…of all the actions a man may take in order to create himself as he wills to be, there is not one which is not creative, at the same time, of an image of man such as he believes he ought to be.

Whenever anyone states that “AI is the future, so…” or “many people are using AI anyway, so…” they are not only expressing an opinion — they‘re shaping that future.

I read this sort of opinion fairly often in my RSS stream, which may be surprising considering my views on LLMs and therefore who I choose to follow.

Perhaps it’s because of a residual respect for exciting new tech among folk who, like me, spend a lot of time with little computing machines. Maybe some don’t like discussion with an adversarial tone, where they’re asked to make a choice about something beyond its supposed technical utility. Maybe they’ve found what’s broadly termed “AI” useful at some point. Maybe they even think LLMs are a societal good, having not experienced any direct harm.

Anil Dash published one of these articles recently, essentially arguing that because many millions of people are freely using LLMs, Mozilla’s choice to integrate “safe” AI into Firefox is a net, protective good.

It bore all the hallmarks of an “it’s the future, so…” argument. Firstly, there was the breezy “this call will be recorded before we move on to your query”-like list of AI’s harms in the first paragraph (Good morning, just a few things before we deal with your question. Genocide, mass unemployment, fascism, misinformation, environmental destruction. Now, how can I help you today?)

Then we have the (unlinked) assertion that “millions of people find AI-generated media images emotionally moving”, which might be true, but doesn’t represent LLMs’ main, more worrying use — the majority of the active weekly users of ChatGPT are students.

And then the appeal to a silent majority of “normal”, reasonable, but ultimately passive users — your friends and family no less — who, I hate to break it to you, while they are there to be persuaded, don’t like your tone. Compare these folk with the “rabid” “zealots” criticising Mozilla for adopting AI.

As ever with these sorts of articles, there’s a profound unwillingness to actually engage with an argument unless it’s in abstract terms, and an equally profound dislike of “ethical” discussion that asserts you have to choose a side. Instead, we resort to objections to tone and appeals to your average man or woman in the street — to marketing, in other words.

The thing is, offering a “safer”-to-Firefox-users version of a technology that runs the catastrophic gamut of “undermining labor to appropriating content without consent to having egregious environmental impacts to eroding trust in public discourse” is weak tea, even if you did trust Mozilla to somehow protect you, rather than simply offer a whitewashed, “ethical” brand of AI to the market, or for it not to be staffed by ex-Facebook execs. It’s like hoping for a nice fascism.

No, there is no meeting the AI behemoth half way: if you feel it’s structurally inimical to a healthy society, you should call it out as such, rather than mithering about people who do object, or appealing to a silent majority of “normal” people in an effort to use it safely.

Mozilla should position itself as the anti-AI and anti-advertising browser, thereby expressing what web they want through choosing the right path. Going all in on a “safe” version is at best shuffling the deck chairs, at worst creating a nice brand for egregious technology.

Similarly, if there is a point to our blogging about the state of the world, it is to express how we want the world to be through our choices about it, whether we’re a blogger with half-a-dozen subscribers, or a tech influencer.