Future of social media what’s the recipe for pluralism, long term user value ..
summary
A multi-platform experiment runs 21 prototype algorithms, selects five and uses a custom browser extension reordering top 50 results on X Facebook and Reddit to test bridging‑based ranking and “surprising validators.” Over four months the intervention reduced polarization by three years on a feeling‑thermometer scale, shows no change in depression/anxiety, and raises reports of seeing negatively framed news; time‑on‑site effects remain unknown pending analysis of ~150 GB of data. Practical blockers: mobile limits, platform resistance, and need for interoperability and enforceable risk assessments.
AI-written summary · about
chapters (9)
- 0:00Social media’s future challenges
- 3:47Pro-social ranking results
- 10:57Algorithms for long-term value
- 19:44User feed control tools
- 27:12Platforms resist empowerment
- 31:29Legal rights to interoperability
- 43:50Q&A and policy debates
- 57:58Regulatory challenges
- 1:04:12Closing remarks
related talks
33mTired of the state of online conversations?
01h 15mSocial Media Recommender Systems Should Deliver Value, not “User Engagement"...
22mOnline Harassment
46mFacebook and Instagram trends in 2022 | Panel Discussion
46mFacebook & Instagram: Strategies, Perspectives and Trends 2022
04mInterview with Evgeny Morozov
49mFacing the Future
29mThe Psychology of Fake News (and What Tech Can Do About It)