Oil and Water

There is an adage that says oil and water don’t mix.  If you fill a glass halfway with water, then pour in a little olive oil, and give it a good stir, you’ll end up with a lot of oil globs that will eventually rise to form a layer that comfortably rests on top of the water.  It happens every time, no matter how much you stir.

I mention this as it comes into play when a company is looking for some capabilities and support for advanced AI research.  The first, and often one of the best, places to turn to is academia – those leading university researchers that know what’s coming next.  But Industry-Academia collaboration is a challenge where each is motivated by different things:

• Universities are driven by “Publish or Perish”, where researchers need to publish in journals or present at leading conferences their latest results to keep ahead and be recognised as top in their field.  Industry needs to have a “Profit or Perish” mentality, or else they simply no longer exist.

• The timelines are vastly different as academics look to what will happen in the next couple of years, whereas companies often look to their next quarter bottom line.

• The fundamental principle of academic research is that it is open and shared.  Corporations want to keep ideas protected and consider IP (Intellectual Property) a valuable asset. 

• It is often said that research is about general solutions looking for problems to apply to (“Tech Push”), whereas developing products for the marketplace is about problems that are looking for specific solutions (“Tech Pull”). 

It’s oil and water.

My experience over the many years of supporting hundreds of industry-university collaboration projects has shown that it can work.  If you get it right, there is a win for both sides, and you can make leaps and bounds forward that you otherwise could not think possible through these unique collaborations. Here are some commonalities and incentives that can create mutual motivation and benefit:

• Master’s and PhD students are the heart of a professor’s lab.  And students love to work on real-world problems.  If companies can divide their ideas into one- or two-year initiatives that a student can focus on, it can be rewarding.  Also, these students can be great hires when they graduate.

• Consider funding both long-term research and quick consulting projects that researchers can apply what AI architectures and algorithms they have readily available.  It’s an excellent way to learn how to work together quickly and support research that can make substantial advances over time.

• Researchers are often faced with having to use “canned” datasets that everyone in the academic world has access to.  One of the things they continue to ask for is real-world data – and companies have that! 

• AI specialist academic researchers are often found in leading machine learning departments of Computer Science faculties – and they are in high demand!  Look to other departments and faculties that are now beginning to work on AI research that is applied to specific areas – such as engineering, medicine, science, and even the humanities. 

• Industry is often looking to “upskill” their engineers, computer scientists, and data scientists with the latest AI understanding.  Consider having small 4 to 6-month capstone projects where your in-house talent works on a university researcher lab problem.  Your researchers get to work with some of the best academics.  At the end of their “secondment” time, they can come back and implement the project in your products. 

• Many universities have very fair IP policies that usually are 50:50 – a shared approach that is a win for both. And, consider allowing some of the results of joint projects to be published in journals or presented together at conferences – success for academics and your staff.

• There are many government funding programs that match industry dollars for projects with universities.  A win at no extra cost.

• Remember that “Proximity Matters!”  The closer industry and academics are, physically and collaboratively, the better the result.

These ideas are what I call “emulsifiers”.  It’s like egg yolks or mustard.  When you add them to oil and water, and shake them lots, the result is that everything mixes exceptionally well.  And, if you add a little spice, you can get some fantastic salad dressings.