System Efficiency vs. Fragility

I just listened to Padmasree Warrior on a Commonwealth Club podcast. Ms Warrior is the chief strategy officer at Cisco. The podcast contains lots of interesting insights and projections (as well as nonsense, specifically that Google was the first search engine company and capitalized on a first-mover advantage). With Cisco being about all things interconnected, she jokingly suggested that at some point in time, every human being could have an unambiguous ID, e.g. an IP address. This would simplify life so much given all the password problems we are having.

Privacy concerns aside, this vision is also difficult from an engineering perspective. Ms Warrior’s thrust is the idea of efficiency. So many things would become so much easier if every human being was unambiguously identifiable. We could build better, faster, cheaper technology and advance humanity through it. However, efficiency breeds fragility. When we look at how humanity got where it is today, I see an abundant amount of redundancy that creates robustness. Taking out redundancy will make things much less robust, as we are creating single points of failure.

I see the value of unambiguous identification but not through intrinsic properties attached to humans. There must be other, better solutions, that don’t make our world more fragile than it already is.

This Curious Word 3D Printing

When I first heard about 3D printing, I thought it was a weird choice of words. It is a good word, because it tells a story familiar to people living today, but why not call it home manufacturing or on-site factories, or the like? I’m not sufficiently knowledgable about the history of technology, but is it possible that the printing press was really the first repeatable manufacturing process? We have always build machinery, but did we ever before the printing press have a repeatable process? If not, then this naming is apt.

Assumption about Longevity and Its Consequences

If you have run into me recently, I may have bugged you with the following question:

Given the rapid pace of development in medical technology, I expect my generation to live to 100 years of age. A child being born today may live to the age of 250 years of age. Under this assumption, what health issues do I need to watch out for most to achieve that age?

I have little scientific fact to backup the assumption; it is based solely on my perception of the acceleration in medical technology today. Once you make the assumption that the average life span may be growing rapidly, you start to wonder how to take advantage of it. Or, put another way: What are the parts of your body should you be caring for most?

For example, I see three layers:

  1. Mechanical stuff. If you have a bad knee, I expect that this will be fully fixable within the next 10-20 years or so. It seems to me to be a purely mechanical issue.
  2. Systemic stuff. More difficult to fix, if anything goes wrong, are systemic issues, for example, arthritis or a bad lung. It is not clear to me how easily this can be fixed.
  3. The brain. At the high end sits the brain. Things that can go wrong are illnesses like Alzheimer or Parkinson, but also loss of energy to live. How to avoid those?

These are all hypotheses, but the question is real. What are the most difficult things for medical technology to tackle and how to avoid that they’l become a problem once we are starting to live longer and longer lives?

The Five Stages of Open Source Volunteering

Abstract: Today’s software systems build on open source software. Thus, we need to understand how to successfully create, nurture, and mature the software development communities of these open source projects. In this report, we review and discuss best practices of the open source volunteering and recruitment process that successful project leaders are using to lead their projects to success. We combine the perspective of the volunteer, looking at a project, with the perspective of a project leader, looking to find additional volunteers for the project. We identify a five-stage process consisting of a connecting, understanding, engaging, performing, and leading stage. The underlying best practices, when applied, significantly increase the chances for a successful open source project.

Keywords: Open source software, open source communities, volunteering process

Reference: Dirk Riehle. The Five Stages of Open Source Volunteering. Friedrich-Alexander-Universität Erlangen-Nürnberg, Dept. of Computer Science, Technical Reports, CS-2014-01, March 2014.

The report is available as a PDF file, on FAU’s OPUS, and as HTML.

Upcoming Public Talks on Open Source

Here are three upcoming talks on open source for the next six months in 2014:

I wish I was more consistent in announcing public talks…

The Economics of Financing Ph.D. Students as Company Contractors

Occasionally companies approach me with the following proposal: If I’m willing to supervise one of their employees for an external Ph.D. thesis, they’ll pay into my University budget an annual lump sum, typically something like EUR 10000. I almost always reject such proposals, unless I can change some of the critical terms, because these proposals are highly problematic. To understand this, please follow along.

The company does the following math: They’ll hire someone with a recent Master’s degree, typically for a research project at the company and on a contractor basis. Then, they’ll promise the contractor that he or she can use some of their time to complete a dissertation, because they argue the project will provide enough research substance. To prove this, they’ll use the professor to confirm to the contractor that they will take them on as a Ph.D. student. A going rate for such type of contractor is (a surprisingly low) EUR 2000 per month. Times twelve months + the professor’s lump sum makes EUR 34000 per year for the company (ignoring company overhead). The official cost of a Ph.D. student at a Bavarian University is EUR 75000. Voila, the company just saved EUR 41000 a year (ignoring other University costs). However, the contractor is much worse off, because no social duties are being paid for them.

Continue reading