AI in Your Hands

Alphabet Inc., Google Pixel 8: AI in Your Hands, November 2023,  commercial video (sound, color), 60 seconds. (Source: YouTube)

The more enhancing the mediation […] the more direct, vivid, and real the experience seems, which is to say the more direct, vivid, and real the fantasy and dependence are.

Wallace, 1998, 75

[W]hen all experience can be deconstructed and reconfigured […] each quantum is as good as the next, and the only standard of a particular construct’s quality is its weirdness, incongruity, its ability to stand out from a crowd of other image-constructs and wow some audience.

Wallace, 1998, 79

I saw Google’s Pixel 8 ad (pixel-ate, geddit?) in the cinema a couple of weeks ago. Even though the movie after it featured probably $50m+ of staggering yet invisible (and exploitative) CGI, the phone ad seemed, in retrospect, (and echoing Wallace) more exciting, weirder, and more, just, wow.

Why?

According to the ad, Google’s latest “phone” promises that “the AI in your hands” will “make it pop”, “make it clear”, “make it sharp”, “make everyone 😃“, “and change everything”. 

In other words (and in the words of the comments below the video) it will render every image taken enhanced, exaggerated, perfected, edited, manipulated, staged, collaged, redacted, in short, spatiotemporally reconfigured, and hence fake, false, deceitful, manipulative, creepy, terrifying, Orwellian, dystopian, uncomfortable, untrustworthy, fabricated, unreal and, basically, bad, but also and at the same time cool, mind-blowing, brighter, awesome, amazing, sick, jaw-dropping, magical, wild, and just wow ❤️. 

There’s quite a lot going on here, much of it not exactly new. (Including the fact that there is actually zero AI inside the physical hardware, the hyped yet widely available features being software services running in massive compute factories). Some tropes and provenances….

Idolatry: images are appearances not reality, hence bad [Plato]

Addictive fantasy: technologically mediated images are enhanced to seem better (sharper, clearer, poppier, happier) than our lived experience, hence are fantastical and addictive, thus making our lived experience actually worse (paler, lonelier, more frustrating) [Wallace]

Proliferation: the exponentially increasing volume / torrent / barrage / assault of fantastical images desensitizes us to / alienates us from / overwhelms us in our lived experience, hence making all images equally suspect and of equally low value [Debord]

Populism: it was OK when a few skilled people with access to specialist tools–a technologically literate elite–could manipulate images, but it’s going to be unfathomably dangerous when everyone with access to a smartphone (82% of the world’s population, over 6 billion people) have the same power [tabloid journalists, unemployed designers]

Vulgarity: the nature of the tools privileges a lowbrow aesthetics of shock and awe, inane grins, lame humor, dangerous contests, daft dances, disgusting recipes, homogenous styles and fast-paced viral fads [parents]

Abuse: the same tools enable malicious strains of vicious treatment–worst takes not best takes–revenge porn, online bullying, blackmail, slander, defamation, abuse, hate speech, hate crime, disinformation, cyberterrorism, etc…. [tabloid journalists, pundits, politicians]

Unattainable ideals: of beauty, lifestyle, creativity, success, luxury, happiness…. [everyone]


Some assumptions and equivalences here are worth questioning.

What’s wrong with a still image that artfully collages still images representing different moments in time? Why must a still image freeze a single moment rather than represent a durational event? What’s so real or important about a few milliseconds when cosmic time stretches for billions of years and general relativity entails that there’s no such thing as an absolute moment anyway?

Why should editing, enhancing, or otherwise manipulating an image make it fake in any derogatory sense? Haven’t artists always used such tactics in pursuit of goals of beauty, truth, or political mobilization?

Why aren’t ‘synthetic’ images real? If physical reality is, say, quantum fields, physical reality is available via abstraction only and not experienceable by humans at all, and all our percepts are unreal in that sense, why should chairs and tables have any ontological privilege over visual representations or, more pointedly, simulated constructs of those objects? The painting of a pipe may not be a pipe but the pipe is not ‘really’ a pipe anyway, it is only expediently perceived and used as such by human animals. An AI generated image of a pipe may involve a new and different supply chain of exploitation, but for what reason is it any less real?

Is the grip of social media not substantively declawed once everyone can easily perfect to their own personal or tribal ideal every image they upload?

More generally, to the extent that we live digital lives already, i.e. our lived experience is in fact highly and increasingly technologically mediated, aren’t we now, contra Wallace, by “enhancing our mediation” actually enhancing our real lived experience, not some disconnected, alienated simulacrum of it?


Who owns the means of production here? The manufacturer of the devices, the consumers who buy the devices, the telcos who provide the credit to the consumers, or the shareholders of those corporations? Who benefits? Who is exploited? Who else is harmed? What exactly is ‘production’ or a ‘means of production’ in this context? The consumers aren’t really producing anything here: their labours are unwaged. When there is AI silicon in the devices–and this will happen very soon–an interesting economic event does cross the horizon, namely the outsourcing by Apple and Alphabet of capital expenditure on machine learning hardware.

But what quanta of rare minerals need to be extracted and human labor exploited for a billion or two people to upgrade to the next, spiralling technological cycle of AI-enabled smartphones?

What dominant forms of power are being embedded and enforced when computational photography and technological mediation ‘enhance’ an image? Are men and women de-androgenized? Are dark skins rendered paler? Are tent cities deleted from the sidewalks? Are autistic people given more asymmetrical faces and curlier hair?! 

Why is this emoji 😃”Grinning Face with Big Eyes” the facial expression promulgated as the ideal in this ad? The emoji in question is “a yellow face with smiling eyes and a broad, open smile, showing upper teeth and tongue on some platforms. Often conveys general happiness and good-natured amusement. Similar to 😀 “Grinning Face” but with taller, more excited eyes. (Emojipedia, 2023)

Why is it so couter-normative to shut one’s eyes or fail to smile in a group photo? Is there here some balefully hopeful projection of joyful communitarian being in a world of atomistic individualism? Why do pain, suffering, exasperation, boredom, serenity, fear, shock, etc. need to be redacted in group shots of families, friends, co-workers? Wouldn’t the richness, complexity, not to say turmoil, of our ‘inner’ lives be more visible in outer expression to others if we were not constantly censoring and self-censoring in the ways that this ad urges us to, and hence experience a less lonely and more genuinely communitarian style of life?


Closing note: it’s fabulous for mega-tech companies and the surveillance states they service that we’re all wringing our hands about harmless deep fakes while AI systems are actually being covertly and unjustly deployed everywhere all the time to minimize welfare payments, increase incarceration, optimize deportation, and escalate genocide. That’s some serious sleight-of-hand we’re falling for….

Bibliography

David Foster Wallace, E Unibus Pluram: Television and U.S. fiction, 1990 (essay, collected in A Supposedly Fun Thing I’ll Never Do Again, Abacus, 1998)

Leave a comment

Your email address will not be published. Required fields are marked *