How do you balance that need for innovation with considerations about bias, ethics, or other concerns?
I’ve seen a lot of technology come and go over 30 years of doing this kind of work. The most important is to balance speed with responsibility, caution, and understanding of the environment.
All humans inherently have some bias built in, so how do we avoid bias in technology? How do we make sure the algorithms that we build into AI are responsible? AI is only as good as the underlying data. If the data is bad, then what the AI suggests is going to be bad.
We formed an entire set of AI principles at the firm and have a governance structure that we've put in place. So, as we are considering, should we adopt this bot? Should we utilize AI? We ask, who does it benefit? What are the potential pitfalls? How are we going to make sure that it is free of bias or unintended consequences? How are we going to responsibly respond to EU data privacy restrictions on GDPR or a whole host of new laws and regulations around bias in the U.S.?
Having that structure and technology allows us to say it’s ok to go fast, but let’s build responsibly, especially in an organization of our size. You do that by building in the principles and governance structures early on. It’s allowed us to move with speed but also with responsibility.
We’ve seen high-profile cases of AI-related layoffs. How do you protect employees as you roll out AI?
It's the question we get all the time from our people. First and foremost, I say this: The robots are not coming for your jobs, they’re coming for your tasks.
And in most cases, they're coming for tasks that you don't want to do anyway. So much of the automation is for mundane, routine work that people don’t want to make the time for. IMF did a piece of research that found 60% of jobs in advanced economies will be impacted by AI. But keep in mind that it's rarely entire functions or entire jobs; it's pieces of jobs.
First and foremost, I say this: The robots are not coming for your jobs, they’re coming for your tasks.
Dan Black | Global Talent Acquisition Leader, EY Global
We’ve been trying to help to reeducate and upskill our people. We want people to understand how this complements their job because we are about human skills at our center. That means that companies need to be responsible and proactive in building those learning opportunities for people to upskill and focus on the human elements of their jobs, and by the end, be more comfortable with the AI technology.
Our global data and AI leader, Beatrice Sanz Sáiz, joined a call with my recruiters to talk about why AI can be scary. The fear is typically that companies could get the same results with 50% fewer people by using AI. But, she argues, who would want to? What company is looking for the same level result year over year? Everyone is looking to grow. Like any company in the world, if you’re looking for better growth, expanded markets, or a bigger customer base, then yes, you need AI, but you’re also going to need more people to get there. The human element is something that will never go away. The key is understanding how to work best and augment that effort with AI.
What are the biggest lessons you’ve learned from EY’s journey with AI?
Two big ones stand out. There needs to be much more experimentation as this is not one size fits all. You can’t bring in this company, this tech, this piece of AI, or this bot, and expect everything is going to be right as rain. You have to experiment to get it right. That applies to both companies and individuals. As fast as everyone wants to go, there's going to be a trial-and-error period and you have to be ok with that.
The second thing is that if you’re going to make the most of AI, you have to understand how AI is going to work alongside the human complement. There’s been a lot of rush to see what or who can be replaced with automation, as opposed to how to create a blended model. In any other facet of your life, even the ones that technology has made exceedingly simple, there's still a need for human involvement. That’s going to be the real key to success for organizatiosn going forward.