以下部分引用 AI 总结,现阶段 AI 仍然有幻觉。请以内容的原文为准。
数据新鲜度群聊 07-14 ✓卖方 07-14 ✓Wrap 断档15天 ✕News 07-15 ✓X 断档12天 ✕页面生成 07-15
Twitter/X
13 atoms · 跨 5 天 · 首见 2026-05-22 · 最近 2026-06-09
三色: 🟦 fact 7 · 🟥 take 6 · stance ▲0/▼5/◆1
来源: X 13
时态: fresh:13
标签: 好观点:5 · 好数字:4 · 好思考:3 · 好问题:1
叙事 (narrative) (3)
- 2026-05-24 · ranking stage scores purely on predicted engagement, no weight for credibility
The ranking stage scores purely on predicted engagement with no weight for credibility, expertise, or track record... This seems sad....
tags: 好观点 · → daily
- 2026-05-22 · 点赞不带来流量提升
when people liked your posts - you got a boost. That makes sense too. This doesn’t happen now.
tags: 好观点 · → daily
- 2026-05-22 · 书签被当做新的流量信号
the new signal they game is bookmarks
tags: 好观点 · → daily
事实 (fact) (10)
- 2026-06-01 · 营收
Twitter/X: $1.8B
tags: 好数字 · → daily
value: qty=$1.8B
- 2026-05-30 · 广告收入下降
Advertising revenue on Twitter/X has dropped 59% since Elon acquired the company.
tags: 好数字 · → daily
value: qty=59% · direction=下降
- 2026-05-24 · auto-translate feature rollout date
before that date, if you wrote in English about, say, the Trump-Xi Beijing summit, you were competing for attention with maybe 5,000 other English-language accounts writing on geopolitics. After that date, your post is competing for attention with other posts on the same topic IN EVERY LANGUAGE ON EARTH.
tags: 好数字 · → daily
value: date=2026-04-07
- 2026-05-24 · ranking stage uses 15 engagement metrics
_Then there's the ranking stage. Each of these candidate posts for your feed is fed through a Grok-based model that tries to understand if you'll engage with the post. It looks at 15 engagement metrics: 1) P(favorite) ... 2) P(reply) ... 3) P(repost) ... 4) P(quote) ... 5) P(click) ... 6) P(profile_click) ... 7) P(video_view) ... 8) P(photo_expand) ... 9) P(share) ... 10) P(dwell) ... 11) P(follow_author) ... 12) P(not_interested) ... 13) P(block_author) ... 14) P(mute_author) ... 15) P(report)_
tags: 好思考 · → daily
- 2026-05-24 · May 15 update added impression bloom filter
The May 15 update added an 'impression bloom filter,' tightening the rule that once a reader has been served a post, the system won't serve it to them again.
tags: 好数字 · → daily
value: date=2026-05-15
- 2026-05-24 · author diversity scorer penalizes multiple posts in same pool
An 'Author Diversity Scorer' inside the ranking stage attenuates the score of every subsequent post of yours that ends up in a reader's candidate pool.
tags: 好思考 · → daily
- 2026-05-24 · repost distribution power vastly demoted in new algorithm
In the new algorithm, that mechanism is vastly demoted: reposts - like every post - need to go through the retrieval and ranking stage mentioned above, so a repost from a big account is a long way from the boost it used to be.
tags: 好思考 · → daily
- 2026-06-09 · 产品发布状态
Holy shit they actually launched it
tags: 好问题 · → daily
- 2026-05-24 · ranking model does not include credibility, expertise, or track record
there is no P(this post is true and well-sourced). No P(the author actually knows what they're talking about). No P(this person has spent a decade building a body of work that has held up). No P(this account has earned the right to be taken seriously on this topic). No P(the author has a large following from credible people). The model does not seem to care - at all - about any of that.
tags: 好观点 · → daily
- 2026-05-24 · low-effort quote tweets often cannot clear retrieval stage
This is especially brutal for low-effort quote tweets, which used to function as cheap amplification: now they often can't even clear the retrieval stage - they simply don't contain enough novel semantic content for the system to match them to anyone's interests.
tags: 好观点 · → daily
时间轴 (近 20)
- 2026-06-09 ·
fact · 产品发布状态 · → daily
- 2026-06-01 ·
fact · 营收 · → daily
- 2026-05-30 ·
fact · 广告收入下降 · → daily
- 2026-05-24 ·
fact · auto-translate feature rollout date · → daily
- 2026-05-24 ·
fact · ranking stage uses 15 engagement metrics · → daily
- 2026-05-24 ·
fact · ranking model does not include credibility, expertise, or tr · → daily
- 2026-05-24 ·
fact · May 15 update added impression bloom filter · → daily
- 2026-05-24 ·
fact · author diversity scorer penalizes multiple posts in same poo · → daily
- 2026-05-24 ·
fact · repost distribution power vastly demoted in new algorithm · → daily
- 2026-05-24 ·
fact · low-effort quote tweets often cannot clear retrieval stage · → daily
- 2026-05-24 ·
narrative · ranking stage scores purely on predicted engagement, no weig · → daily
- 2026-05-22 ·
narrative · 点赞不带来流量提升 · → daily
- 2026-05-22 ·
narrative · 书签被当做新的流量信号 · → daily
← 实体目录 · 系统日志