研究内容为英文. 本页的研究内容(标题、简报、来源)以英文撰写,未经机器翻译。只有界面已翻译。

新想法Consumer social

Algorithm Mirror

Upload a few screenshots of your feeds (TikTok For You, Instagram Explore, Pinterest, YouTube) and get an AI portrait of "who the algorithm thinks you are"; then invite friends to answer six quick questions and get a side-by-side "Algorithm vs Friends" card to share.

3周前 · 评分于 2026年10月20日 · ID HBV-261020-05

摘要

简报假设

New idea. Existence check: Searches (20 Oct 2026) found no dedicated generator for "how the algorithm sees me" from feed screenshots, and no product that pairs it with friend answers. Closest analogs: ChatGPT caricature prompts (manual trend), Spotify Bedroom (internet-bedroom.com, room from Spotify data), EraBoard and…

阅读其余部分

Decor8 (aesthetic quizzes), oomfmap.com (AI personality grid from an X handle; X mentions 76% below peak), CutNoodle X Recap ("X personality"), Wimo brain-map template, NGL and Tellonym (anonymous friend questions). Partial overlap with HBV-261019-04 Rage Index (feed analysis), which scores rage bait rather than identity. X: 2,949 posts/7d (Up 6% Wed→Mon (full UTC days: Wed 14 234 → Mon 19 249); peak Fri 16 762; Mon 19 is 67% below peak — ≥50% below peak: POP 5.3 window-closed signal; window 2026-10-13 10:30 UTC → 2026-10-20 10:00 UTC (Tue 13 Oct and Tue 20 Oct are partial UTC days; full days Wed 14 → Mon 19)). Revenue hypothesis: ~$2K–$5K/month at month 6 (HYPOTHESIS: 200k cards/month × 0.5–1% buyers × ~$3).

描述假设模型

Upload a few screenshots of your feeds (TikTok For You, Instagram Explore, Pinterest, YouTube) and get an AI portrait of "who the algorithm thinks you are"; then invite friends to answer six quick questions and get a side-by-side "Algorithm vs Friends" card to share. Who: Gen Z and young millennials (16–30) who post identity memes and "how X sees me" trends on TikTok, Instagram and X; first in English, then Portuguese and Spanish.…

阅读其余部分

Pain: People are curious and uneasy about what recommendation algorithms "think" of them and enjoy comparing that to how friends see them, but today they have to improvise with ChatGPT prompts and cannot compare with friends. Monetization: Free card; one-off paid HD poster and "deep read" (about $2–$4, hypothesis); optional monthly re-scan; No approved price. Build: 1–2 agent-weeks (screenshot upload with on-device redaction, vision model labeling, portrait generator, friend-invite form, card renderer, checkout); stack: Next.js on Vercel; vision LLM for labeling; image model for the portrait; Postgres; Stripe; server-side cards. GTM: Share cards on TikTok, Instagram and X; friend-invite loop (each user needs friends to answer); faceless trend accounts; SEO pages ("how TikTok sees me").

目标客户

Gen Z and young millennials (16–30) who post identity memes and "how X sees me" trends on TikTok, Instagram and X; first in English, then Portuguese and Spanish.

痛点

People are curious and uneasy about what recommendation algorithms "think" of them and enjoy comparing that to how friends see them, but today they have to improvise with ChatGPT prompts and cannot compare with friends.

为何是这个分数

分数较低的原因: Fad decay: the core X trend fell 67% from its peak within three days, so the friend-invite loop must carry growth after the trend moment.

护城河假设

Friend-answer graph and card library: The moat is the invite graph and the best-known card format, not the model calls. Repeat scans and friend comparisons create a reason to come back; competitors start without friend data. Status: Hypothesis.

需求信号

X 上的帖子(7 天)

2949

信号趋势

Up 6% Wed→Mon (full UTC days: Wed 14 234 → Mon 19 249); peak Fri 16 762; Mon 19 is 67% below peak — ≥50% below peak: POP 5.3 window-closed signal

网络需求已验证

The "how the algorithm sees me" wave peaked on Thu 15–Fri 16 Oct (747 and 762 posts) and fell to 249 on Mon 19; the "how others see me" prompt family is steady at ~100 posts/day. Korean saju brain maps (A1 today) show the same appetite for identity images. Other X counts (same window, Verified X API): 765 posts: ("how …

阅读其余部分

my friends see me" OR "how others see me" OR "how people see me" OR "describe me in one word" OR "first impression of me" OR "what do you think of me") -is:retweet

市场与竞争对手

TAM

Not sized this run — Scope: Paid personality and identity images. Method: No market report fetched. (Basis: n/a)

SAM假设

~$24K–$60K / yr (hypothesis) — Scope: English-language cards. Method: 200k cards/month × 0.5–1% × ~$3 × 12 (HBV assumption). (Basis: HBV assumption)

SOM假设

~$24K–$60K ARR (hypothesis) — Scope: Month-6 run-rate if 200k cards/month. Method: as SAM. (Basis: Low (hypothesis))

竞争对手

Manual ChatGPT/Grok image prompts; Spotify Bedroom; EraBoard; Decor8; oomfmap.com; CutNoodle X Recap; NGL, Tellonym, Sendit; platform-native recaps (Spotify Wrapped, Pinterest Predicts).

商业模式

变现方式假设

Free card; one-off paid HD poster and "deep read" (about $2–$4, hypothesis); optional monthly re-scan; No approved price.

收入假设(每月)假设

~$2K–$5K/month at month 6 (HYPOTHESIS: 200k cards/month × 0.5–1% buyers × ~$3).

待完成的任务

When an identity trend is all over my feed, I want to get my own version in seconds and see if my friends agree, so I can post something that feels like me and talk about it with friends.

HXC

Maya, 19, Manchester (illustrative) · Student who posts "how X sees me" trends and moodboards on TikTok and Pinterest. Buying signal: Tried the ChatGPT version of the trend; would send the friend link to her group chat to see whether friends agree with the algorithm, and buy the HD poster for her wall.

构建与 GTM

技术栈

Next.js on Vercel; vision LLM for labeling; image model for the portrait; Postgres; Stripe; server-side cards

GTM 渠道

Share cards on TikTok, Instagram and X; friend-invite loop (each user needs friends to answer); faceless trend accounts; SEO pages ("how TikTok sees me")

风险

主要风险

Fad decay: the core X trend fell 67% from its peak within three days, so the friend-invite loop must carry growth after the trend moment.

全部风险

Value: Attention — Trend falling 67% from peak | Usability: Attention — Privacy of screenshots | Feasibility: Supported — Standard APIs | Business Viability: High — Low willingness to pay

下一步验证

Ship in 7 days; target 50,000 cards and a 20% friend-invite completion rate in 14 days; kill if invite completion is under 10%.

来源

  1. X API counts (window above)
  2. existence web searches (20 Oct 2026)
  3. internet-bedroom.com
  4. oomfmap.com (X news, Oct 2026)
  5. cut-noodle.com
  6. wimouniv.com

ideaidentityshare-cardai-portraitfriend-invite-loopb2cgen-zglobalbrazillatam

想在评分当天收到下一个想法吗?

分数、来源、风险和下一步,通过邮件发送。

免费。通过 beehiiv 发送。一键退订。

这是研究,不是投资建议。数字保留其发布时的标签。