We're considering a proposal to revamp the current homepage feed to make it primarily focused on news stories about current events. We would limit the news stories we display to ones from high quality publiclation that we've algrorithmically determined will be relevant to your interests, and will display 1-3 contextually relevant markets below each story. Read more here.
At the end of the year, all people working on Manifold will take a Yes/No vote. If the majority vote the Newsfeed as the most impactful new feature from 2023 this market will resolve YES. A split vote resolves NO.
If we decide not to implement this feature (or if the feature we do implement looks sufficiently different), I will resolve this market N/A (even if we do end up implementing it later).
The typical Manifold user right now only seems to care about 30-year AI speculation, so I'm not sure how much value those users will get from headline news.
But, well-known news stories might give a better impression of the site to new potential users who aren't immersed in AI hype or meme markets.
So my question would be, how would this be presented to new users who aren't logged in and who don't have any interest data in the system? Top news stories overall?
Take the average embedding of last n markets user bet on, show similar ones on home page.
Let people bet $1 with their IP address (one per market per hour, and ~5-10 bets per ip) then transfer when they sign up
turn on word, user, and soft-embeddings muting. Have good defaults
Add a chat prompt and let people interact entirely with chat ☠️
Reflects how I advertise Manifold to my friends. "Hey, I have weirdly specific and high-quality forecasts about current events because of this site."
bullish on the feature, don't think it will revolutionise Manifold.
@DavidChee idk what else you guys have planned, but i think it's more likely to revolutionize manifold than leagues or private groups
@DavidChee Bearish on this feature, but think if it works it would revolutionize manifold
@MichaelWheatley I strongly agree. Risky but would vastly improve the site if it works.
From the link, it seems like variations of the idea involve user curation of news stories and markets, and that seems like a safer bet.