OpenAI details its approach to AI safety, including evaluating AI systems, improving safeguards based on real-world use, protecting kids, and respecting privacy
OpenAI is committed to keeping powerful AI safe and broadly beneficial. We know our AI tools provide many benefits to people today.
OpenAI
Related Coverage
- OpenAI reveals safety measures used when building AI models like GPT-4 Tech Monitor · Matthew Gooding
- OpenAI shares AI safety approach amid ChatGPT controversy Silicon Republic · Leigh Mc Gowran
- OpenAI details how it's striving to improve the accuracy and safety of its AI systems SiliconANGLE · Mike Wheatley
- Amidst controversies, OpenAI insists safety is mission critical Mashable · Cecily Mauran
- OpenAI just promoted its approach to #aisafety in my LinkedIn feed. — This is almost certainly in response to #potus' recent remarks calling … Bill Bourdon
Discussion
-
@lukolejnik
Lukasz Olejnik
on x
“after our latest model, GPT-4, finished training, we spent more than 6 months working across the organization to make it safer and more aligned prior to releasing it publicly”. Yet, they still have not published a privacy impact assessment :) https://openai.com/... https://twitt…
-
@andrewcurran_
Andrew Curran
on x
OAI's new post. Only three things really. Under 18 are not allowed to use ChatGPT. Apparently it has always been this way? News to me and every kid on earth I guess. They are looking into ‘verification options’. This is all because of the Italy case. https://openai.com/...
-
@ruchowdh
@ruchowdh
on x
👀 this part of the OpenAI post - “policymakers and AI providers will need to ensure that AI dev and deployment is governed effectively at a global scale... This is a daunting challenge requiring both technical and institutional innovation” https://openai.com/...
-
@liron
Liron Shapira
on x
Here's OpenAI 𝗴𝗮𝘀𝗹𝗶𝗴𝗵𝘁𝗶𝗻𝗴 about AI safety today, continuing a scary pattern. It's a post called “Our approach to AI safety” that shows no acknowledgement that many observers are worried they're elevating humanity's near-term existential risk. https://openai.com/...
-
@openai
@openai
on x
We're sharing details on our approach to safety https://openai.com/...
-
@officiallogank
@officiallogank
on x
To deliver on @OpenAI's mission (AGI for humanity), we need to: - Building increasingly safe AI systems - Improve system guardrails - Protect children - Respect privacy - Improve factual accuracy And much more. Read more abut our approach to AI safety: https://openai.com/...