It is unclear to me if the explosion of Internet access has shown the dark side of humanity or it has turned us into terrible people. If you remember the Internet even as recent as 2008, it was not the case. There was no organized mob doling instant murder and rape threats to anyone who chose to voice their opinions, there was no ‘darknet’ marketplace for the vilest things possible, not every online message you received was fake news, foreign interference into elections of an entire large country was unheard of. Community platforms like Reddit provinging semi-anonymity are increasingly seeing toxic behaviour from people. It is as if people just go crazy when they feel no one can tie their posts to their public identity.
At a previous job back in 2013, I got curious about who these people abusing everyone were and their interaction graph, tweeting patterns, and related analytics on Twitter. It helped that it was a media startup and I was able to run my experiments just before the general elections with the Twitter streaming API collecting Gigabytes of data on multiple AWS instances to be processed later. One of the key insights was the graphs for the two polarities were quite separated (though had an overlap, you got to pay those bills) and the interaction graph could easily be used to figure out the political affiliation of any ordinary Twitter account to a great accuracy. I was also able to identify more than 25,000 bot accounts solely made for spewing political agenda.
Democracies in the time of ‘Whatsapp Elections’
The Internet has provided a cheaper way to reach the masses instantly. Coupled with data and analytics around demographics that can be targeted with fake news for manipulation, this is a much better medium for promoting one’s political agenda. The Wall Street Journal’s revelation this month that Facebook overlooked their hate speech policy to allow politicians of a certain political party in India to spread hate speech is unsurprising. This comes at a time when Zuckerberg wrote a note in June this year claiming to better clamp down on hateful content. As a public company, it is expected that Facebook will do everything it can to maximize value to its shareholders. This is the time where morality and ethics have been relegated to electives even in universities. Thieves and dubious behavior like Theranos and Cambridge Analytica has been renamed to hustling in the startup culture. This is not the first time that Facebook has been involved in something like this. Just a couple of months ago, Facebook had to fire an employee for exposing its preferential treatment to posts towards a certain political inclination. And this is not a problem for Facebook alone.
Closed social networks like Whatsapp groups have been analyzed in the past to have significantly influenced elections in Brazil, India, and Nigeria in the past. The spending on IT cells by Indian political parties increased more than 10 folds from 2014 to 2019. These groups provide cheaper massive reach in a shorter duration, data on demographics to target and manipulate with fake news. A certain political party in India has more than 30 lakh ‘volunteers’ that are part of various Whatsapp groups ready to share and promote whatever content comes down from their overlords.
Twitter has constantly downplayed the number of spammy bot accounts it has on its platform. Let’s just be real for a moment. Identifying bot accounts is not a very hard problem for Twitter. What is needed is the willpower to reduce the count of users, maybe by 15% by some estimates.
How altruistic of companies like Facebook to have attempted programs like ‘Free basics’ to provide restricted Internet access to their platforms in developing nations like India. And Facebook is not alone. Everyone wants to be in the Internet provider business. The recent massive investments in Jio platforms is an indirect solution to the same problem.
At this point, I have become so used to fake news that I tend to verify anything dramatic before believing it. At some point in the immediate future, this verification might also become difficult. I am sure that Deepfakes, neural voice cloning and GPT-3 as the natural successor to manual photoshopping and content generation would be mainstream by the next set of elections in five years to peddle agendas. With reupted sources going behind paywalls, all the fake news portals might remain as the only sources of free information. Fake news has led to multiple lynchings and innocent people losing their lives, agenda that have promoted non-scientific information leading to loss and elections that have been indirectly influenced by purely maniuplate strategies.
The Internet has become a terrible place. And unfortunately, it might just be beginning to get worse.
Other things that matter
A significant portion of the Internet experienced an outage a few hours ago, even disrupting the Chess Olympiad finale happening today and forcing FIDE, the official chess body to announce India and Russia as the joint winners. Slack’s desktop app had a remote code execution exploit reported almost six months ago, disclosed this week. As an expected step in the Apple-Epic fight that I wrote about in the last post, Apple has terminated Epic’s game developer account on its platform. Musk’s Neuralink came out with an interesting progress video (It involved a pig too.) Would be interesting to see the future security implications of an embedded BCI device that can communicate via Bluetooth or similar wireless tech.