Giving AI chatbots human feedback on their responses seems to make them better at giving convincing, but wrong, answers. The raw output of large language models (LLMs), which power chatbots like ChatGPT, often contains biased, harmful or irrelevant information, and their style of interaction can seem unnatural to humans. To get around this, developers often get people to evaluate a model's responses and then fine-tune it based on this feedback.

Humans have always been trying to enhance machines and give them artificial intelligence so that they can exhibit features similar to humans and …

Twitter is proposing a handful of new features designed to help its users spot "synthetic" or "manipulated" media, including deepfake videos. The social networking giant last month announced plans to implement a new policy around media assets that have been altered to mislead the public. Today heralds Twitter's first draft proposal, alongside a public consultation period, as it works to refine the rules and how they will be enforced. "When you come to Twitter to see what's happening in the world, we want you to have context about the content you're seeing and engaging with," said Twitter VP of trust and safety Del Harvey in a blog post. "Deliberate attempts to mislead or confuse people through manipulated media undermine the integrity of the conversation."