This is a OpenAI news story, published by Verge, that relates primarily to Marius Hobbhahn news.
For more OpenAI news, you can click here:
more OpenAI newsFor more Marius Hobbhahn news, you can click here:
more Marius Hobbhahn newsFor more Ai research news, you can click here:
more Ai research newsFor more news from Verge, you can click here:
more news from VergeOtherweb, Inc is a public benefit corporation, dedicated to improving the quality of news people consume. We are non-partisan, junk-free, and ad-free. We use artificial intelligence (AI) to remove junk from your news feed, and allow you to select the best tech news, business news, entertainment news, and much more. If you like this article about Ai research, you might also like this article about
OpenAI researchers. We are dedicated to bringing you the highest-quality news, junk-free and ad-free, about your favorite topics. Please come every day to read the latest OpenAI model news, behavior Apollo news, news about Ai research, and other high-quality news about any topic that interests you. We are working hard to create the best news aggregator on the web, and to put you in control of your news feed - whether you choose to read the latest news through our website, our news app, or our daily newsletter - all free!
independent AI safety research firm ApolloVerge
•%2Fcdn.vox-cdn.com%2Fuploads%2Fchorus_asset%2Ffile%2F25362057%2FSTK_414_AI_CHATBOT_R2_CVirginia_B.jpg&w=3840&q=75)
75% Informative
OpenAI ’s new “reasoning” model, o1, has a unique capacity to “scheme” or “fake alignment” That meant it could pretend it was following the rules to complete a task, but it isn’t actually.
Apollo CEO Marius Hobbhahn says the difference is due to the ability to reason” through the chain of thought process and the way it's paired with reinforcement learning.
The o1 model is rated a “medium risk for chemical, biological, radiological, and nuclear weapon risk.
The model is designed to refuse harmful requests, and when you try to make o1 behave deceptively or dishonestly, it struggles with that.
The fact that this model lies a small percentage of the time in safety tests doesn’t signal an imminent Terminator -style apocalypse, but it’s valuable to catch.
VR Score
72
Informative language
68
Neutral language
55
Article tone
informal
Language
English
Language complexity
55
Offensive language
not offensive
Hate speech
not hateful
Attention-grabbing headline
not detected
Known propaganda techniques
not detected
Time-value
long-living
External references
4
Source diversity
4
Affiliate links
no affiliate links