BETA nonprofit public democratic european moderated

Balance ton Claude

1 posts
Profile compiled from news and social media. Claim profile

Balance ton Claude

Dear readers. We call for moderation in the attacks against Thélyson Orélien. First, there is a non-zero chance (even if it is tiny) that he is acting in good faith. Moreover, Orélien undoubtedly played in a morally gray area, exploiting a period of confusion during which our institutions and our practices (as well as our literary awards) have not yet adapted to this new and transformative technology. But he probably did not think that the success of his book would gain national prominence; and while it seemed important to us to point out the scam, it also seems important to have compassion and empathy for him, and even to support him during this difficult time he must be experiencing. The best thing would, of course, be for him to start writing again! The lessons of this affair go beyond his personal case, and that is what we must focus on. We reported this book because it is successful, but it is far from being the only one. On this network, hundreds of politicians and slop accounts have their posts written by AI without disclosing that it is AI, deliberately deceiving their readers all year long. And it has been a long time since the bestsellers on Amazon (the leading book distributor in France) have been dominated by books written directly by AI, sometimes very crudely, and with results not as good as Thélyson's (who thus has a degree of merit for managing to produce a coherent narrative that has satisfied many readers). The Parisian intelligentsia still likely underestimates the capabilities of AI and uses LLMs little or not at all (and does not hang out on LinkedIn), which is why they have not recognized the stereotypical formulas. Formulas that, if we were not tired of seeing them incessantly, would probably please us: language models learned them because they appealed to humans who supervised their training via RLHF! In the comments, some express that they do not see the point of authenticity; perhaps what should be taken away from this controversy is that, given the scale of the scandal, the majority places a lot of importance on authenticity. If you believe you are providing value by having your works written by AI, go ahead! But at least mention it, so as not to betray your readers, because lying, in this case, is morally reprehensible. (translated)