4 ms·

Safety from AGI, sure. But I've could argue there will still be safety concerns from the deepfakes and scams OpenAI's tech is sure to proliferate

pid-1 2y ago
I wonder if Adobe Photoshop has a safety team
- Nullabillity 2y ago
  Yes: https://en.m.wikipedia.org/wiki/EURion_constellation https://en.m.wikipedia.org/wiki/EURion_constellation
swatcoder 2y ago
But if you assemble a team of leading public figures and researchers caught up in AGI catastrophe scenarios, and then try to direct that team to just implement some guardrails against deepfakes/scams/taboos, you're quickly going to find them frustrated. It's like taking a star NFL/NBA team and sending them to play a season of exhibition games in a foreign market instead of continuing to compete in the peak league of their sport.
- BadHumans 2y ago
  This analogy doesn't track because Lebron would run circles around any foreign basketball team while OpenAI is not keeping up with deepfake detection.
londons_explore 2y ago
> safety concerns from the deepfakes and scams Those will happen whether or not you have a safety team. The best way to protect against these is to widely distribute deep fakes of famous people so everyone is aware that faking things is possible.
- passwordoops 2y ago
  This would certainly raise awareness, but how exactly would it protect against deepfakes?
  - londons_explore 2y ago
    Don't protect against deep fakes. Make everyone know that deep fakes are possible and then they lose their power to influence.
nroets 2y ago
Social media was already rife with fake news and half truths before ChatGPT. And there are already open source LLMs that can be trained to scam a significant number of people. OpenAI can't do anything about it. The only remedy is social media warning people.
cyanydeez 2y ago
We need safety from darkpatterns worse than twitter and facebook.combined.

pid-1 2y ago

I wonder if Adobe Photoshop has a safety team

Nullabillity 2y ago
Yes: https://en.m.wikipedia.org/wiki/EURion_constellation https://en.m.wikipedia.org/wiki/EURion_constellation

swatcoder 2y ago

But if you assemble a team of leading public figures and researchers caught up in AGI catastrophe scenarios, and then try to direct that team to just implement some guardrails against deepfakes/scams/taboos, you're quickly going to find them frustrated. It's like taking a star NFL/NBA team and sending them to play a season of exhibition games in a foreign market instead of continuing to compete in the peak league of their sport.

BadHumans 2y ago
This analogy doesn't track because Lebron would run circles around any foreign basketball team while OpenAI is not keeping up with deepfake detection.

londons_explore 2y ago

> safety concerns from the deepfakes and scams Those will happen whether or not you have a safety team. The best way to protect against these is to widely distribute deep fakes of famous people so everyone is aware that faking things is possible.

passwordoops 2y ago
This would certainly raise awareness, but how exactly would it protect against deepfakes?
- londons_explore 2y ago
  Don't protect against deep fakes. Make everyone know that deep fakes are possible and then they lose their power to influence.

nroets 2y ago

Social media was already rife with fake news and half truths before ChatGPT. And there are already open source LLMs that can be trained to scam a significant number of people. OpenAI can't do anything about it. The only remedy is social media warning people.

cyanydeez 2y ago

We need safety from darkpatterns worse than twitter and facebook.combined.