TE
TechEcho
Home24h TopNewestBestAskShowJobs
GitHubTwitter
Home

TechEcho

A tech news platform built with Next.js, providing global tech news and discussions.

GitHubTwitter

Home

HomeNewestBestAskShowJobs

Resources

HackerNews APIOriginal HackerNewsNext.js

© 2025 TechEcho. All rights reserved.

Ask HN: How would you design a cringe detector?

2 pointsby mib32over 2 years ago
I have a dream of making a cringe detector - a program that you could input a tweet or image, and it outputs how cringy it is.<p>I am thinking, it could use a neural net, or not necessary.<p>I was trying to imagine, how could that thing work? And actually out of clue.<p>What do you think?

3 comments

armchairhackerover 2 years ago
look up “sentiment analysis NLP”. That is what you want, except instead of detecting whether text has a positive or negative opinion, you detect whether it’s “cringe” &#x2F; “not cringe”.<p>I would start by building a sentiment-analysis model by following a tutorial like [1]. Except instead of a dataset with data labeled “positive opinion” and “negative opinion”, you need data labeled “cringe” and “not cringe”. The hard part is finding and classifying that data: remember that if you just take a bunch of “cringe” and ordinary tweets, you may be training for spurious correlations like “hot take” which your model will misinterpret as being “cringe”<p>[1] <a href="https:&#x2F;&#x2F;huggingface.co&#x2F;blog&#x2F;sentiment-analysis-python#3-building-your-own-sentiment-analysis-model" rel="nofollow">https:&#x2F;&#x2F;huggingface.co&#x2F;blog&#x2F;sentiment-analysis-python#3-buil...</a>
red_Seashell_32over 2 years ago
Based on what it would calculate it? It’s a spectrum just like “how funny is this”
newbieuserover 2 years ago
if trained via reddit, i think you could call any content cringe. maybe by finding people from different segments to rate whether certain content is cringe.