Are Google Phones Fascist?
Yeah, yeah. Clickbait headline. Except I am not sure that it’s not actually accurate. This is going to get a bit ramble-y so you might want to grab some snacks and a drink before you dive in.
Sarah Jeong at the Verge has an excellent article on the Pixel 9 line of phones and the incredible easy editing tools they added to them. In Jeong’s words:
A cockroach in a box of takeout. It took less than 10 seconds to create each of these images with the Reimagine tool in the Pixel 9’s Magic Editor. They are crisp. They are in full color. They are high-fidelity. There is no suspicious background blur, no tell-tale sixth finger. These photographs are extraordinarily convincing, and they are all extremely fucking fake.
The Pixel 9 lets anyone create their own reality, at least in pictures. Yes, Photoshop has been a thing forever, but it took more time, money, and skill to mangle pictures than is required to do the same damage with these phones. Jeong does a great of laying out the fact that photos are entering a space where they have to be assumed to be false or fake.
That is fine with Google, by the way. They designed the phones this way. From the same article:
the group product manager for the Pixel camera described the editing tool as “help[ing] you create the moment that is the way you remember it, that’s authentic to your memory and to the greater context, but maybe isn’t authentic to a particular millisecond.”
That is horrifying. Pictures used to be a record of what happened, or, at least, what people photographed while it was happening. Now Google, who used to be dedicated to organizing and access all the world’s information, is creating and encouraging the use of tools that flatter your memory instead of telling you what really happened. No wonder they don’t seem too concerned with the flood of imitative AI slop poisoning their search results.
Bullshit is not a new problem. Fox news regularly misinforms its audience. If you watch Fox news, you are the least informed news consumer, and it recently settled a defamation suit around its personalities telling flat lies about voting machines in 2020. Fox New’s business model is based on telling its audience what they want to hear, not telling them what is actually going on. And that, I think, is the big problem with the Pixel 9 phones.
Jeong rightfully points out that the next atrocity, like police brutality or Abu Gharib, will be buried under AI generated fakes, and thus much less likely to generate the outrage required for real change. But I think that is the lessor problem. I think this new flood of unreliable photography is going to only help authoritarians. Fascists don’t care about being caught in lies or telling a version of the truth. They want to make it so that their followers cannot know the truth. In the words of Hannah Ardent:
It has frequently been noticed that the surest long-term result of brainwashing is a peculiar kind of cynicism — an absolute refusal to believe in the truth of anything, no matter how well this truth may be established. In other words, the result of a consistent and total substitution of lies for factual truth is not that the lies will now be accepted as truth, and the truth be defamed as lies, but that the sense by which we take our bearings in the real world — and the category of truth vs. falsehood is among the mental means to this end — is being destroyed
They are already partially there. They have, as noted, a news source that won’t tell them inconvenient truths. And, as the fact checking scandal of the DNC proves, they are well on their way to making the truth a partisan issue. Many, many fact checks of the DNC were utter nonsense, with this perhaps being my favorite (its PolitiFact and I am not linking to it directly):
Harris: Trump “plans to create a national anti-abortion coordinator and force states to report on women’s miscarriages and abortions.”
Mostly False.
What Harris describes is Project 2025. Although the 900-page policy manual makes such recommendations, it isn’t Trump’s plan. The project, led by conservative Heritage Foundation, contains proposals for the next Republican administration, and got input from dozens of Trump allies. But Trump and his campaign have repeatedly said they were not involved in the project and Trump is not listed as an author, editor or contributor.
This requires, as Scott Lemieux points out, and willful disregard for the meaning of, well, words:
“It’s not Trump’s plan, it’s the plan of policy blueprint put together by the Trump allies who will be responsible for staffing the executive and judicial branches and setting their agenda if they win. Bur Trump backed away from it when it became politically toxic so Democrats cannot mention it.”
This was not just one organization. Most fact checkers went out of their way to ignore the meaning of words, of how the English language was used, of how people speak. It was so bad that the Washington Post even mocked its own fact checker. These, also, are not isolated incidents. The New York times, for example, while covering Biden’s age issues extensively, almost obsessively, largely ignored the cognitive decline of Trump. In all these cases, the mainstream news organizations appeared to go out of their way to not tell the truth. By insisting that telling the truth, or providing proper context, or pointing out when someone lies, is partisan, the right-wing has done a good job of reinforcing the already poorly handled “he said, she said” tendencies of journalism to muddy the very idea that one thing could be considered true over another.
By doing so, they create an information space that reinforces the notion that their followers can only trust them. After all, if there is debate about basic facts, if no one can trust the news organizations to state the truth plainly, then you are better off getting your information from someone you trust — like the leader of your political movement. That plays into the hands of fascists and other authoritarians and weakens democracy. If we cannot agree on the shape of the world, we cannot have democratic conversations. And I don’t think it is going to get better.
Cole Haddon (who you should be reading if you are interested even a tiny bit in art during the late stages of capitalism) recently highlighted Warner Herzog’s concept of the “ecstatic truth” approvingly. I am not so enamored of the concept, in part because of the idea that facts are not the basis for truth. As I said at the time:
Of course, not all facts are created equal. We are learning new things about history all the time, and we will be amazed tomorrow at what we knew to be true today. And movies are entertainment with limited scope for telling their stories. Not every fact is going to fit, and every fact left on the floor is a choice made to shape the story told. I do not deny these truths, nor do I think movies are lessor because of the need to bow to these realities. But I am concerned that completely ignoring facts in historical movies gives a weight to the “truth” that the artist wants to convey that it does not deserve.
That is what I believe authoritarian and companies like Google are doing: they are privileging their concept of the “truth” over the reality of the facts. They are allowing anyone to create a mythical world that fits their preconceptions, facts be damned. And they are doing so deliberately.
OpenAI could, if they wanted, make it clear which ChatGPT outputs were generated by them. They choose not to because they are afraid a significant portion of their users will stop using their products. In other words, they prioritize bullshit over people. If OpenAI could mark their outputs, so could Google. But as far as I can tell, no such protections exist for the Pixel 9 manipulations. After all, who is Google to question the ecstatic truth of its users?
Maybe I am overreacting. Maybe we will adjust, in the same way we adjusted to the possibility of Photoshop manipulating phones. But as Jeong points out, we now move from a world in which photos were assumed good until proven otherwise to a world in which photos are assumed to be manipulated unless proven otherwise. And given the incentives that fascists, news organizations, and tech companies have to privilege fake content, I am afraid that I am not concerned enough.
There are things we can do as a society — make companies liable for the output of the AI manipulators, require them to mark their outputs, provide grants for local news, break up monopolies so that a “truth phone” has a chance to be introduced to the market, etc. But since most of the money and power line up against those actions, it is going to be a long fight.
A fight long enough, in my worst fears, that the new status quo is the ecstatic truth of the fascists permanently polluting our world.
K.C.,
from Cole Haddon's 5AM StoryTalk that you pointed us to: "And there you go: Herzog’s ecstatic truth as a kind of performance art predicated on a lie to express a deeper truth."
How does this sentiment compare with the title of your substack, Metaphors Are Lies ? They seem awfully close to me, and I speak as one who is quite friendly to metaphor.
I'm skeptical of language like "a deeper truth", though, because it implies truth is a thing with dimensions (huh! a metaphor!), an inside and an outside, depths, and so on. I don't find that a helpful way to think about truth. I prefer to think of it as something that brings greater illumination and more insight, something that enlarges our understanding. That is why I’m friendly to metaphor, because that illumination and enlargement is what I think good metaphors do.
Which makes for a nice segue, I think, to thinking about what we call facts. An isolated fact is a pretty meager thing, even a disputable thing. The disputability, if there is such a word, of facts comes from most of them being a product of interpretation. This is even more obviously true of what we would call a ‘scientific fact’ because those depend on a scientific framework for their meaning. We can use them readily because of how widely many conceptual frameworks are shared, although – as we’re all too aware these days – not all the frameworks are universally accepted. But scientific facts are like other facts in that what makes them substantial and meaningful is their connection to other facts in a contextual framework.
When it comes to photographs as facts, the practice of altering photographs started almost with photography’s earliest uses. The current AI manipulations are faster and ‘truer to life’ than the earlier and cruder alterations, but that’s a difference of degree, not kind. The sheer representational facticity of photographs has never been beyond question even if most are accepted at face value. [btw, the Hannah Arendt excerpt on cynicism was a good find.]
I do agree with you that Google’s new offering is not a good thing, but I would frame it a little differently with respect to different practices of photography. I think the speed and potential verisimilitude of its artificial images will disrupt the remaining credibility of the snapshot. Who knows, maybe there will be an unintended good consequence of undermining the credibility of the ‘Mona Lisa selfie’ that will lead to its social media devaluation and less crowded museums? But the devaluation of the snapshot will bleed over into the practice of photojournalism, which shares with snapshots the aura of capturing a moment.
But in any of these cases, the provenance of candid photographs is what matters: photographs are never incontrovertible proof by themselves; they always require human witness and often require information outside the frame that the witness, er, bears witness to. That brings us back to issues of human credibility, which are very old indeed. With the current rage for activism in journalism, credibility readily becomes suspect (but that’s already happened, and Google didn’t do it, for all that G may help it along).