AI use in businesses: myth vs. reality
Artificial intelligence (AI) has revolutionized every aspect of our lives, including how we work. That said, there are still a number of barriers which make some businesses wary of adopting AI. While these valid concerns are relatively wide-ranging, companies do share a few common worries when it comes to adopting AI.
The ROI of AI: Miracle or mirage?
AI projects have exploded in recent years thanks to a wave of optimism and interest in new technologies. But how did we get here? Not so long ago, AI experts (e.g. engineers, data scientists, mathematicians, etc.) were being snapped up on the job market, surprising partnerships and acquisitions were announced every day and the R&D community was booming.
Though AI was officially declared a scientific field in 1956 at a conference in the United States, it’s not until 2010 that we begin to see it used in the business world by pioneers like Google and Facebook. As with any new technology, it was hard to determine the necessary resources and return on investment. The effort required and the R&D-related risks diluted the advantages. AI adoption requires a significant amount of preparation and investment. Companies need to define their objectives, plan their business intelligence strategy and evaluate their capacity to take on this process. Québec and Canada as a whole have made AI adoption part of their economic development initiatives. There are many federal and provincial subsidies available to organizations of all sizes and all business sectors. These programs encourage companies to adopt AI, all while reducing the associated risks.
Last but not least, AI is democratizing, which means we’re seeing more turnkey tools that require no technical knowledge to deploy. Employees “just” need to be trained to use these new tools to reap the benefits of proven technologies with a sound ROI.
Two companies, Altus Drones and Parrot in France, have developed drones that inspect wind turbines. These drones use an AI program to detect visible issues or defects on all parts of the wind turbine, including the tower, nacelle, blades, screws and bolts. The drones are also synced to an AI system, which makes it possible to analyze the collected data and generate a report based on the selected parameters. In the past, this type of inspection fell within the purview of engineers specialized in wind turbine maintenance. Now thanks to AI, we just need technicians to fly the drones.
AI will replace humans: science fiction or reality?
Of course, your company’s ability to adopt AI depends on whether or not your employees embrace it. Your role is to quell any fears they may have, starting with the idea that they could lose their job.
Some dream of increased automation, while others dread it. Either way, there’s no easy answer. Sometimes the fear is well-founded, for instance with repetitive tasks. However, let’s be clear: Nothing can replace human intelligence! AI should be used as a tool to free up work time, enhance our ability to analyze and understand a given issue or to process more information. Science has repeatedly shown that the best approach is one that combines artificial and human intelligence. End users should therefore view AI as an ally, not an enemy.
Automate to free up resources
AI can be a great tool to combat current and future labour shortages. An over-saturated job market is forcing companies to rethink their workforce needs. Automating certain processes and tasks frees up your staff for higher value-added assignments. Still, managers should make sure that the new roles or tasks align with employee expectations.
Recent data shows that both prospective and current employees consider a company’s digital maturity when deciding whether or not to apply to or leave a job. According to 93% of employees polled, a company is more attractive when it integrates new technologies into its work processes.
In a Harvard Gazette article, Joseph Fuller, Professor of Management Practice at Harvard Business School, argued that AI helps employees work better, make fewer mistakes and showcase their expertise to share it more effectively throughout the company. For many jobs, AI can not only be used to automate repetitive tasks, but also to enhance more strategic tasks and decisions.
AI: ethical or amoral?
Becoming allies with AI requires trust, which has proven very difficult to date. For one, algorithms are abstract entities. There have also been a number of recent scandals involving the ethics of AI. In 2018, Amazon came under fire when their AI recruitment tool seemed to show biases against women.
That’s why it’s crucial to remember: AI isn’t ethical in and of itself. Like any other technology, algorithms are amoral—it’s the ways they’re used that can be problematic. That’s why we should take a closer look at the ethics of the strategic decision-making that happens before the AI tool is developed. If we feed biased data to the algorithm, it will unwittingly replicate the bias. Amazon’s algorithm was given data that characterized a good candidate. Since most Amazon employees were men, the algorithm considered being male to be a key characteristic of strong candidates. The AI algorithm therefore simply perpetuated an inequality that already existed.
One thing is certain: the worst way to take on AI challenges is to do nothing. Introducing AI tools will inevitability raise questions around responsibility and ethics. The answers are not straightforward. Yet companies now have an opportunity to gather and consult the right people, whether it be employees, suppliers or customers, in order to determine a path forward within a clear ethical framework. If there is an ethical violation, the (social, more so than legal) fallout can be quite severe. However, when the questioning process is carried out in good faith, the benefits are worth the risks. According to Andrew Burt of the Harvard Business Review, companies that use AI have an obligation to be transparent about how they choose to use it. They must also stay abreast of any recent advances to ensure that their AI complies with ethical standards and data governance, and to make any necessary changes.
At a time when the idea of corporate AI adoption still inspires debate and uncertainty, it’s important to stress that AI offers real, concrete benefits. The key to AI adoption success is to learn from past projects and to respect the following 3 pillars: 1) Technological: Plan your business intelligence strategy. 2) People: Back your transformation with change management. 3) Social: Establish clear ethical AI guidelines.
Successful digital transformations require understanding and anticipating your technological choices.
Download the 7th Edition of Tech Radar to learn more about how to adopt and integrate new technologies into your business processes.