Meta is surprising us by currently testing AI-generated comments. This type of commentary is appearing on some Instagram pages. A feature that raises questions about the very notion of "engagement." on social networks.
As we know, Meta has been banking on artificial intelligence-boosted features since the beginnings of generative AI. But the latest innovation The feature that the social media giant is currently testing on Instagram is sure to raise eyebrows. These are pre-made comments generated by AI, which users can post directly to certain pages with a single click.
The feature is somewhat reminiscent of the reaction suggestions offered by platforms such as LinkedIn. According to the site SocialMediaToday, the option is only available on certain accounts, pages, and content. In concrete terms, the new feature is represented by a new icon next to of the comments field on eligible Instagram content.
Is a feature that automatically generates comments really necessary?
The icon in question looks like a pen with a star on it. When you tap this icon, a menu offers you a series of three pre-made comments, generated based on the context of the publication. All with necessarily rather superficial wording. On paper, this new feature should encourage members of the social network to achieve more engagement with content.
As a reminder, the notion of engagement is a key metric for estimating the success of an account or brand on social networks. It is in itself more important than the mathematical count of the number of subscribers or followers. Engagement refers to the propensity of a page's content to encourage Internet users to comment or react. The more engagement an account generates, the more likely it is to be highlighted by the algorithm.
Using AI to generate comments then seems to pose a major problem associated with this commitment: that of the quality of user reactions. If a page, a brand, or an account generates a lot of reactions in the form of comments generated by AI, is it really the same thing as comments written spontaneously by Internet users?
Beyond that, the implementation of ever more artificial intelligence on social networks can raise some fears. Especially on American platforms in a troubled period, where generative AI is already helping certain actors to spread mass manipulation. And even though actors are already relying on on bots and other large-scale distribution techniques.
For now, in the absence of official information from Meta, it is difficult to know if this new option is intended to become anything more than a limited-time test.
0 Comments