OpenAI’s voice cloning AI model only needs a 15-second sample to work

https://www.theverge.com/2024/3/29/24115701/openai-voice-generation-ai-model

No-one blocked ads anymore. They were so much better than the generic stream of franchise movies and series broadcasters were still trying to push. Watercooler conversations about the latest release had been replaced by private moments, content relationships. These weren’t ads, they were time machines. Since she’d uploaded the fragment she found herself reaching for her screen throughout her day. Sometimes it was times of stress when she needed to hear the voice, sometimes it was just a passing moment that she wanted to share with him. Even if all he was saying was about toothpaste, it was still him.