4 ms·

I see it as not unlikely that there'll be a campaign to sigmatize, if not outright ban, open source models on the grounds of "safety". I'm quite surprised at ho

I see it as not unlikely that there'll be a campaign to sigmatize, if not outright ban, open source models on the grounds of "safety". I'm quite surprised at how relatively unimpeded the distribution of image generation models has been, so far

int_19h 3y ago
This is already happening, actually, although the focus so far has been on the possibility of their use for CSAM: https://www.theguardian.com/society/2023/sep/12/paedophiles-using-open-source-ai-to-create-child-sexual-abuse-content-says-watchdog https://www.theguardian.com/society/2023/sep/12/paedophiles-...
AlanYx 3y ago
What you're predicting has already started. Two weeks ago Geoffrey Hinton gave a speech advocating banning open source AI models (see e.g.: https://thelogic.co/news/ai-cant-be-slowed-down-hinton-says-in-call-to-ban-open-source-models/ https://thelogic.co/news/ai-cant-be-slowed-down-hinton-says-... ). I'm surprised there wasn't an HN thread about it at the time.