“We are all headed to creepy – and the people in the room here are making it happen.”

That was the provocation Salesforce CEO and tech VC Marc Benioff directed at a room full of marketers and business leaders at this week’s Fortune Brainstorm:Tech dinner, co-hosted by gyro at the kickoff to CES in Las Vegas this week.

He was talking about the use and abuse of data by marketers – with his own much loved Fitbit as an illustration of how the relationship between consumers, creators and marketers is changing.

The latest Fitbit can access location, heart rate, steps taken, whether the user is sleeping or awake, and soon, a whole new host of information including blood pressure, insulin levels, and hydration. This data is layered on top of customer information such as email and payment details and integrated with social data. In Benioff’s case, the result is that friends, like Michael Dell, can joke with him about how much exercise he’s getting on his holidays. In an interesting aside, he acknowledged that for a man who runs the fifth largest publicly traded software company in the world, allowing the public access to his personal health data could also carry unintended consequences – but even for the rest of us, this scenario is one with the potential to slide into uncomfortable territory.

According to Benioff, “The AI Spring”, this year’s investing fad will bring with it further challenges: Venture capitalists and big companies alike have been funneling investments into research on machine learning and “Deep Mind” computing that hopes to replicate human pattern identification and judgment. The resulting technology allows computers to “learn” how to reach entirely novel (not pre-programmed) conclusions based on large data sets, and continually improve predictions and modelling.

For instance, IBM’s Watson is already being put to use far beyond the game show circuit. It is being integrated into apps and services with purposes ranging from health and fitness optimization, recipe creation and Veterans’ PTSD counseling, to skin cancer detection. With Watson Analytics, marketers have direct access to the computer’s predictive power. Financial intelligence startups such as Sentient Technologies seek ways to better predict market behavior. And MetaMind, itself a Benioff-backed startup, is working to give enterprises advanced tools in sentiment analysis, visual computing, and database prediction.

Predictive tools can make people feel uncomfortable – an unsolicited email or notification can alert consumers to the fact that they’re being monitored, and untimely, persistent, or unwanted sales messages can be pushy and irritating. That’s why Benioff’s “heading to creepy” is a challenge to marketers.

So how can we use the wealth of new data we are able to access to be more empathetic and human, instead of less? As gyro ceo+cco Christoph Becker told the audience, technology can also enable companies to be more humanly relevant. A hotel can use its knowledge about repeat customers’ preferences to help employees act more like a host welcoming an old friend. A hospital can be more helpful to patients needing assistance in staying healthy. Most importantly, they can do all of this without feeling like a prying neighbor.

As marketers, we have the power to decide what the future of AI and data will feel like. Are we inevitably headed towards creepy? Or will we take advantage of this big and exciting opportunity to be humanly relevant? It’s up to us.