(C) Daily Kos
This story was originally published by Daily Kos and is unaltered.
. . . . . . . . . .
Branding, AI, And The Death of Language [1]
['This Content Is Not Subject To Review Daily Kos Staff Prior To Publication.']
Date: 2025-06-04
A short one today, and probably not where you think this is heading. To cut to the chase, I am not actually blaming the AI dweebs for this specific issue. Rather, today, our enemies are the true villains of humanity — the marketing dweebs.
There has been a kerfuffle in the online publish world around Kobo, the leading contender, as far as I can tell, to Amazon in the e-book space. They recently updated their terms of service to allow them to use what they call AI for certain purposes. You can read the terms here. They are clear about not feeding books into imitative AI, what most people call generative AI, the technology behind ChatGPT, for example. But they are also vague on whether or not the models they use for their purposes are or are not based upon or built off of the kind of Large Language Models that are at the core of imitative AI. Authors are concerned — these models have been trained in large part by stealing the work of authors and using them for model training and then turning around and profiting off that work. So far, so usual. But this time, I think the dispute is also one of language — specifically, how we have allowed language to lose precision in the pursuit of all powerful branding.
Kobo itself has a thoughtful and reasonable take on the issue. They state that they use algorithms, machine learning, and other tools that get lumped into the AI label to do things like find books that have child sex abuse material in them and corral them for human vetting. This cuts down, they say, on the amount of psychologically damaging material that humans have to moderate. They also intend to clean up metadata to make books easier to find, and to try and counter the tidal wave of AI slop that they receive. All of the above could be done with systems that do not use LLMs but are called AI regardless. However, the Kobo terms and the Kobo employee on Bluesky NEVER say that they won’t ever use LLM based tools and the terms of service, by sticking with the loose term AI, leave that possibility open.
And that is where the warping of language comes in. They could use more precise terms — machine learning, non-LLM based algorithms, etc. That is not a panacea, as even those terms can be considered generic. But they world thinks of algorithms, now, as AI. This is in part because the people pushing technology have decided that AI is sexier, in the same way they decided a few years ago that machine learning sounded sexier than statistical analysis. By using AI as a term, Kobo is both covering a wide blanket to ensure that what they do is supported under the terms of service AND creating ambiguity about what, specifically, they are actually using.
The use of the AI branding, then, sends different messages to different people. To lawyers, it is a potential cover for any algorithms that might be useful down the line. To CEOs, it is a magic word that means fewer employee costs and more money for them. For many other people it is a burden to be dealt with at work or the theft of their own hard-earned creativity. Language has always been malleable, but it is interesting to see how the marketing world manipulates connotation to serve their ends. Obfuscation and imprecision serve their needs, so we get vague terms, and everything connected to an algorithm gets labeled AI, often to the detriment of clarity.
Which is probably the point. I have discussed the problems with branding before, but this little adventure in language manipulation reinforces those concerns. Branding is not a morally neutral exercise. Branding almost requires that you use the languages in ways designed to obscure rather than illuminate. Even if it is just trying to focus on one aspect of your work/company/personality, it is still meant to obscure. I was once told that my fiction writing’s brand should be “sarcastic little stories about fighting capitalism and monsters”, and that is somewhat fair. But it is also not complete. And if I tried to push that line about my writing, then I would either warp what I write to fit it or warp the definitions in that sentence to fit what I write. Manipulation and obfuscation are unavoidable in branding.
No shattering insight here, I know. But there is value in pointing out the effects of the bad things we are encouraged to do in our lives. And while this is a minor dust up— some confusion and hard feelings appear to be the worst outcome — it is still a good example of how we, as a culture, have abandoned precision for, well, marketing.
[END]
---
[1] Url:
https://dailykos.com/stories/2025/6/4/2325777/-Branding-AI-And-The-Death-of-Language?pm_campaign=front_page&pm_source=more_community&pm_medium=web
Published and (C) by Daily Kos
Content appears here under this condition or license: Site content may be used for any purpose without permission unless otherwise specified.
via Magical.Fish Gopher News Feeds:
gopher://magical.fish/1/feeds/news/dailykos/
