A casual search for issues with LLMs:
Biases in Large Language Models: Origins, Inventory, and Discussion | Journal of Data and Information Quality
Biases in Large Language Models: Origins, Inventory, and Discussion | Journal of Data and Information Quality
Similarly to physical appearance and disability biases, religious bias can be detected more easily than other biases, including via probing techniques. Nevertheless, large language models have been found to exhibit religion bias consistently in different tasks and uses.
Large Language Models pose a risk to society and need tighter regulation, say Oxford researchers | University of Oxford
Leading experts in regulation and ethics at the Oxford Internet Institute, have identified a new type of harm created by LLMs which they believe poses long-term risks to democratic societies and needs to be addressed by creating a new legal duty for LLM providers.
www.ox.ac.uk
Associate Professor and Research Associate Dr Chris Russell, Oxford Internet Institute said: 'While LLMs are built so that using them feels like a conversation with an honest and accurate assistant, the similarity is only skin deep, and these models are not designed to give truthful or reliable answers. The apparent truthfulness of outputs is a ‘happy statistical accident’ that cannot be relied on.'
Sophisticated AI models are more likely to lie
Human feedback to AIs makes them favor providing an answer, even a wrong one, while making the answer more convincing.
arstechnica.com
When a research team led by Amrit Kirpalani, a medical educator at Western University in Ontario, Canada, evaluated ChatGPT’s performance in diagnosing medical cases back in August 2024, one of the things that surprised them was the AI’s propensity to give well-structured, eloquent but blatantly wrong answers. Now, in a study recently published in Nature, a different group of researchers tried to explain why ChatGPT and other large language models tend to do this. “To speak confidently about things we do not know is a problem of humanity in a lot of ways. And large language models are imitations of humans,” says Wout Schellaert, an AI researcher at the University of Valencia, Spain, and co-author of the paper.
The more difficult the question and the more advanced model you use, the more likely you are to get well-packaged, plausible nonsense as your answer.
Last edited: