#i'm getting very good at spotting AI images in the wild
Explore tagged Tumblr posts
Text
I don't think that making or using AI art/image generation is morally wrong, as you guys know, but I have to admit that slotting a wibbly-lined low effort image selected from the first result set from Midjourney in your content is incredibly tacky. At least select something that looks good. Maybe something without the boring AI "sheen" look either.
You see this a lot in clickbaity content like web spam and youtube shorts attempting to algorithm game.
#i'm getting very good at spotting AI images in the wild#or at least i think i am lol survivorship bias#and yeah it's usually a shit look that lessens my interest#now that the novelty has worn off#assuming that nothing interesting is being done with it#ai
1K notes
·
View notes
Note
Please could you talk about the weird and specific visual language of person of interest? I haven't noticed it before but now I'm intrigued
Hi, Anon. You sent this ask 8 months ago. I took this long to answer because I wanted to really lay it all out for you, with the right screencaps and reference images and a whole theory of why the show looks the precise way it does. But the fact is, that’s not going to happen. I don’t have the time or mental energy to do it properly, and I have finally accepted that because the almost completely written long-ass answer to a different question about a different show that’s been sitting in my drafts for about as long is still languishing! and it’s almost done! wild what grad school can do to a person.
So since I have accepted that I can’t do the Full Treatment I wanted to, I’ll just do the quick explanatory version.
POI has 5 main visual ideas it returns to over and over. These are:
1. Surveillance footage. This is notable because it tends to look, by normal framing and cinematography standards, bad; they are very deliberately putting out “bad” images to tell the story, which after all is in large part about surveillance and privacy. I can’t remember now where I read/heard this, but the EPs definitely talked about the struggle to get cameras into these really inconvenient spots to get the properly terrible angles that would play as plausible CCTV-type footage.
2a. Edward Hopper’s style. By this I mean less his use of color than his tendency to show one or a few static figures, small in relation to their (usually urban) surroundings, often locked off by a frame-within-the-frame (someone framed in a window within the larger setting, for example); and significant contrast in lighting. The emblematic example for me is the wide shot of John cradling Carter’s body next to the phone booth, right after the moment of her death. But a lot of the comedic exterior shots of a bar or other establishment inside of which John is wreaking havoc actually fit as well.
2b. Noir, which was probably influenced by Hopper in the first place (or at least was responding to similar aspects of the zeitgeist). “POI noir” is a very popular thing in the fandom, and I assume you don’t need me to explain it. I will say, though, that as a film genre noir is about the failure of institutions, loneliness, isolation, and doom, so the choice is both telling and appropriate. (The same applies to the Hopper point.)
3. This is related to 2 and 3, but there is a broader theme of small isolated figures against a big, empty (or sometimes empty by virtue of being anonymously crowded) urban backdrop. (This is something noir does a lot also.) The reason I’m separating it here is that POI does this a lot of the time without making it noir; it creates something lighter-colored, lower-contrast, and more contemporary-looking, but maintins a similar effect. Takeshi Miyasaka’s work evokes it very directly for me. What’s interesting about his paintings as well as these compositions in POI is that they often come off as unstudied, not overthought or excrutiatingly composed. This is not the case; they are very carefully thought out (in both cases). But they convey a sense of naturalism you don’t get from a traditional noir composition (or most of Hopper’s noirish work--some of his other stuff is different, but not relevant here), while maintaining that feeling of a small subject in a big, uncaring, not particularly beautiful world.
4. Comics. POI is often compared to Batman; in some ways it more closely resembles Batman’s even noir-ier antecedent, The Shadow. It obviously leans into this in 4x06, “Pretenders,” but it’s there in a lot of the action sequences. I’d cite the one where John goes after Quinn in “The Devil’s Share,” a lot of “Relevance,” or this as examples.
5. The “interior,” thinking/processing shots of the Machine and Samaritan. For these they obviously invented a lot. They definitely drew on what was at the time cutting-edge data visualization as well as the marketing materials of some leading tech and security companies to create them, but I think it’s one of the ways in which the show was most original. (Interestingly, you can see some of this idea being worked out before the show even began in the movie Eagle Eye, which had the same producers. It’s not a good movie, mind you, and its notion of AI is super simplistic, but some of the visualization in this area is clearly prototypical of what would play out on POI.)
A lot of these visual choices don’t stand out as noticeable (aside from the AI visualizations, which of course are unmissable). There are two main reasons for this. One is that none of them are constant. POI never set out to make every shot a stylistic masterpiece, the way a show like Hannibal more or less did. I assume this was partly because of the logistical realities of a 22-episode season, but it also works with the show’s storytelling. The idea they want you to get, that they reiterate over and over, is that the world looks normal but isn’t underneath. So the 5 stylistic ideas I mentioned above tend to appear in short spurts--an action sequence, a shot framing Harold against a New York skyline or the whole team by a bridge, etc--stitched together by pretty standard TV framing (shot/reverse shot, close-up, medium). This allows them to kind of ramp up and ramp down the level of visual intensity, using noirish or comicky compositions, or particularly intense AI visualizations, in line with the storytelling. It is also a kind of metaphor for the entire premise: things look regular most of the time, but they aren’t if you pay attention.
The second reason is that we are so incredibly used to surveillance footage! It doesn’t stick out to us anymore! This is actually very significant because, as I noted above, surveillance cameras are not positioned for aesthetic value. The point is coverage, not composition or image quality. So this is a cinematic product (a television show) that is working overtime to give you technically “bad” images. But those images don’t really stand out as ugly, per se; instead they denote truthfulness. So for example, the Machine’s POV (as opposed to its thought process, which is digitally animated) is almost always in this surveillance style. This reminds us of what the Machine is and how it works, and it also tells us that what we’re seeing isn’t subjective. It is not someone’s memory or their perception. It is a literal recording. This allows the Machine to act as our guide through the whole timeline, moving us back and forth through that horizontal scroll, zooming in on a moment or incident for replay. We never have to question if the Machine is lying to us or mistaken (aside from 5x02), because its memory is a video archive presented to us with all the hallmarks of video that means “proof,” not video that means “feelings” or “perceptions.” Once we’ve been transitioned into a flashback scene, that style can go away so that we can abandon the distance that surveillance introduces to engage with the emotion of the flashback scene, but they use that device to move us around because it automatically tells us “this is true.” Simultaneously, the ubiquity of surveillance-style shots reinforces that same message I talked about above: that we are being watched (because the world has gone sideways).
(Perhaps I should mention that Nolan and Plageman said a big goal of theirs with the show was to make people more aware of how profoundly the technological and therefore social world around them was changing with little notice or fanfare. One of the obvious ways they were trying to direct our attention on this point was to issues of privacy. Finch helping John out of a sticky situation by exploiting a homeowner’s smart TV isn’t just a matter of making Finch look clever. They wanted to let us know that if your smart TV has a camera, someone can remotely turn it on. [I know this is common knowledge now, but that episode was like 2012. I found out what Palantir was because of POI!] Just to contextualize why to me, “you are being watched” and “the world has gotten very weird” are basically the same message in the context of POI.)
One of the show’s key influences is The Naked City (the show and the film). The film version of The Naked City is notable because it was one of the first post-WWII movies to shoot on location instead of on a film set, and it was shot in NYC. It’s also one of the earliest police procedurals, in the form we know the genre today. It’s not exactly noir, but it’s not exactly not noir. It’s based on the photography of the famous ambulance-chasing, poverty-documenting photographer Weegee; you could say it’s an early example of a “gritty” film. It doesn’t have the elegant devastation of The Third Man or the deep shadows of Double Indemnity. It was all about bringing audiences the amazing spectacle of a real place; the paradoxical insistence on authenticity in cinema goes way back. The Naked City to me is somehwere between noir and cinema verité, and in that sense, POI is true to it as an influence.
26 notes
·
View notes