Technology doesn’t solve problems; intent does

I went to a startup conference over the weekend (24 June) and chatted up a host of familiar faces. Many conversations inevitably turned towards Aadhaar. I heard two recurring sentiments:

1. Aadhaar is fundamentally a good idea, whatever the current problems may be.
2. It is a mistake for tech entrepreneurs to get involved with government.

I respectfully disagree with both sentiments and I’d like to explain why. Be warned, tech jargon ahead.

Technology is only a tool

We in the tech community recognise this implicitly when it comes to our own work. Take JavaScript for example: over the last 15 years, the language has transformed from being a gimmick into a serious ecosystem that underpins much of how the modern web works. We even remember the most important milestones of this transformation:

1. The first rich web apps, Gmail (2004) and Google Maps (2005)
2. API standardisation with jQuery (2006)
3. High performance JS runtimes with the V8 engine (2007)
4. Server-side JS with Node.js (2009)
5. Modern front-end frameworks, AngularJS (2010) and React (2013)

As a tech person building web apps, what did you get when each of these milestones occurred?

1. Did Gmail and Google Maps make your website a rich web app? Nope, they only offered proof that this was possible. You had to figure out how to do it for yourself.
2. Did including jQuery in your website automatically make your JS better? Nope, you also had to use jQuery. It made coding easier, but you still had to write the code.
3. Did Chrome give your website users a free performance boost? Nope, as a new browser, users weren’t using it, and their current IE or Firefox browsers didn’t magically get a speed boost from Chrome. Users had to consciously switch to Chrome, or wait for their browser makers to upgrade their own JS engines. You as a website developer couldn’t afford to sit back and wait for the world to catch up—unless you simply didn’t care for your competition.
4. Did Node.js give your front-end developers an automatic upgrade to back-end developer status? Nope, the language is just one piece of the puzzle. There was an entire new ecosystem to learn about; a dizzying pace of development that meant everything you knew was obsolete every few months; fractious community relationships that resulted in the IO.js fork and subsequent re-merger; and continued drama and instability with the NPM package manager. Node.js is great, but it only works if you’re an experienced developer who can keep up with all this.
5. Adopted AngularJS because it was backed by Google? Too bad they messed up the API and had to release a backwards-incompatible 2.0. The years you invested into building websites with Angular 1.0 are now your liability to upgrade. Feeling smug because you backed React instead? The Vue.js developers are smirking from their corner.

Let’s face it. Technology is not magic pixie dust that you sprinkle on your problems to make them go away. To solve a problem, you must have the intent to solve the problem, and back it up with sufficient experience and resources. Technology is merely a tool in your resource toolkit. It may be a useful or useless tool, an expensive or cheap tool, or a substitute tool because your preferred tool isn’t available. But it’s a tool in your hands. You matter more.

You didn’t need this lecture. You knew this. Why, then, must you elevate Aadhaar to the status of technology that will magically enable inclusion, solve corruption, and take India into the league of global superpowers? Shouldn’t these things require intent by responsible officials, in whose hands Aadhaar is a tool that may or may not be useful? If, as the refrain goes, there has been no progress for 70 years, exactly how does Aadhaar instill intent in the hearts of apathetic officials? If venture capital can’t turn any idiot into a successful entrepreneur, how will Aadhaar solve any problem by itself?

In 2017 India, Aadhaar has become a hammer in the hands of central government that makes every problem look like a nail. At startup conferences, prospectors have their eyes peeled for an Aadhaar-enabled startup that will become the next unicorn.

As the eminently quotable Jamie Zawinski once said:

Some people, when confronted with a problem, think “I know, I’ll use regular expressions.” Now they have two problems.

Try s/regular expressions/Aadhaar/

“But I know all this. Aadhaar detractors like you are spreading fear and confusion and hampering responsible officials from using Aadhaar effectively”

Let’s examine this sentiment with one popular use of Aadhaar, for the Public Distribution System (PDS) that distributes subsidised food rations. As is well known, PDS is plagued with inefficiencies and leakages from decades of shoddy implementations.

Under India’s federal structure, almost all citizen interactions are handled by state governments. The central government is—to use technical parlance—a backend services provider that redistributes tax money to support states that need assistance, offers a central forum to resolve disputes, provides common services that all state governments need, and handles foreign relations. Day-to-day life is almost entirely the responsibility of the state government, and that includes the PDS.

Given this structure, whether PDS works or not is a function of how well the state government is organised. The South Indian states report eliminating leakage to over 90%—and that’s before introducing Aadhaar—while Rajasthan has a series of horror stories of people excluded by Aadhaar. In an efficient system, Aadhaar is yet another tool that may improve efficiency, used at the discretion of officials. In a weak system, Aadhaar is a blunt instrument that allows officials to blame the technology’s failings to deflect attention from themselves.

These public officials are not Aadhaar’s detractors on Twitter and in newspaper opinion columns. Those are concerned citizens pointing out the technology isn’t magic.

If PDS is a state subject, and it’s up to state-level authorities to decide if Aadhaar is a useful tool or not, why is the central government pushing for mandatory use of Aadhaar for PDS? What magical results do they expect by imposing a tool? Why does this idea—that Aadhaar is magic—exist? Let’s turn to familiar web tech again.

Technology is often marketed as magic

Look at the website for any popular web framework. Look at how simple they make everything seem. Here are a couple extreme examples (and corresponding criticism):

From the Sinatra and Flask web framework home pages. These two are interesting because they’re often compared

Technology marketeers have perfected the art of convincing you that their technology is the solution to your problems. They have beautiful websites and careful attention to onboarding detail, with every effort made to ensure a customer is retained and cohort charts look better than ever.

They only need to keep you on the ride until you’re too committed to walk away, and then you’re coopted into supporting the dream, whether by patronage in the form of fees or code contributions, or by preaching the gospel to the unconverted. “Yes, there are problems,” you’ll say as a fresh convert, “but we’re better than the other, and if you join us we’ll get even better.”

As tech folk, we hate this phase of any technology. Will this lead to eventual glory, as with Node.js, or head into the gutter as with AngularJS 1.0? We don’t know yet. We’ll attend conferences and pay critical attention to sermons of the converted. We’ll ask tough questions to tease out the inevitable gotchas. We do this because we’ve been burnt by half-baked technology before. As insiders in the tech industry, we don’t buy into the dreamy marketing we produce for gullible outsiders.

Aadhaar is no different, and they found a gullible customer in government. The dream of good governance using Aadhaar has somehow morphed into the reality of governance using Aadhaar—alone.

The makers of Aadhaar are not fools. They don’t believe it’s magic. Occasionally they’ll cry out in public that they too have been asking for a privacy law all this while, and their work can’t be blamed for the lack of said privacy law.

This leads one to wonder: are the people in government who bought this dream the fools, or did they believe they were buying something else?

What they bought

“We made this powerful technology. Just like nuclear, if misused this will be dangerous like a bomb, but we designed it carefully to only be used for peaceful purposes. Because we care.”

“A bomb, you say?”

“No! This is only a truth machine. If you speak to it, it will tell you whether the statement is true or false. It will not say anything else. We even put a blindfold on it, so it doesn’t know anything about who is asking and why they want to know. It is dangerous if the machine knows too much.”

“Noted. We’ll keep the blindfold on like you asked, and make it mandatory for everyone in this country to speak to this machine in front of us.”

The blindfold. Federation via ASAs and AUAs is supposed to keep you safe as no single entity has a comprehensive profile of your activities. But watch out for dominant providers, consolidation, or Sub AUAs leaking their identity—and thereby your activities—via registered devices. A related project, the State Resident Data Hub, is an entire other can of worms: it’s effectively Aadhaar minus the protections of the Aadhaar Act

Click/tap through for the full thread

UIDAI can be blind to usage, but so what? Build an all-seeing wrapper around UIDAI

Coda

Fellow denizens of the tech industry: wake up. The portal has been opened. Aadhaar won’t go away if you pretend it’s not your problem. As people who understand how this stuff works, it’s now imperative on us:

1. To preach responsible use of technology as a tool, not solutionism.
2. To acknowledge that we know far too little about how the world works.
3. To listen more. Engage with people from other domains. Learn from them.
4. You build tech to empower yourself and your customers. That power is real. Be aware of how you’ve upset a delicate balance of power in society when you deploy your technology. You are responsible.
5. Don’t drink your own kool-aid like these folks. The technology isn’t magic. It will have flaws like every other piece of tech. To have blind faith is to do society a disservice. Question it. Demand accountability from the promoters.

Addendum

Kentaro Toyama, formerly of Microsoft Research India and one of my inspirations, said it succinctly back in 2001:

In project after project, the lesson was the same: information technology amplified the intent and capacity of human and institutional stakeholders, but it didn’t substitute for their deficiencies. If we collaborated with a self-confident community or a competent non-profit, things went well. But, if we worked with a corrupt organization or an indifferent group, no amount of well-designed technology was helpful. Ironically, although we looked to technology to attain large-scale impact into places where circumstances were most dire, technology by itself was unable to improve situations where well-intentioned competence was absent. What mattered most was individual and institutional intent and capacity.