I think that’s part of the problem, their new algorithm essentially railroads you into the same results that you’ve looked at before which becomes self reinforc

I think that’s part of the problem, their new algorithm essentially railroads you into the same results that you’ve looked at before which becomes self reinforcing. Part of the magic that’s missing now is the kismet of searching for something and finding a tangentially related result that happens to be better than the thing you were looking for. Now you have to be aware that your results are being filtered by your past behaviour and that of your cohort and actively fight against it which is terrible user experience.