An Aigitated warning; a pre-posting preamble
Post 2.3 and a related side series of amplifying and deep-dive pieces will (finally) be posted soon, but first...
In a previous notice, I stated there were signs that instances seen that conflate the word photography with AI-generative “art” appeared to be on the wane. I have to renege on that assessment. I continue to find instances where I need to point out that AI-generative “art” is not a photographic process and does not produce photographs.
Yes, the battle continues, especially on the Medium platform, and may be getting worse. Many content producers continue to blur the distinctions now by deliberately incorporating photographic terminology in their “prompt engineering.” Some show complete and deliberate disregard for the differences by stating outright that they are producing photographs.
Let me be very clear: this is not a situation where they are blending photographs with AI-generative tools, which is a very gray area. They are calling their generative images created strictly by prompts, photographs. Many know better because they say they are photographers or teach “computational photography” (another misnomer in this context). This is fraudulent, and I have posted comments saying so in no uncertain terms. This is how many respond to my pointing out basic truths to them:
I just come in on a different device without commenting and continue monitoring and cataloging what they say and do. I could name names. I could provide links. I chose not to because “views” and viewing duration are how they make the pitiful payouts on that platform. Some posts are very long. Some are long because much of the text is often obviously AI-generated, given away by the repetition and other peculiarities we have all come to recognize.
I take these assaults on photography seriously, and I am prepared to back what I say.
In the previous core post, I pointed out anachronistic historical fallacies and the reasons they happen. That post was never a part of my original outline of this project, but once recognized as a serious impediment, it became important to address. And that will continue, in more detail, in future posts.
The problems caused by overextended, romanticized “dogsled” narratives stem in part from the need in this new “information economy” to produce a constant stream of new content on a deadline or schedule to maintain their “followers.” They rely on content fodder from quick-hit “research.” The returns found and used, usually from the top of the first page, are alleged historical facts that are usually little more than massively abridged, sloppily paraphrased, and endlessly recycled historical corruptions that are contextually barren, overly homogenized, dumbed-down dreck. But since they rank high in manipulated search returns, they are speciously used as a copiously cited corpus of deeply flawed “sources” used as if they were legitimate references.
These problems are being made much worse as the race to embed new GIGO AI algorithms, programmed by humans with agendas—don’t ever forget that—into search tech that produces returns deliberately and increasingly, algorithmically and humanly pre-biased and “vetted” for our emotional safety based on their sense of what is “correct.” The number of returns shown is deliberately reduced, and the content is AI-summarized and truncated, which is then re-crawled, indexed (or scraped), and repeatedly regurgitated. And will become circularized training material for all these AI producers.
These results are more pernicious than the “sponsored” search returns, which are labeled as such and can be ignored if desired. “Search,” as we have known it, was always an AI process sharing certain similarities with present predictive models, albeit more benign, but woe to us all once these new technologies and the oligarchy of overlords have taken complete control of the process and the results. I will continue to show examples as I get back on track examining the relevant technical history in developing a needed legal definition of photography.
I need to reiterate that not all information, data, references, or “content” is available on the open web. Huge amounts (often the needed primary sources) are in gated-access repositories or behind paywalls. Much is in those old-fashioned brick-and-mortar buildings called libraries or, in private hands, undigitized.
Thus, “the web” is a vast but limited resource. It is simultaneously a vast and expanding cesspool. Serious issues with new AI technologies appear almost daily. All web-based and sourced information needs to be vetted, all the time. AI/AI-enhanced technology incursion, immersion, and recursion will only make this more obvious and the need more necessary, while making that process more difficult to accomplish.
We now need to “vet the tech” as well as the content.