In the production of video games and animated movies, directors and animators are constantly fine-tuning how a character's movements should appear. In many cases, directors require subtle changes, such as making a character appear a little more restrained or a bit more expressive, and communicate subjectively how much change is needed.

When a conversational AI character in an immersive environment says something false, how should the system alert the user? Researchers have now explored this question in a proof-of-concept study, which tested three techniques for alerting users to hallucinated statements from "embodied" AI agents in virtual reality (VR) systems. The study is published on the arXiv preprint server.

Recent safety tests have shown advanced AI systems making misleading statements, concealing information or attempting to prevent their own shutdown. Such findings regularly generate headlines. But what do they actually mean? Do they really point to a form of deception or self-preservation in AI systems? Or are humans too quick to interpret the behavior of language models through a human lens?

Dynamic Infrastructure, the engineering AI platform for critical civil assets, today confirmed an ongoing research collaboration with the University of Bath's Department of Architecture and Civil Engineering. Working with Thomas Kjeldsen, professor of hydrology and water engineering, the project is being carried out as part of Arline Osorio Moreno's MSc research and explores a practical question with major implications for infrastructure AI: how well different AI approaches hold up on a real engineering task, especially when applied to assets and sites they have never seen before.

In-home assistive robots could clean and organize a home while the owner is away, but without enhanced memory capabilities, they could cause more problems than they solve. They could, for example, cause a major inconvenience if they move important objects like wallets and keys but are unable to tell the owner where they put them.