Allyson Jacobsen
Global Marketing Director, Edison Portfolio
GE Healthcare



Artificial Intelligence, despite its potential to transform an industry, is not a product by itself. There is always a person behind the AI. They are gathering the data, influencing the data leveraged, building the models, and training the ultimate intelligence.

AI can transform our daily lives, with connected devices predicting our needs and tackling our to-do lists. In the professional world, however, all of this data fuels innovation and product development. It opens up endless opportunities for corporations to improve customer outcomes, increase sales, and improve the bottom line through AI based intelligence around our products, services, and customers. But with all that data and accelerated scaling, we can run into new challenges.

As the famous quote from Spiderman goes, “With great power comes great responsibility.” That power comes in the form of being able to leverage AI to understand what our customers need, and to build innovations faster than ever. We get our products to market and increase sales at a fraction of the time it used to take. But at what cost?

Some organizations depend too heavily on Artificial Intelligence that was created using a succinct data set, not on an extensive data set that is rich in variety and veracity to build those innovations. In other words, the AI is working off a simplified environment without all the variables in play. There is also a gap in the humanization of that data. Organizations must look beyond the process of simply optimizing products and workflow. They must ensure the AI innovations are inspired by human context, emotions, and are intended to improve people’s lives.

So, where does the “responsibility” part of that quote come in? Here are a few things organizations can do to improve the odds of success in scaling with AI:

  1. Take a broader view. Bring data from various parts of the world into your modeling. This will improve the end result of the solution you’re building. Technology, training, and the availability of valuable information exists depending on the maturity of the country. These varying maturity levels will help you train an AI model to be intelligent. For example, in healthcare something as simple as taking a chest x-ray is very different in the US than in many other parts of the world. Even the age of the x-ray device and the training of the technician varies greatly around the world, which impacts the image itself. These are very important things to take into account.
  2. Bring in the human element. Teach your model to think with more contexts: to learn emotional reactions, consider the various maturity levels of your end consumer, expect the unexpected and leverage the deep and varied context a human would make to ensure accurate responses. A while ago, a Microsoft AI Twitter “bot” was corrupted by trolls on Twitter because the bot wasn’t trained to process words and sentences. It just parroted offensive language, and didn’t have discernment built into it.
  3. Ensure you are principled. Make sure you are using your data in a secure and principled way. Don’t store or share your data in a way that goes against the greater good. We all know the trouble that Facebook got in by handling data recklessly. Data is powerful, and it needs to be used for good, not evil.
  4. Go to market with integrity. Data and AI are more powerful in sales and marketing than just about anywhere else. However, there is less and less room for error than ever before. We use all kinds of data, bots, and automation tools to streamline our marketing and sales activities. This increases reach and the speed in which we reach the audience, but used without humanization, it can go very wrong.
  5. Consider people. Employees, customers, users, experiencers, everyone. People need to be at the top of every consideration. How will your employees benefit or be hurt? Do they understand the value of AI? Are AI applications in the workplace eliminating personal relationships or improving the quality of life? Do customers feel like people don’t exist in your organization? There’s no robot substitute for human connection.

In addition to all of that, how does the AI fit into the existing environment? I believe that AI is most powerful when it is integrated into our existing workflows, processes, and innovations, providing seamless product enhancements, improvements, assistance and guidance while being completely invisible.

AI technology can certainly be proprietary, but if we’re committed to making the world better through technology, we have to willingly share our findings as it relates to humans flourishing. Tools like open source development help democratize data and improve the world around us.

I’m excited to be participating in the upcoming 14th Annual New Product Innovation & Development: A Frost & Sullivan Executive MindXchange. For those attending, I’d love to discuss this topic and others in greater detail.

Allyson is the Global Marketing Director of Artificial Intelligence and Analytics Solutions at GE Healthcare. In her role, she is responsible for the strategic direction and execution of all global marketing functions.

A highly adaptable marketing executive with over 15 years’ experience, Allyson has worked internationally with successful global brands. At IBM, she directed globally influential teams, providing integrated marketing programs and strategies to increase awareness and revenue. With, a B2B2C training marketplace, Allyson led a global marketing and e-commerce team focusing on brand awareness and growth campaigns, exceeding annual targets by 200 percent in 2015. Most recently, at an information technology consulting firm known as Trianz, Inc., Allyson developed and implemented global omni-channel marketing strategies responsible for exceeding revenue goals by 22 percent.