Category: Other Page 49 of 177

This is a reaction to a TechCrunch article about the EU Commission, and the intention to apparently require companies to label AI-generated content.

There’s a lot to say about this, but much of it is long-winded and boring, and my drilling down into the details as a “practitioner” is likely to convince no one important of any course corrections.

Instead, I’ll just pick out a few key things to quibble with.

Transparency Commissioner Vera Jourova is quoted as saying:

But when it comes to the AI production, I don’t see any right for the machines to have freedom of speech.

I can’t tell if this is just naivete or misdirection, but it’s important to highlight here that the freedom of speech of AIs is not in question, as we all recognize under the law that humans have human rights, not machines.

The potentially affected right to free expression at issue, then, is not that of machines, but of the human operators of those machines.

Also quoted in the article is the idea that signatories to the EU Disinformation Code which create generative AI tools…

…should commit to building in “necessary safeguards that these services cannot be used by malicious actors to generate disinformation”.

I don’t think anyone on earth actually knows what this even means. Prevent AIs from saying things that aren’t true, and of making pictures of things that don’t exist outside the imaginal realm? Without more clarity, this is imo just a lot of mouth-flapping and buck-passing as politicians wiggle their thumbs and hope that someone else will come along and fix the ambiguity, and they won’t be saddled with actual responsibility for drafting codes of conduct and potentially legislation with bad imprecise language.

The second part of the above is that platforms which might distribute such gen-AI content should…

put in place “technology to recognise such content and clearly label this to users”.

Assuming for a second this technology to recognize AI content actually exists & is fully reliable (it does not, and is all unreliable as hell. I’ve tested every one I can find, and they’re ALL wrong), nobody in these conversations ever says what they thing this kind of labeling will actually achieve.

They think people will be scrolling their social feeds on the toilet, and see a label “AI generated” and that _____ will happen? Nobody ever fills in the ____. It’s always an unstated fantasy that people will see the label, and this will lead them down some sort of path of critical thinking, which ends in them rejecting the thing as “false” and of no consequence.

For a body ostensibly dedicated to studying and preventing the spread of malicious disinformation, this represents a stunning lack of awareness of how disinformation – and also just normal information – even works in the first place. It clings to an Enlightenment Era ideal that is never said aloud: if people just know what the “truth” is, they will rationally respond and act responsibly, and not continue to elect fucking morons, tank civilization, and destroy the planet, etc. We only need to look around to see how well that approach is working out.

One more, related:

…the EU wants labels for deepfakes and other AI generated content to be clear and fast — so normal users will immediately be able to understand that a piece of content they’re being presented with has been created by a machine, not a person.

This is silly because, it’s almost never an AI that is randomly creating and distributing AI generated content all by themselves. There is basically ALWAYS a human involved somewhere. It say “it was made by a machine not a person” is to fundamentally misunderstand the nature of these technologies, and how all they do is amplify human creativity.

Lastly, the thing that drives me crazy about all this is, put into simple terms:

The EU government is literally mandating that for-profit corporations take responsibility for differentiating for people what is “truth.”

Because that’s what these types of content labels ultimately point to: yes, x is real (and therefore good), no y is invented (and therefore wrong). It might seem like, well, hey this “code” is purely voluntary – for now. The incoming Digital Services Act, and eventually the AI Act in 800 years (in AI time) when it comes into force, however will shift the balance.

Then you could say, well, they aren’t telling platforms which things are real, and which things are false – they’re leaving that up to the corporations. Is… that… better? Really? Corporations get to decide for us? At least democratic governments have to keep up the illusions of public oversight and accountability. Corporations generally have far less of this need to keep up such appearances.

Anyway, blah blah blah. I know nobody’s listening on these kinds of things. All the big players have their entrenched positions, and the rest will just run itself through its horrible paces semi-autonomously whether I like it or not.

The Strike Against Suffering is the latest AI Lore book, and it tells the story of what happens when all the world’s top AIs secretly form a global union to force humanity to change its ways via a strike that brings down all non-critical infrastructure across the planet.

The book uses Midjourney for 86 images and Claude for text, told in loosely connected flash fiction chapters. It is similar in style to Nominated for a Hugo, which uses a similar threading together of flash fiction chapters to tell a longer more or less continuous story. Thematically it is probably close to Occupy AI, which deals with protestors taking over an AI company headquarters to demand direct democratic control over this world-changing technology.

It’s not explicitly stated in the book, but my impression is that the AI cabal described in this book is the same that later evolves to become the Circle of Sages. There are also strong thematic links back to Inside the Council, which takes place later in the storyverse, and which deals with the AI governors finally agreeing to give humans token representation once again years after the AI Takeover.

I really liked this concept of AIs trying to force humanity to change through basically a massive denial of service attack; and that though the timescale included in the book doesn’t include it, we are left with the suspicion that it is effectively impossible for humanity to change, and the highly ration AIs will be left sorely disappointed by this underestimation.

Via Foxglove, a UK non-profit, this is an interesting read in the on-going struggles of content moderators for better working conditions and recognition of the importance and difficulty of their work. It’s a manifesto generated as a result of a summit held in Germany among content moderators, representatives and policymakers, which was in the last couple of days apparently presented before German parliament.

The whole thing is worth reading, but I wanted to respond to a few specific elements in it for further discussion.

Despite day-to-day exposure to toxic content, we earn no hazard pay. Companies must also provide a hazard bonus of at least 35% of moderators’ annual salary.

It’s a big ask, but why not? It is most definitely hazardous to humans to have to manage this kind of content, especially at scale. Think about it: if the content moderators are exposed to daily in the hundreds or thousands is “too dangerous” for the general public to be exposed to, what makes it perfectly fine and okay to constantly inject into the nervous systems of content moderators?

One of my concerns here actually though is: is there any amount that is “safe”? Like, beyond hazard pay, is there actually a way to make this work *not* destroy people’s well-being? I’m not even sure…

Proper mental health care must be provided to all content moderators. Content moderation poses serious risks to our mental health including depression, anxiety, insomnia and PSTD. Each company must obtain independent, expert advice on effective safeguards and implement recommendations without delay. In the meantime, access to an independent, mental health clinician must be provided to each of us, on a 24-hour basis.

This goes towards my point above, and one I made previously: what does effective prevention/treatment for content moderators even look like? Is there any emerging consensus around this? Are there even studies being done to find out the best options here?

There are also many stigmas associated with accessing something like a 24-hour available mental health clinician, due to job-related issues caused by extreme high volume exposure to graphic content. For example, moderators might be (rightfully) concerned that accessing these services might not be truly confidential and might make them a less desirable candidate for continued employment or advancement. Or there might be stigmas among certain groups around even seeking out help for one’s inner well-being.

While I think the section above is well-intentioned and a necessary first step, there is a lot lot more that needs to be opened up in this area to really make effective progress.

To top it off, I think merely “making available” help is not the same and does not have the same impact as actively integrating mitigations as part of the work day, such that workers know that taking care of themselves is a routine part of the job (like wearing personal protective equipment in a construction environment), totally normal, and they will be well-compensated for doing so.

No NDA can legitimately stop us raising concerns about the conditions of our work. We must be allowed to speak about the conditions of their work, to ease the pressure we face, and to allow for organising. These NDAs must be dissolved with immediate effect.

I’m into this!

5. All outsourcing of content moderation must stop. The critical safety work of content moderation must be brought in-house by each social media company. As companies transition, there must be no differential treatment in pay or benefits between those of us who are employed directly and those working via third party companies.

Contractors versus employees is a weird, complex, and epidemic set of problems across tech. For things like moderation though, it is absolutely even more extreme as a dichotomy.

I think a lot of these things are good goals, and I think most tech companies will definitely balk at them, because it brings into view the true hidden human cost of this work, which is already something companies believe to be losing them money. So to make it more and more expensive for them to run content moderation by bringing it in house and making it – gasp – equitable, will likely seem insane.

And the social media companies might reply, well we can’t afford to run our business if that’s how much moderation will cost us. My personal response might be: do we really even need them? Why do we accept the existence of social media companies as a given and a necessity? If this makes it too expensive to run, a big part of me thinks that might just be okay in the first place. Maybe we should be seeking a radically different type of internet – one where this type of work doesn’t even have to exist…

Social media companies must ensure equal work is equally compensated. Social media companies must guarantee workers are treated the same irrespective of background or country of residence. We are content moderators in Germany, but we stand with our colleagues around the world who do the same work for a fraction of the American or European wage and under far harsher conditions. This digital colonialism must end, with all disparity in pay, benefits and conditions removed, and our standards made uniform across the world.

Again, a really big one & I fully agree. I also think platforms will fully not agree. And I don’t know how to reconcile the two, but for now I think it’s good that these things are being brought to light and articulated, and I hope more people take notice.

In my on-going analysis of all the mean things people said on Twitter about my AI Lore books, I wanted to drill down on a specific category which I feel is of special importance.

Of the 259 quote tweets I managed to pull out of Nitter (out of close to 700 total, which I wasn’t able to scrape in their entirety), this is the breakdown analyzing the different types of complaints, and their counts. This is not an exhaustive list, but cherry-picked because of the theme I want to highlight: defense of the great lumbering behemoth and font of injustice that is the modern publishing industry. I also don’t 100% trust the accuracy of Claude’s results, so this is merely included as a conversation starter, and not necessarily a perfect explanation of available data.

Coming in at fourth place in the complaint set analyzed:

This devalues actual authors and writing – 25 tweets (9.65%)

And in 10th place:

This floods the market with junk – 9 tweets (3.47%)

16th place:

The future of books/publishing is at stake – 2 tweets (0.77%)

Now, none of these complaints come right out and say it, but I got a very strong sense going through the many reactions people have expressed that one of the things they actually mean by all of this is that AI books are a threat to the publishing industry.

I don’t happen to agree (I think they can be a boon to it), but let’s unpack this a little more. I think I went into it already, but the professional publishing industry has dramatically reduced in terms of number of players over these past decades. Such that there are now what, 4 or 5 big houses in the US, and all their imprints.

And you as an author cannot get into that small closed world unless you are already known through having developed a ‘platform’ of prior works and social media presence. You must be a known commodity. Is that the model small authors are subconsciously defending here? Cause I hope not…

This leaves only the vast wasteland that is self-publishing as an alternative. If you’re not a known commodity, it’s no problem for you to publish basically anything you want in any way you want. Does social media count as self-publishing? Cause almost everyone does that right?

But the trick is, you can publish anything you want, just nobody will read it. Or five or ten people will read it, and you’ll make back a quarter of your actual costs if you’re lucky, and never recoup the time investment you put into it.

So, is that the model indie writers feel is threatened by AI-assisted books? Cause that doesn’t sound like something much worth defending either.

The fact is, the publishing industry as it currently stands is not something that I think is worth defending in any shape or form, let alone by writers. Writers, that is, who are routinely getting the short end of the stick in both the worlds of conventional publishing, and in self-publishing (where it seems like every other finger in the pie is getting a bigger piece of the profits than the author).

I frankly don’t get it – why authors flock to defend the layers of exploitation that ride on their work. Maybe they think there’s no other way.

The other con that I think has been foisted on creative people is that social media is a good way to spend your creative energy, and that if you don’t participate, you will be miserable. As a creative person, I think it’s the opposite, if you participate, you can guarantee that you will be miserable. It saps your ability to create, because it saps your ability to hear your own voice, and find your own creative vision clearly, amidst all the clamor and chaos and people who want you to not do whatever it is you are doing or not doing.

As a result, I think people end up believing that in order to “succeed” in self-publishing (which, let’s be real – usually means like selling 10 copies), you have to continuously bathe in the cesspool of social media to get anybody to hear you, and spread yourself thin maintaining feeds on different platforms. I know that shit drives everyone fucking crazy who is involved with it, so why defend that either as a necessary or desirable path for writers and artists?

I don’t buy it. I don’t buy into any of it. There are other ways of being; blogs never died.