←back to thread

Show HN: A website that makes your text look cool anywhere online using Unicode

(fontgenerator.cool)
108 points liquid99 | 5 comments | 14 Mar 25 02:15 UTC | HN request time: 0.457s | source
Show context
gryfft ◴[14 Mar 25 02:33 UTC] No.43359090[source]▶
It's been mentioned elsewhere recently but this presents an accessibility nightmare for screenreaders and similar assistive technologies.
replies(4): >>43359382 #>>43359523 #>>43359799 #>>43360499 #
1. itake ◴[14 Mar 25 03:39 UTC] No.43359382[source]▶
but great for fraudsters trying to side step content moderation models!
replies(3): >>43359439 #>>43359657 #>>43359658 #
2. waltbosz ◴[14 Mar 25 03:52 UTC] No.43359439[source]▶
Are these models really able to be fooled by text tricks like this?
replies(1): >>43359713 #
3. h4ck_th3_pl4n3t ◴[14 Mar 25 04:46 UTC] No.43359657[source]▶
Ding ding ding! Billion dollar unicorn startup found!
4. scripturial ◴[14 Mar 25 04:46 UTC] No.43359658[source]▶
Unicode obsfucation tricks trigger modern content filters faster than you can blink. Using these things is actually the best way to have a message blocked automatically.

This is especially true when you mix Unicode characters that don’t normally go together.

(Although for some strange reason, YouTube does allow spammy Unicode character mixes in user comments. I don’t know why)

5. itake ◴[14 Mar 25 04:58 UTC] No.43359713[source]▶
It depends on what you mean by "models".

LLMs? No. But LLMs are too slow for content moderation at scale.

Custom trained models? Maybe. Is the unicode characters in the training data?