221
Youtube AI filter making it dangerous to watch for people with Epilepsy due to a bug
(www.youtube.com)
this post was submitted on 26 Oct 2025
221 points (98.7% liked)
Technology
76415 readers
3706 users here now
This is a most excellent place for technology news and articles.
Our Rules
- Follow the lemmy.world rules.
- Only tech related news or articles.
- Be excellent to each other!
- Mod approved content bots can post up to 10 articles per day.
- Threads asking for personal tech support may be deleted.
- Politics threads may be removed.
- No memes allowed as posts, OK to post as comments.
- Only approved bots from the list below, this includes using AI responses and summaries. To ask if your bot can be added please contact a mod.
- Check for duplicates before posting, duplicates may be removed
- Accounts 7 days and younger will have their posts automatically removed.
Approved Bots
founded 2 years ago
MODERATORS
you are viewing a single comment's thread
view the rest of the comments
view the rest of the comments
If you have epilepsy, watch out so you don’t randomly see something that could potentially kill you.
Great job google.
FWIW, photosensitive epilepsy is typically only triggered at flash rates between 3-30 hertz. The rate of flashing shown is extremely unlikely to cause seizures even in generally susceptible individuals.
I don't know the type of flashes causes by this glitch but the post specifically warns for epilepsy.
That part of the post might be wrong, that’s completely fair and i trust you have less to gain from lying versus a post craving clicks.
Though my point is, if it is dangerous. What are they supposed to look out for? The “recommendation” should be google stop doing this shit that puts people at random risk.
If it’s just annoying, then its just a glitch, the same for other photosensitive but not suffering epilepsy people.
The danger is honestly pretty minimal for people who are aware they have photosensitive epilepsy; those who are prone to it but unaware of it are not likely to heed warnings even where they exist since they won't typically perceive the risk until after experiencing it.
It takes several minutes from triggering exposure to actual seizure onset, so those who know of their susceptibility have time to stop exposure and make sure they're in a safe position if a seizure does come. There are many ways of mitigating the seizure risk by stopping exposure, closing one eye and facing away from the light source, keeping screen brightness at the lowest level you can still easily read, etc.
That's not to say I think warnings aren't useful, but the intensity of many of the warnings people use is disproportionate to the actual risk and can cause people to be much more worried than necessary IMO. Google et al really need to stop messing with videos and such via AI without any sort of notice or warning for a whole host of reasons, including broader non-epileptic photosensitive since becoming intensely nauseous or getting a migraine over it is still pretty annoying.
tl;dr I think the warnings are a good idea, but maybe a little broader and less "OMG the epileptics are gonna all die". And fuck companies silently manipulating content they didn't even produce with AI in general.
Well said.
Did they ever give any reason why they even consider it an acceptable thing to do?
Imagine making an art piece to be displayed in a museum only to find they allowed an interim with next to zero experience to paint over it.
Not that I'm aware of, unfortunately. They seem to be trying to pretend they aren't even doing it, so telling us why they obviously are modifying content doesn't seem likely until they're backed in to a corner by popular outrage.