I had fun analyzing the comments on a recent article from The Verge about AI’s energy consumption. The article itself stems from the publication of a new study on the subject, and I won’t delve into the question of « Is AI an energy sinkhole or not? », but rather on how people react to such articles.
Using Gemini, I analyzed the comments (strictly moderated by The Verge, which means not all reactions appear—I assume the most virulent or off-topic ones—but they are probably still representative).
3 Types of Reactions: Antis, Pros, and Disillusioned
Three groups stand out clearly:
The « Worried » Group
They are the most numerous, 57% (17/30). For them, AI is a danger that worries them. Their main arguments are as follows:
- They denounce a form of resource privatization. According to them, Big Tech is seizing energy sources.
- And this, for activities they find trivial at the expense of more useful activities like medicine or science. I note that no one here mentions job loss, which is a major concern for AI worriers, as the story of AI-generated ads clearly shows.
- Their demand is that instead of diverting electricity or water from the community for their data centers, big techs should build their own energy sources, like nuclear power plants, for example (which is already the case, in fact).
The « Anti-Worried » Group (30% – 9/30).
They are not really ardent defenders of AI, but they oppose the arguments of the fearful.
- They remind that the water used in servers does not disappear but is only evaporated.
- They put energy consumption figures into perspective by comparing them to other sectors, notably agriculture (with a nod to … cultivation in California, a notoriously water-poor state).
- They are convinced that AI will solve its own energy problems and remain optimistic.
The « Disillusioned » Group (13%, the rest)
These are the ones who seem tired of these endless debates. They have two reactions:
- For them, anyway, it’s too late. AI is here, and we will never go back. It’s like a steamroller that won’t stop.
- The irony: we read these articles and have these debates on devices (phones, computers) that are far more polluting than AI, hence their irony.
I also analyzed the tone of the commentators, and here too, it’s interesting.
A Striking Contrast in the Tone of Arguments
Again, and it’s striking, the tone divides into three groups, exactly the same as before.
The Worried: Highly Emotional Tone
The style of arguments is often very dynamic and very emotional. They are often filled with aggression, but more towards Big Tech than other commentators. In any case, urgency, indignation, and anger dominate. Their arguments are rather short and expressive.
The Anti-Worried: Calm and Superior Tone
What stands out in the second group is the calm and composed tone of the arguments. These are often well-argued (and longer) and delivered in a lecturing tone. Even if they are not aggressive in form, they are in substance and directed towards other commentators, notably with contempt for their technical knowledge of the subject.
The Disillusioned: Calm and Ironic
The third group, almost unsurprisingly, uses irony with short and terse sentences. The tone is not aggressive and merely marks disinterest or weariness for the subject.
This article is in no way representative and is not a statistic, but I think it clearly shows how the debate around AI is established. What dominates is still a negative sentiment (if we count the disillusioned and the worried). But no one is really anti-AI (I might not say the same if we were talking about Coca-Cola or McDonald’s ads).
There is a sort of acceptance or resignation that the advent of AI is inevitable. No one seems to really deny that AI consumes a lot of energy. But two camps oppose each other (and I’m sure if we could do a social analysis of the comments, we would have a clear separation between upper socio-professional categories and others) on the fact that some think big tech will self-regulate, relying on a kind of wisdom (not really proven in facts), while others probably rely more on state authorities to force big tech to source their energy elsewhere and spend less.
Gemini concludes, on its part, that the two camps seem desperately irreconcilable, but… isn’t that the life of social networks today (which does not please me, I tell you frankly).
Laisser un commentaire