The TikTok saga and why the net must run on democratic ideals5 min read . Updated: 21 Sep 2020, 08:35 PM IST
India might have a good opportunity to lead the world in ushering in a golden age of a democratically-controlled internet
As I write this, TikTok has narrowly averted being banned from app stores in the US for at least another week. The White House demanded in early August that the popular short-video app be sold to American owners by the weekend, and while Microsoft and later Oracle turned up as suitors, such a sale did not go through. In a last-minute reprieve, US President Donald Trump gave his “blessing" to a deal for 53% of the TikTok company to be sold to a consortium of Oracle, Walmart and others.
Many see the TikTok drama in the US as Trump’s personal revenge against an app whose users had wrecked a rally of his in Tulsa, Oklahoma, in June. The week before that, Trump had tweeted: “Almost One Million people requested tickets for the Saturday Night Rally in Tulsa, Oklahoma!" One local official said over 100,000 supporters would show up near the arena. But come rally day, registered attendees didn’t fill the venue at Tulsa’s Bank of Oklahoma Center. Empty seats at the rally became a laughing stock and deeply embarrassed the Trump campaign.
A coordinated effort was underway on TikTok in the days leading up to Trump’s rally, encouraging people to register online for the event and not show up. Until that rally, TikTok was seen as a platform for teenagers entertaining one another and not as a tool of political action. Ever since Trump threatened to ban the app (and China’s WeChat) from app stores in the US, advocates of internet “laissez faire" have been complaining loudly.
But too much is being made of Trump’s alleged motives by the tech world’s laissez faire camp. To insist that TikTok and other Chinese apps must be left alone by the US and Indian governments is ludicrous. Countries have the right to take justified action, especially when it comes to an app from a country that has violated so many basic norms of the internet.
The US drama follows events in India, where the app was banned in June, and where the sale of its assets (read its user base) has proven thorny. Mint reported earlier this month that technology transfer curbs by China and the rise of rival platforms in India could make any acquisition of TikTok India a tough affair.
Mint had reported that SoftBank, a minority investor in TikTok’s parent company ByteDance, had begun talks with potential partners to jointly bid for the India unit, which arguably had the app’s largest user base anywhere in the world. Meanwhile, Chinese rules that prevent homegrown companies from transferring technology algorithms to foreign entities could come in the way. This would prevent the transfer of its successful content recommendation algorithm, the one behind its “For You" page, which presents personalized content for each user by using text analysis, voice recognition and other techniques to hone its recommendations.
I argued at the time that the real question was not necessarily where the algorithm resides. In fact, it is possible to reverse engineer “black box" algorithms by studying their output. Researchers are now proving repeatedly that while black box engines may be inscrutable, they aren’t above being reverse-engineered; not unlike pharmaceutical drugs whose individual active pharmaceutical ingredients (APIs) can be broken down and reconstructed. Interestingly, in the artificial intelligence world, this is accomplished by a term that shares the acronym API, which in computer speak is short for “application programming Interface".
A few years ago, a team of computer scientists at Cornell Tech in New York City, Swiss Institute Ecole Polytechnique Federal de Lausanne and University of North Carolina published a paper titled Stealing Machine Learning Models via Prediction APIs. APIs are built into a computer application to let programmers and other applications access it. These researchers found a way to create their own artificially intelligent interface for a black box and then use this box’s output to reconstruct its internal workings, thereby reverse-engineering it.
The laissez-faire approach to data privacy has allowed the growth of business models based on attention capture, surveillance and user opinion modification. Indifference for the past two-and-a-half decades to what happens on the internet has destabilized political systems, even allowing for foreign interference in American elections.
If tech laissez faire has run its course, what are the alternatives? The first is to ape China and impose net nationalism, which makes the state the predominant authority over all things online. This will inexorably lead to the global internet splitting up into several national ones. The clear and present danger is that such national internets can become instruments of state power. They might end up serving as a means of disseminating state propaganda, monitoring dissent and furthering crony capitalism.
Tech-nationalism is not the only alternative, though. We also have democracy, which holds that matters of public importance should be decided by the people—and that people should control the excesses of both private and government power. We have laws against child labour, for instance, and against the excessive use of force by police authorities. Similarly, when it comes to the internet, democratic principles hold that legitimate governments can make rules so long as these serve the interests of the people.
This is the concept of tech democracy. The only justifications for any non-democratic intervention would include national security, the defence of institutions, the preservation of markets, the need for retaliation against a belligerent foreign power, and so on. But it is the interests of the public and not the whims of a leader, nor the interests of corporations, that should guide us. Getting the tech policy balance right is a key challenge for democratic governments around the world.
How we could do this in an orderly fashion is beyond me, especially given that Big Tech internet companies—both Chinese and American—have now become powerful enough to sway public opinion. But I take heart in India’s approach to net neutrality, where the Telecom Regulatory Authority of India, after a public referendum, put an end to a thinly-veiled attempt at stifling net neutrality during the Free Basics fiasco back in February 2016.
While America’s TikTok drama rages, maybe India has a chance to lead the globe in ushering a golden age of a democratically-controlled internet. The next step should be to look within our own borders, where dangerous dragons also lurk.
Siddharth Pai is founder of Siana Capital, a venture fund management company focused on deep science and tech in India