The importance of open source software (OSS) in university-industry collaborations is on the rise – generating an ever growing two way flow of advanced code and high tech talent, spinning up a wide range of startups as well. The best practice open source communities has three parts: (1) open access to code, (2) a commercially friendly license, and (3) open governance in a neutral third-party foundation that specializes in catalyzing and nurturing OSS community growth.
Recruiting & Readiness:
Established firms and startups study the GitHub profiles of candidates to hire the best. For example, the hottest technologies are open source in areas such as Artificial Intelligence/Deep Learning (TensorFlow, PyTorch, ONNX), Containers (Kubernetes), Data Science (R, Python, Spark), Graph DB (JanusGraph), and Blockchain (HyperLedger).
Research & Regional Entrepreneurship:
In addition, candidates who are Kaggle Masters and/or who have experience in submitting open source solutions to top algorithm challenge leaderboards may have multiple offers even before graduating – is that good or bad? Both corporate and university teams compete on leaderboards, such as SQuAD for question answerinq (i.e., https://rajpurkar.github.io/SQuAD-explorer/). A growing number of challenges include a student section on a subset of the data, such a CVPR Moments challenge (e.g., https://winstonhsu.info/winning-third-place-in-cvpr-2018-video-recognition-challenge-moments-in-time/). For example, IBM recently ran the Call For Code, and developers from around the world submitting solutions of how AI systems can help in case of natural disaster – see https://callforcode.org/global-prize/
Who has Open Source Software (OSS) programs today?
A small sample of noteworthy university OSS centers include: 1. RPI RCOS (https://science.rpi.edu/computer-science/programs/undergrad/bs-computerscience/rensselaer-center-open-source-rcos) contact: Wes Turner (RPI) <turnew2@rpi.edu>, 2. Berkeley RISE: (https://rise.cs.berkeley.edu/), 3. UC Santa Cruz CROSS (https://summerofcode.withgoogle.com/organizations/5730466942943232/), 4. Stanford (https://opensource.stanford.edu/) – and DAWN: https://dawn.cs.stanford.edu/, 5. MIT (http://mitopensource.sourceforge.net/) – and MIT-IBM Watson Lab see: https://mitibmwatsonailab.mit.edu/, 6. OSU Open Source Lab (https://osuosl.org/) – and IBM PowerAI see: https://developer.ibm.com/linuxonpower/cloud-resources/
Role of UIDP
UIDP (https://www.uidp.org/) role might be to provide an improved overview and help answer these basic questions: 1. Which university programs exist and what are the best practices for curriculum, interns programs, etc.?, 2. What are the best practice in setting up and runnings challenges, across university, industry, and government challenges?, 3. What government programs exist in the USA and around the world to enhance open source collaborations?
IBM and Open Source AI
CODAIT (Center for Opensource Data and AI Technologies) – see http://codait.org – is an example of an industry-based open source center seeking university and industry collaborators on projects that improve the ethical use of AI in enterprise AI solutions. For an overview, of CODAIT please watch this 15 minutes video, by CODAIT Chief Architect, Dr. Frederick Reiss, during his engaging recent ApacheCon keynote – see https://feathercast.apache.org/2018/09/26/from-adam-to-zara-enabling-the-ai-lifecycle-for-your-enterprise-using-open-source-fred-reiss/