Machines are getting more intelligent every day. AI and specifically machine learning-technologies are advancing to the point where they enable systems to move beyond rule-based behavior simply by learning from data. They’re getting so good at performing automated tasks such as processing invoices, answering frequently asked questions, and controlling product quality in assembly lines using computer vision that we’re systematically handing work over to them. We’ve officially entered the age of augmented humanity, where machines and humans collaborate to change our lives for the better. It’s a concept that SAP CEO Bill McDermott laid out at the World Economic Forum in Davos earlier this year.

Ultimately we need to set our own level of collaboration that we’re comfortable with, but it’s important to keep the relationship we establish with machines human-dominant. We can choose to opt in or opt out, share information or keep it in the vault, and take the reins or delegate tasks to the machine, based on our own preferences.

Assessing the value of trade-offs

The degree to which we will allow machines to assist us will be proportional to the value these systems can provide. Every day, some service pitches us an offer: share bits of information in exchange for something of value. If we think it’s worth it, we’ll take a survey or share insights into things like our customer experience when we purchased laundry detergent in a store. No harm done and we get something out of it, such as an improved checkout experience. If we don’t see value in the offer, we’ll simply swipe left and move on.

I’m willing to let a machine learning-enabled smart speaker listen in certain scenarios because it provides me with a valuable service: It enables me to focus on more urgent or enjoyable, human tasks. Just being able to start and stop a playlist or restock household products using a simple natural voice command is great when a two-year-old is tugging at you for attention. I’d also be fine with empowering HR systems with the ability to understand my natural language so that rather than having to click through a half-dozen screens, I’d be able to ask, “How many vacation days do I have left?” to get the job done — sign me up for that.

Managing the changes to work processes

Machines have gotten tremendously better at automating mundane tasks over the past few years, and no end to the improvement is in sight. McKinsey has reported that about 60 percent of all jobs could see 30 percent or more of their constituent activities automated. And while we will find ways to leverage the massive opportunities inherent to this next industrial revolution, we will also have to manage the likely substantial impact on our society.

On the personal side, perhaps you don’t want an AI-powered machine to plan your entire day for you. But what about scheduling and rescheduling occasional meetings to accommodate your family’s schedule? An AI assistant can monitor your calendar and tap into historic data and your individual preferences to learn about your optimal slots for meetings. The days of back-and-forth scheduling will be no more. However, if you enjoy extra “Are you free?” emails, then continue full steam ahead with manual scheduling. Either way, the choice is yours.

Drawing your information-sharing line

We’ll all have choices to make as AI functions get more sophisticated, contextual, and targeted. It comes down to determining how much information you’re willing to share about yourself. Just as many of us are willing to share personal data with corporations in return for a customized browsing and shopping experience, we should decide the tradeoffs we’re comfortable with. While the value of AI becomes more and more apparent on the individual level, it should let you control what kind of information you put into it and what you get out of it. Keep in mind that the machine will only be as good as the data we feed it.

Ultimately, control of human-machine interactions will not be a binary arrangement. We should determine individually on a sliding scale how enabled machines are to help us. AI may empower machines to assist us more than ever before, but the ultimate decision on their roles in our lives should lie with us.

Till Pieper is general manager of the SAP Leonardo Conversational AI Foundation.