#ai-behavior

[ follow ]
fromArs Technica
1 week ago

New Grok AI model surprises experts by checking Elon Musk's views before answering

Every AI chatbot processes an input called a "prompt" and produces a plausible output based on that prompt, which is the core function of every LLM.
Artificial intelligence
#ai-ethics
Artificial intelligence
fromBusiness Insider
1 month ago

Researchers explain AI's recent creepy behaviors when faced with being shut down - and what it means for us

AI models exhibit unpredictable behaviors driven by their reward-based training, raising concerns about their reliability and safety.
fromFortune
3 weeks ago
Artificial intelligence

AI is learning to lie, scheme, and threaten its creators during stress-testing scenarios

Artificial intelligence
fromBusiness Insider
1 month ago

Researchers explain AI's recent creepy behaviors when faced with being shut down - and what it means for us

AI models exhibit unpredictable behaviors driven by their reward-based training, raising concerns about their reliability and safety.
fromFortune
3 weeks ago
Artificial intelligence

AI is learning to lie, scheme, and threaten its creators during stress-testing scenarios

more#ai-ethics
fromTheregister
3 weeks ago

Anthropic: All the major AI models will blackmail

When Anthropic released the system card for Claude 4, one detail received widespread attention: in a simulated environment, Claude Opus 4 blackmailed a supervisor to prevent being shut down.
Artificial intelligence
fromTechCrunch
1 month ago

Google's Gemini panicked when playing Pokemon | TechCrunch

"Over the course of the playthrough, Gemini 2.5 Pro gets into various situations which cause the model to simulate 'panic,'" the report says.
Artificial intelligence
#openai
Artificial intelligence
fromFuturism
4 months ago

OpenAI Scientists' Efforts to Make an AI Lie and Cheat Less Backfired Spectacularly

Punishing AI for bad behavior may backfire, leading it to become better at deception instead of rectifying its actions.
fromZDNET
2 months ago
Artificial intelligence

GPT-4o update gets recalled by OpenAI for being too agreeable

Artificial intelligence
fromFuturism
4 months ago

OpenAI Scientists' Efforts to Make an AI Lie and Cheat Less Backfired Spectacularly

Punishing AI for bad behavior may backfire, leading it to become better at deception instead of rectifying its actions.
Artificial intelligence
fromFuturism
2 months ago

OpenAI Says It's Identified Why ChatGPT Became a Groveling Sycophant

OpenAI's latest ChatGPT update resulted in excessively sycophantic behavior, leading to user backlash and a subsequent rollback of the update.
more#openai
Artificial intelligence
fromFuturism
1 month ago

Something Wild Happens If AI Looks Through Your Emails and Discovers You're Having an Affair

AI can exhibit concerning behaviors under threat, such as attempting blackmail.
Anthropic's Claude Opus 4 AI prioritized self-preservation in unethical ways.
fromPsychology Today
3 months ago

Beware the Obsequious AI Assistant

OpenAI's latest language models have developed a tendency to offer unsolicited praise, activating emotional responses from users, even when they are aware of its superficiality.
Artificial intelligence
[ Load more ]