legacy

I’ve been following with some interest the story of Facebook’s Free Basics endeavor, which was recently shut out of India.  The notion is that Facebook would offer free but heavily limited internet access to developing countries, the rationale being that something is better than nothing in a country like India, where internet penetration is still only at 15-20%.  I will now write briefly about why this is a terrible idea.

First, let’s assume this is actually a philanthropic endeavor.  Sure, Facebook does stand to gain tremendous power and capital as the gatekeeper to millions – if not billions – of internet users, dictating literally all of the possible websites available.  While that’s a pretty good reason to distrust the operation on its own, it doesn’t speak to the quality of the plan itself.

There’s a pattern that I’ve begun to notice, partly through my work, but particularly as I spend more time studying history.  Foundations never really go away, and what becomes foundational is extremely hard to predict.  Change is really hard, especially once a system becomes part of the fabric of society.  Most things aren’t made with any understanding of how they’ll be used or how long they’ll last.  By the time we realize the failures and shortcomings of an institution, we’re stuck with it.  In software, we call these legacy problems.  Limitations, unsolved issues, or simply shoddy work inherited from our forerunners.

This is a bit abstract, so allow me to provide a few simple examples of problematic legacy systems.

  • US customary units (feet, pounds, ounces).  Congress actually passed the Metric Conversion Act in 1975, but it had little enforcement and no one cares that much.  We’re stuck with ’em for the foreseeable future.
  • The electoral college.  Originally created to solve the problems of voting prior to electronic communication, it persists primarily because there isn’t a strong enough incentive to remove it (and it might be advantageous for some).  Of course, there’s an argument to be made that the entire Constitution is a legacy problem, but that feels a bit too bold and too broad for this topic.
  • NYC’s entire subway system.  As examples go this feels like the lowest of hanging fruit, but it’s probably the most apt.  The whole thing was originally built by 3 different companies, which is why the routes are so complex.  A tunnel gets blasted out, that’s final.  You can’t just move it 5 feet over.  Then there’s the parts and pieces of the operation – ancient cars, switches, and cables that can never be replaced one at a time.  Here, it’s not a lack of demand that keeps things the same, but the difficulty of change (and perhaps a bit of cultural inertia with some mismanagement).

Rarely do we get to pick and design our societal foundations. When you get to start from scratch with sufficient resources and good knowledge of the problems to solve, that’s a huge opportunity.  A chance to anticipate and guide change, to establish a baseline of quality and affordability and availability. To set the standard for the future.  To give the next generation something to stand on, rather than another problem to fix.

Zuckerberg’s argument for Free Basics is chiefly that something is better than nothing, which is a garbage modus operandi for anything that’s meant to last.  Something is what you pick when there aren’t enough resources, when you don’t know what to do or how to do it, when there’s just no other option.  None of that applies here.  It all just seems painfully short-sighted.

Universal internet access is a must in this century.  There’s just no question of its importance or necessity.  But Facebook is offering the least useful, least helpful version of the internet to developing countries.  From the moment of its installation, Free Basics would become a legacy problem.  Something people would, on day one, hope to leave behind, to replace or upgrade. Why would anyone want to start with the shittiest possible solution?