriptions that reflect their revised roles
- Collaborate on objectives that employees can focus on post-automation
When automating any task, there’s a risk of focusing entirely on results and ignoring how it was calculated within the “black box.” As such, it’s becoming increasingly important to have an AI transparency policy that accounts for any government regulations and internal algorithm-related procedures. Beyond the compliance considerations, transparency resonates with many customers and honesty is always a safe bet for brand success.
AI has the potential to revolutionize our workplaces in ways that are difficult to imagine. But the overarching goal shouldn’t be to automate for automation’s sake, it’s to encourage a higher-performing, more effective workplace. Humans will have a crucial part in that role, from assisting with AI integration to augmenting their work with machine learning strategies.
Garry Kasparov learned first-hand that machines can surpass humans in certain tasks. He also learned that machines can make humans more effective than ever before. That dynamic will likely drive business and human development into the 21st Century and beyond.