I like the conclusions they come to in their letter to social media, where they are requesting allowing researchers more insight and control, essentially enabli

I like the conclusions they come to in their letter to social media, where they are requesting allowing researchers more insight and control, essentially enabling them to sample/simulate the recommendation engine and get more insight into how it clusters or relates contents. Furthermore they want to reduce limits on API keys so that more comprehensive academic research can be performed. They know the issue isn't so much the algorithm as those that are gaming it to empty extraction of peoples attention (== money), extreme, or harmful ends. In doing so we might better understand the problem and then provide tools or recommendation to social media and the public on the topic.