Updated: Jul 18
After 13 incredible years working at Duke’s Center for Advanced Hindsight and 6 amazing years with Pattern Health, I’m joining Apple to lead behavioral science for their Health AI group. I’ve gotten a bunch of questions, so I pulled together some FAQs here to quench your curiosity.
Are you moving to Cupertino?
Nope. My team is primarily based in Seattle and Durham, and I’ll be staying in Durham with a bit of travel when and where it makes sense.
Are you excited? Nervous?
Why get into machine learning / artificial intelligence?
I’ve been feeling dissatisfied for a while now with some of the traditional methods of behavioral science. Don’t get me wrong; I'm still head over heels with the RCT, and I think it is the absolute best tool for some situations. But no tool is perfect, and I’ve just had this nagging feeling that as a field, behavioral science is missing out on the flexibility and predictive power of machine learning techniques.
There are some behavioral scientists already doing very exciting work in this space, but let me just list a few that I know are paving ground with forward-looking research methodologies:
Linda Collins (here’s my interview with her: On Hard-Nosed Empiricism And Optimizing Interventions) and this incredible quote, which I have seriously taken to heart:
“There is an interplay between research questions and research methods. Of course, research questions must drive the choice of research methods, but it is also true that the emergence of new methods opens up possibilities and stimulates new questions. Therefore my advice is: Keep your methodological training up to date so that you are always aware of what the possibilities are.”
Susan Murphy, renowned for her work with micro-randomized trials
Ganna Pogrebna, behavioral data scientist
But Aline, you are a complete novice in machine learning!
This is true! That’s why it’s so incredible that Apple is supporting me with its AI / ML Residency Program, which includes machine learning and AI courses, mentorship and training, and collaboration with seasoned experts on high-impact projects. The program runs each year, inviting domain experts (including but not limited to behavioral scientists) to build up their AI / ML skills and gain from their interdisciplinary expertise in exchange. I am thrilled to be working with Nick Foti, Guillermo Sapiro, and the rest of the Health AI team to get up to speed on machine learning methods so that I can more successfully leverage behavioral science in our projects.
What can behavioral science offer machine learning / artificial intelligence?
So much! As an applied behavioral scientist passionate about shaping positive health behaviors through digital technologies, I believe that machine learning and behavioral science can work together as complementary tools. Machine learning can take traditional behavioral science experimentation and analysis techniques to the next level, while an understanding of human behavior can help generate better predictive models and inform interventions so they are poised for success.
What will you do as a behavioral scientist in Health AI at Apple?
The ultimate goal is to make Apple devices a driver of healthy decision-making. Using AI and behavioral science, I’ll be working to help people make better health decisions.
What shall we call you?
Behavioral Science Lead, Health AI
Does Apple publish its work?
Yes! You can find some of the exciting health-related machine learning research here.
Will you keep up your writing and podcasting?
Is Apple hiring more behavioral scientists?
Though my team will likely grow in the future, we are not currently hiring. I will share any and all openings that I hear of on social media, and will also spread the word when the next AI / ML Residency Program is accepting applicants.
What kids song is stuck in your head right now?
So glad you asked!
How can I talk to you about life, work, etc?
I’m trying something new. On the last Friday of each month at 11am ET (starting July 29th), I’ll be hosting a monthly AMA to meet new connections, catch up with old ones, and just be generally helpful if I can. The invite is open to anyone and everyone, and I can't wait to see you there. Here’s the zoom link.