How to get employees on board for an AI transition
- Written by Grant Case, regional VP, strategic accounts, APAC at Dataiku
When introducing artificial intelligence (AI) into an organisation, it’s common for employees and business leaders to fear for their jobs or that AI will make their jobs more complicated. With just 39% of organisations in APAC having invested in AI technology to date, this barrier may very well be what is preventing companies from going down the AI path.
To ease this fear, businesses can adopt an augmented intelligence (AIA) or human-in-the-loop strategy. AIA is an AI approach combining the power and strengths of AI with human capabilities. AIA integrates AI systems into employees’ day-to-day work to help them make better decisions. When implemented successfully, AIA can help businesses create accurate and scalable solutions.
Here are three ways in which you can get your employees up to speed with AI in your organisation, using AIA as an approach.
1. Push teams to upskill
Before employees can leverage AI and AIA, organisations must empower employees to tap into the technology. Organisations can do this in three steps.
First, leaders must have honest conversations with their employees about the future of the company and where they see AI playing a part. They must communicate that advancing the business is a shared responsibility, emphasising that AI and employees both play a critical role.
Second, leaders must invest in training for their employees. This training should cater to individual employee levels of technical experience. This ensures employees have the required education and can ask the right questions when making decisions with AI.
Finally, business leaders must invest in tools and platforms that allow employees to use AI. This means tapping into commercial solutions that democratise AI use across the business.
Leaders that have direct conversations about AI’s importance and provide the requisite training and software tools will see employees become more engaged, take on greater responsibility, and think more critically about the business.
2. Showcase AI and how it helps achieve business results
AIA means using AI as a guide where people are at the helm of insight-driven decision making. Adopting an AIA strategy empowers employees to make decisions while taking context into account. A great example of AIA at work is Next Best Action systems. These are machine learning-based systems that create personalised recommendations based on pre-existing data.
One of Dataiku’s banking customers uses this technology for impressive effect. By leveraging millions of data points, the AI system shares client portfolio recommendations to financial advisors. This process is not unlike how Netflix uses AI to make recommendations for what the viewer should watch next. Like Netflix, the AI surfaces recommendations to the financial advisor based on the client’s portfolio and other clients like them. From there, similar to how the viewer controls what to watch next, the financial advisor decides whether to accept the AI recommendations, modify them, or make their own for the client. This is a perfect example of how machines are streamlining processes and providing options, while employees still make the final decision.
3. Educate and involve teams to overcome bias by being intentional and accountable about AI
Employees may feel a sense of apprehension toward AI because of the many stories of bias that have come to light in recent years. However, it is this very case study that showcases the importance of an AIA approach. To engender trust in AI, organisations should create AI that is intentional and accountable. One of the most critical functions of humans is to review the entire system at regular intervals. Humans can better understand when AI is not behaving in the way it’s intended and can thus change the system.
To illustrate, research has shown that 63% of companies plan to invest in AI hiring solutions in 2022, making it a key part of recruitment processes. However, there are dangers involved in relying only on AI to make talent recommendations. In 2018, Amazon went through this exact scenario and its experience illustrates why humans are important to the process.
In its effort to scale, Amazon had trained AI to read resumes and recommend candidates with a goal towards hiring. During the testing process, humans in the loop found troubling patterns with the AI. Given Amazon’s historic hiring of males, the AI had learned to favor male candidates thus penalized female terms in resumes and even discounted the graduates of two all women universities. Once recognized by the humans working and overseeing the project, researchers disrupted this bias by making the AI more gender neutral. This led the AI to be more inline with Amazon’s stated goals and desires for fairness in recommendations among genders. However, without a way to ensure bias could not return, Amazon could not assure accountability and thus closed the project.
Instead of failure, we should consider this project a success as it shows what it takes to build trust in AI. Amazon clearly understood and articulated what it’s goals for the program were and when they were not being met in the desired way, researchers made necessary changes. When Amazon could not guarantee the AI to be accountable for its methods, it shuttered the project.
It’s natural for employees to feel a sense of apprehension about AI, including a concern about it replacing them. Leaders must communicate the importance of AI to the organisation and the benefits to employees in helping them perform their jobs and progressing in their career. Leaders must be ready to show the value of AI. Finally, applying an AIA approach will create better business results, more employee engagement, and ease AI fears.
https://www.dataiku.com/
