What I found Interesting
- They wanted to investigate if AI would be able to replace some of the existing staff in videogame development or if it could lead to more expansive options in games that have multiple options of conversation with non-playable characters
- When they spoke to a games designer who had planned to use AI to develop a new game on his own, it was interesting that he described his experience with AI as frustrating and not as helpful as he thought it would be
- It lacked creativity and a cohesiveness of approach which is what he eventually realised he needed from a creative assistant
- It could generate ideas in a disconnected way, but not ‘understand’ the wider aims of the project – even though the game was a relatively simple concept
- When they spoke to game developers about using large language models to create more lifelife, richer non-playable characters I was surprised to hear that the industry considers this kind of approach risky. The reason they gave made sense – that game players often try to break systems or test their limits, and having NPCs who can ‘think and speak’ for themselves, means that they may say or do things that might be offensive, immoral, or potentially harmful and no games company would want to invest in this kind of risk
- My Conclusions
- Generative AI in this context is not good enough to do much on it’s own and is also considered to be a liability. In this sense – it is probably not going to replace people in the process
- Where it might be useful is in generating ideas or content that people can work with and this is probably true of many of its applications in industry at the moment.
Reference
BBC. (2025). The Artificial Human – Can AI Level Up Video Games? – BBC Sounds. [online] Available at: https://www.bbc.co.uk/sounds/play/m0026vvz.
