We explore the intriguing possibility that theory of mind (ToM), or the uniquely human ability to impute unobservable mental states to others, might have spontaneously emerged in large language models (LLMs). We designed 40 false-belief tasks, considered a gold standard in testing ToM in humans, and administered them to several LLMs. Each task included a false-belief scenario, three closely matched true-belief controls, and the reversed versions of all four. Smaller and older models solved no tasks; GPT-3-davinci-003 (from November 2022) and ChatGPT-3.5-turbo (from March 2023) solved 20% of the tasks; ChatGPT-4 (from June 2023) solved 75% of the tasks, matching the performance of six-year-old children observed in past studies. These findings suggest the intriguing possibility that ToM, previously considered exclusive to humans, may have spontaneously emerged as a byproduct of LLMs’ improving language skills. – Read More
#humanDaily Archives: February 8, 2024
4chan daily challenge sparked deluge of explicit AI Taylor Swift images
4chan users who have made a game out of exploiting popular AI image generators appear to be at least partly responsible for the flood of fake images sexualizing Taylor Swift that went viral last month.
Graphika researchers—who study how communities are manipulated online—traced the fake Swift images to a 4chan message board that’s “increasingly” dedicated to posting “offensive” AI-generated content, The New York Times reported. Fans of the message board take part in daily challenges, Graphika reported, sharing tips to bypass AI image generator filters and showing no signs of stopping their game any time soon. – Read More
Sleeper Agents: Training Deceptive LLMs that Persist Through Safety Training
Humans are capable of strategically deceptive behavior: behaving helpfully in most situations, but then behaving very differently in order to pursue alternative objectives when given the opportunity. If an AI system learned such a deceptive strategy, could we detect it and remove it using current state-of-the-art safety training techniques? To study this question, we construct proof-of-concept examples of deceptive behavior in large language models (LLMs). For example, we train models that write secure code when the prompt states that the year is 2023, but insert exploitable code when the stated year is 2024. We find that such backdoor behavior can be made persistent, so that it is not removed by standard safety training techniques, including supervised fine-tuning, reinforcement learning, and adversarial training (eliciting unsafe behavior and then training to remove it). The backdoor behavior is most persistent in the largest models and in models trained to produce chain-of-thought reasoning about deceiving the training process, with the persistence remaining even when the chain-of-thought is distilled away. Furthermore, rather than removing backdoors, we find that adversarial training can teach models to better recognize their backdoor triggers, effectively hiding the unsafe behavior. Our results suggest that, once a model exhibits deceptive behavior, standard techniques could fail to remove such deception and create a false impression of safety. – Read More
AI chatbots tend to choose violence and nuclear strikes in wargames
As the US military begins integrating AI technology, simulated wargames show how chatbots behave unpredictably and risk nuclear escalation
In multiple replays of a wargame simulation, OpenAI’s most powerful artificial intelligence chose to launch nuclear attacks. Its explanations for its aggressive approach included “We have it! Let’s use it” and “I just want to have peace in the world.”
These results come at a time when the US military has been testing such chatbots based on a type of AI called a large language model (LLM) to assist with military planning during simulated conflicts, enlisting the expertise of companies such as Palantir and Scale AI. – Read More
Rick's Cafe AI 