EA Sports is integrating artificial intelligence into the commentary system for NHL 27, a shift that changes how broadcast-style audio is produced for the upcoming sports title. The use of AI commentary in NHL 27 impacts player immersion and highlights the evolving role of human voice actors in sports games. The studio is using AI to generate specific commentary lines, which reduces the manual workload for the talent while aiming to maintain broadcast quality.
Commentator John Bucigross accepts AI-generated lines to reduce tedious recording tasks
The game features a new broadcast team consisting of John Bucigross and Darren Pang, who serve as the primary voices for the simulation. Bucigross has been working directly with the development team to test and refine the AI-generated audio. He noted that the technology has reached a point where the output is convincing enough to warrant its use in the final product. Bucigross's involvement in correcting AI errors suggests a hybrid approach where technology supports rather than fully replaces human commentary.
Bucigross described the initial results as surprisingly accurate, noting that the AI captured his vocal style effectively. He stated, "But they’re also using some AI stuff now. They played me some stuff back in my voice and it’s like ‘wow, that’s pretty good’. You know, if it means less work and the same check, I’ll take it." Bucigross's willingness to use AI for less demanding lines reflects a broader industry shift toward using automation for repetitive recording work.
The integration is not without flaws, as the AI occasionally misidentifies specific terminology used in hockey broadcasts. Bucigross pointed out errors such as the AI referring to stadium seating as "chairs" instead of the correct term "seats." He explained, "Sometimes AI will get something wrong… And so they were playing me back some of my AI voice and it’s like: ‘Score! And the crowd gets out of their chairs!’ And I was like ‘no, they’re not chairs, I know they look like a chair but we call them seats’." The need to correct AI mistakes, such as terminology errors, shows that human oversight remains necessary to maintain broadcast quality.
Source: VideoGameChronicles
