Deepfakes herald the end of mass-media and the public internet as we know it. Can we reclaim truth on the internet or is it too late?
“You are free to be a drunkard, an idler, a coward, a backbiter, a fornicator; but you are not free to think for yourself.”
– George Orwell
The airwaves are buzzing after Trump decided to ban TikTok in America: everyone has realized that the internet is ready to be carved up into separate pieces. In the decades to come, we will see that the Chinese firewall was the shot heard around the world; every region – Eurasia, America, India, Russia, and Africa – will have their own version of the internet.
Likewise, long-time subscribers may remember my piece about Video is Eating Software and how video will consume everything because it’s higher bandwidth – text loses its luster by comparison. Yet, what if these videos are fake as many increasingly are? Will we have the capability and courage to discount what our primate brains want us to believe? I don’t think so.
“The confidence that individuals have in their beliefs,” as Daniel Kahneman wrote in his groundbreaking book, Thinking Fast and Slow, “depends mostly on the quality of the story they can tell about what they see, even if they see little.”
Therefore, the proliferation of deepfakes and bots encompassing the internet are worthy of our attention. Is there anything worse than the internet breaking up? What if, drastically, the internet loses all its signal and is, instead, deluged with noise from these deepfakes?
Don’t believe me: watch this recent commercial aired during ESPN’s documentary series – The Last Dance – what do you see?
(Let’s be honest, you were probably convinced it was real until he mentioned that “it’s gonna be lit.”)
For good reason, this commercial from State Farm has become one of the most hotly debated commercials in recent history (spoiler: it’s a deepfake and not many people could tell!) The public is revolting.
By pattern-matching our behavior, deep-learning algorithms understand how we’re going to act before we do so. How have deepfakes gone from an amusement to the harbinger of the digital apocalypse in only three years?
If everything is susceptible to manipulation, can we somehow reclaim truth on the internet before it’s too late?
Worse Than a Crime, a Photo 📸
Understanding the modern era requires a depth-first search into the past. And so the birth of deepfakes, as we understand it, interestingly starts with photo-editing.
Instagram began as a tragedy and ended as a farce. The app was birthed as a place to share glossy photos – users could harness filters to simulate the best parts of themselves – essentially, easy Photoshop. It was clear that nobody desired genuine images of themselves and slowly, every photo we see has become artifical in order to compete with the other pictures. But, this innovation in photo manipulation has been centuries in the making.
The first instance of retouching photographs is said to have happened in 1846: back then, specialists would carve their film with knives, draw on top, and even splice multiple images together to create a single print. And like many issues today, the ethics of such tampering were hotly debated. Because of the intense manual labor and capital involved (retouching desks were mandatory), it was less common. But just as it has become easier to take photos, it has become easier to manipulate them.
Every technological innovation begins on the fringes of society and first used by hackers who apply the inventions in unorthodox ways. This famous photo of Lincoln in the 1860s, taken a couple of decades after Niepce created the first photograph in 1814, was fake; the photographer affixed Lincoln’s head upon John Calhoun’s body. That didn’t change the response – people saw what they wanted to see – Lincoln as the archetypical hero as well as the enigmatic, saviour of the union.
Of course, it didn’t take long for photo editing to become the de facto tool of cultural satirists. If you could edit the past, why not do so in a manner that supported your political beliefs?
John Heartfield, a pseudonym for the German graphic designer Helmut Herzfeld, heralded the age of political photomontage, or political photo manipulation. Despite growing up in Germany, he was a strong dissenter of the Third Reich and fought back by masterfully combining political satire in photographs. Viewers outside of Germany viscerally connected with his images because it, again, confirmed their prior beliefs: Hitler as a greedy, cunning dictator who’d stop at nothing to get his way.
Overdosing on Lies 💊
But, photomontages didn’t stop there; undoubtedly, it was bound to enter the beauty industry. Starting in the 1950s, Playboy crafted and perfected the notion that women must ‘look’ a certain way – these techniques were heavily relient on the editing skills, and later software, of their time. They culturally commodified the process; it became acceptable to airbrush models and adversely inspire generations of women to reach for an impossible ideal.
However, lurking under the surface was the natural human incliniation to play up societal stereotypes. Time’s infamous photo of OJ Simpson was rightfully pilloried by other members of the press after it was found that they had artifically darkened the Newsweek photo to play to their readers’ fears of African-American violence (coincidentally, a few years after the Rodney King riots).
Deepfakes in this timeline are the final evolution of photomontages. We’ll use algorithms to generate our own symbolic ideas of reality and thereby collectively imprint our own bias on the events of the time.
Do we take photos for record-keeping? I don’t think so – I suspect that between the choice of the truth (the event that actually took place) and an artificially simulated photo, we’d choose the latter. Instagram isn’t real life and nor do these simulated photos display an accurate record of the past. The media has been whitewashing reality for centuries; it’s only recently that our society has understood the extent of this charade.
Seeing is believing 👀
The internet is suffocating under the weight of its own lies. Why? Because it’s subjective truth. Wikipedia, news reports, and publications can all be edited ex post, changing the nature of reality as we remember them. What will our ancestors remark about the Epstein case – will they believe any of the reporting that occurred?
Primary sources in history are given extraordinary weight only because we believe the authors are recounting true, unedited events of the time. We trust their lived experience. Yet, that’s not the case anymore – documents are doctored in the guise of “easy collaboration” and “listening to the audience.”
This shambolic turn of events isn’t nascent; however, the velocity of communication has changed the game. Almost immediately, events will be constructed to fit overarching narratives, backed up with proof engineered by deepfakes. George Berkeley once remarked, “perception is reality.” As we saw with political photomontages, they don’t need to be truthful, these bits and pieces of media need only be persuasive.
The problem with listening to the mob is that they burn the books and people who fought them, brand those they hate with the scarlet letter, and cast off opinions that besiege their ideals. Winners in the public realm have always written history, but never have they written and edited it as fast as they have now.
Regardless of how moral we may consider ourselves to be, our base instincts of combating those with different opinions has not changed; in fact, cancel culture has proved the opposite – those with heretical views are burned at the stake writ large.
Is there Objective Truth?
Make no mistake, the splintering of the internet has already begun and all the technology giants are coming to terms with it in their own way. Their platforms are drowning under the weight of bots, fake videos, and deepstate actors propelling their own narratives to influence elections.
Let’s take a step back: If we think of text as the modality of subjective truth, then what’s the modality of objective truth? To me and many others, it appears to be math, crowdsourced at scale. Incentivizes bend reality, which is why the media apparatus ingests truth and spits out lies. What if there was a machine that could do the opposite: feed it lies and have it generate truth?
It is in this framework that Bitcoin might go down as the biggest technological invention since the printing press – in an inverse manner. Gutenberg wanted to decentralize the power of the Catholic Church and successfully did so. Now, upon witnessing the havoc caused by too many simultaneous opinions with no cohesive narrative, we are trying to go back to a world where truth isn’t an endangered animal.
Truth, like money, is a scarce commodity. And like money, we collectively agree on what it is and what it isn’t – we can’t wish the truth away much as we can’t wish away debt or credit (though we’re trying). Thus, the pushback is coming; people want the truth no matter the cost and the only way they can get it is through the blockchain.
So, if I had to make any predictions about the future it would be that all online content will eventually be fake because of deepfakes. The technology cabal isn’t strong enough to monitor the content on their own platforms, much less that of the entire internet. Perhaps the only medium to fix this is a truth-first internet, underpinned by the blockchain, where people will discern what is real through proof of work.
Secondly, the public internet will die and be split up into discrete parts of exclusive information silos where people can police the quality of the information they receive. Look at the new cambrian explosion of social networks: all of them are highly-curated, selective, and intimate. The experiment of a universal internet has failed and has wrought its own demise.
If we choose to stay in the past, the future will hit us like a sledgehammer. The powers that be are too distracted by their surface-level squabbles with China to dig deeper; to comprehend these monumental tectonic shifts requires a mindset focused on the truth.
And sometimes like Aldous Huxley said, “ending is better than mending.” We can’t allow structures to become our shackles – we must release our ideological shackles and start anew.
Thanks to Gabe Leydon for his feedback on this post – when he talks about the future of technology, I listen.