> It's incredible that these shitty ML hueristic systems are the best these "genius" FAANG developers, who get paid $150-300k a year, can come up with up. I th

> It's incredible that these shitty ML hueristic systems are the best these "genius" FAANG developers, who get paid $150-300k a year, can come up with up. I think it gets personalized / fingerprinted fairly quickly TBH. For example, I have 2 identical (Linux + Firefox + uBlock Origin) VMs that I use over the same VPN connection. One of them gets almost no captcha challenges and the other gets them continuously. My subjective experience is that it's based on what you're searching for on Google or maybe due to hitting sites that Google might have flagged as malicious (guilt by association). It's tough to tell, but it sure feels like there's some type of cumulative score based on activity.