What confuses me the most is that nobody talks about the nature on how to exploit this vulnerability.

This has been a well known vulnerability for years but considered too complicated to exploit. It is the type of work that would have taken an expert weeks to hack into a single system and the idea that an AI can speed this is up is not unfounded but currently without evidence.

Yes, the attack surface is huge but when compared to Log4J this is materially more difficult to execute and the countless PoC that have been seen were AI slop.

As a security expert I see why you would not want to downplay this, but I think the involved actors have all the incentive to use this for their typical messaging.

As a tech community we have a greater responsibility to not be driven by hype but understand what is going on and let make ourselves some tools and be swayed by the messaging and instead talk about this in a more nuanced way. Perhaps I am wrong because I am not a cyber person myself but this all strikes me as similar to the whole Specter outrage.