A trader uses a trading platform and a service which uses algos to give broad analysis of a range of stocks that may be undervalued by the market. There's no genius involved there. The algos might have a stock on the radar for some factors like discoveries, but rate low a stock that misses targets. But the reason for the missed targets isn't given, as algos aren't perfect, they're only as good as the imputs. That's why more information is needed.
The trader selects a few possibles and starts asking leading questions in chat, giving the devil's advocate position.
LT holders fall over themselves to address their concerns, effectively doing their research for them. It's magic! They dig deeper, so then actual research to verify and then they may well decide to take a position.
The timing and depth of that later part is far from transparent. Certainly there is potential to argue vehemently against whatever they're actually doing in terms of a trade. That's why there is a disclaimer in HC when you sign on. Because often things are not as they appear.
Paradoxically, the intensity of the attacks can be inversely proportionate to how much they actually agree with you. Hence is it possible that decoded, the statement might read: "I like and respect you, so forgive me while I viciously attack you in every possible way but only because I like my shares cheap and you're getting in the way of my trades".
Counter signalling trading might be highly effective, and it's not as if I haven't thought of it. But there's some things I simply won't do, regardless of the money. Because I don't think self-loathing is a good enough residue once the dust settles. Some might be immune to this, but I'm saying I'm not. IMO.