It seems to me that this would actually be easier for software to bypass than a regular image based CAPTCHA, it just consumes a bit more bandwidth. Pulling indi

It seems to me that this would actually be easier for software to bypass than a regular image based CAPTCHA, it just consumes a bit more bandwidth. Pulling individual frames out of a video is simple, so you just pull a few frames out of the video and run your regular image CAPTCHA breaking algorithm on them. Since you've got multiple frames, your software has multiple attempts to compare which makes it even more robust. Meanwhile it doesn't solve any of the actual problems with CAPTCHAs that others mentioned.