Information warfare and its limitations

Business Standard, 15 May 2023

The alleged exploits of Sanjay Rai Sherpuria highlight the place of information warfare in the modern world. Technological advances (WhatsApp, Facebook, Twitter) have created a whole new world of possibilities for information warfare which many bad actors are exploiting. Authoritarian states are able to meddle in foreign countries through these tools, even without collusion with local actors. These problems are about to get a lot worse owing to improvements in machine learning and large language models. There are many pathways through which things are improving.

In a recent article in The Print, Ananya Bhardwaj tells the story of Sanjay Rai Sherpuria, who allegedly created a public-domain information set about himself, which could be Googled by the credulous, and then used it to achieve outsized influence.

When miraculous powers in the form of a smartphone first landed in our palm, we were all delighted. Our early communications were with our trusted people, and it was human nature to apply that level of trust to everything that happened through the smartphone. Bad actors rapidly caught on to this, and a new age of information warfare began, aiming to make us believe things that are false.

Information warfare methods have been scaled up and industrialised. Thousands of workers are toiling, all across the world and round the clock, lying to us, pulling a world down in front of our eyes. As an example, a Russian firm named `Internet Research Agency' (IRA) was created by Yevgeny Prigozhin, who is now more famous as the military leader of the `Wagner' mercenary army fighting in many countries including in Ukraine. The IRA and other Russian agents have run campaigns to push the Brexit referendum in favour of `Leave', to push US presidential elections in favour of Donald Trump, etc. These methods could arrive into Indian politics.

An important insight from investigations into the Brexit referendum and the US presidential election is that there need not be collusion between the campaign and the apparent direct beneficiary of the campaign. For example, Vladimir Putin did not collaborate with Boris Johnson, who was keen on Brexit. Mr. Putin supported Brexit because he felt that it was in Russia's interests to engage in this asymmetric warfare, where tiny sums of money were spent on persuading voters, and a strategic Russian objective (degrading European unity) was fostered.

Here in India, the essential game of New Delhi has always been the portraying of proximity and then trading up. In the age of information warfare, many people have run internet-based campaigns to make it seem they are in favour. It starts as unsolicited proclamations of love. Once the campaign secures tangible milestones (photographs, retweets, other measures of proximity), these hard facts are used to trade up into bigger tangible proofs of proximity and ultimately into influence. Bombay knows less than New Delhi about who is in favour and who commands the power, and is therefore more credulous. Small investments in information warfare have yielded outsized returns.

How will information warfare play out in coming days, and how could we fare better?

  1. The media is splitting into an elite media supported by subscriptions vs. a mass market that is based on clickbait. The elite media (e.g. the New York Times) is more immune to information warfare. Even right-wing figures who revile `the mainstream media' turn to it for truth. In an age where Google searches yield bad information, there is a business opportunity to offer trusted information for a subscription fee. It is in our self-interest to buy subscriptions for information and opinions that are by authentic skilled humans, and to stop using mass media or social media. Households that don't pay for media subscriptions will carry a systematically different worldview, and suffer from more conspiracy theories.
  2. There is an arms race between information warfare and human defences. Information that arrived on WhatsApp was once treated as gospel truth. Now it is treated as something on a par with Twitter or the shouting on Indian television. In time, the people realise they are being gaslighted. There is less naivete in the masses about information warfare in India today than was the case about 10 or 20 years ago. Similarly, the Republicans fared poorly in all elections in the US after 2016, despite Mr. Putin's information warfare contributions to their cause.
  3. A powerful productivity gain for information warriors lies in modern machine learning software, which is able to generate realistic-looking text, images, and video, at a lower marginal cost. In India, there used to be a cost barrier in getting someone who can write fake text which would pass rudimentary muster. A lot of the fake information was identifiable through the embedded cultural markers. These protections are on the decline.
  4. These developments are not in favour of the internet giants such as Google. A lot of searches on Google or YouTube take you to information warfare content, which reduces the usefulness of Google. Similarly, when Facebook feeds are flooded with information warfare, it makes Facebook less interesting. Thus, the plague of information warfare is bad for trillions of dollars of market capitalisation. A good deal of brainpower is thus working on fixing these problems.

It looks bad, but it is not all gloom. The internet will be islands of LLM-free knowledge floating in a sea of derivative sewage. The value of trusted sources -- ranging from the mighty New York Times to the obscure The Leap Blog -- will go up. Grandmothers reading WhatsApp will steadily become more sceptical of taking content from strangers. Deepfake tapes will evoke less excitement after the first one million of them. Google and Facebook will get a bit better. The best days of information warfare are likely to be behind us.

Back up to Ajay Shah's 2023 media page
Back up to Ajay Shah's home page