Research Philosophy

Why do we do research?

1. because the world is interesting: we are a curious species and we, collectively, cannot keep ourselves from knowledge building.

2. because the knowledge we produce can sometimes be deployed to effect positive change in the world

Why do I do research?

Because I have been given the opportunity and - through the investment of my parents/teachers/advisors - the skills to convert indulging my own curiosity into a career.

What about affecting positive change in the world?

While I am indeed hopeful that my work contributes to positive change in the world, I don't hang my motivation for research on this point. The impact of a new idea/technology only comes via its deployment into the world - and that process is sometimes unpredictable and also largely out of my control:

1. New technology sometimes means new problems - the existence of a profitable and beneficial use-case today does not guarantee a net positive impact in the long term.

2. Deployment is less-well tethered to technical merit than we might often like to believe - there are social and political stress tests that can fail even where technical success is achieved.

Can new technology make the world better?

Of course - the history of human innovation is nothing but a chain of embarrassment and humiliation for Luddites and Malthusites. But we should caution ourselves that (a) past performance is no guarantee of future success, and (b) not every idea is embraced purely on the basis of its merits - we as a species can be short-sighted, and solution-aversion is a very real phenomena.

To realize the potential of a technology that is being developed - or has already been developed - I think that scientists have important roles to play as advocates outside of the research setting:

1. We need to advocate the technical merits of ideas that failed to deploy for purely social/political reasons, even if the motivation for our own research is predicated on viewing those social/political barriers as insurmountable. We can certainly accept social/political obstacles as realities outside of our immediate control, but as it pertains to our research and outreach we should not predicate our work on validating social/cutural entitlements in the status-quo that are unsustainable and/or inequitable.

2. We need to cultivate an interest in the deployment process itself, paying careful attention to what industry experts and scholars in the humanities are saying about the problem.

Thoughts on Collaborating with Industry

I worked as a Tech-Aide at 3M for three years during my undergrad, and my postdoc work has been done in close collaboration with Unilever. I am very comfortable working with industry, and it is my view that collaborations between academia and industry can be extremely productive for a number of reasons:

1. For industry, in-house R&D expertise is expensive and therefore focused on areas of core competency. Academic collaborations are a comparatively flexible and inexpensive way of contracting valuable expertise (and labor) outside of that core competency.

2. In academia, research groups are often revenue-limited, and funding from industry collaborations allow otherwise un-tapped ideas to be fully explored.

3. For students headed to industry - and for the companies funding the collaboration - these projects can create a productive hiring pipeline with a more seamless transfer of desirable skills.

4. For students headed to academia, collaborations expose students to some of the real-world limitations that influence the selection of research topics: regulatory shifts, anticipated changes in supply/demand, and the challenges that one encounters during scale-up.

Thoughts on Experiments vs Modeling

During my time with 3M - prior to my PhD - my work was primarily experimental in nature.  I enjoyed that work and, I dare say, I was pretty good at it too.  But I think that theory and modeling work come more naturally to me.  I enjoy working with experimental collaborators, and I am confident in my ability to support an experimental project if the need were to arise.  For as much as I would personally like to develop both areas of expertise, it is my view that the world today primarily works through specialization and collaboration.

Thoughts on Machine Learning and Data Science

Machine learning and Data Science are extremely powerful tool for identifying patterns in data sets that are too large, too noisy, and/or too abstract for human comprehension.  However, in my area of research - constitutive modeling of complex fluids -  machine learning tools seem somewhat reckless.  My work involves problems that (with a few possible exceptions) do not seem too large/noisy/abstract for human comprehension - but the outputs from machine learning are themselves inscrutable, and there is no way to gauge the limitations of the model that results.  In other words, there is no a-priori guarantee that predictions will be accurate for fluids/flows outside of its training set - the very conditions where a model becomes most useful.

For my field, I think that the most interesting research directions for machine learning will focus on bridging the gap between the microscopic degrees of freedom that overwhelm us (e.g. from particle-based simulations) and the mesoscopic abstractions that we already make use of.  This could help us glean insight to important questions like "what is an entanglement?", "what is the tube diameter?", "how does chemistry define a packing length?" and so on.  In effect, I think that these powerful tools can help us interpret particle-based models and then we can find ways to recast those interpretations as PDE descriptions, which is where my research has been focused thus-far.