Artificial intelligence continues to evolve and bring about breakthroughs in many areas. AI media creation is perhaps one of the most interesting areas I have taken notice of recently. It’s fascinating yet a little scary at the same time. A couple of weeks ago, Nvidia highlighted their work on expressive speech synthesis under project ‘Flowtron’. A system that synthesised an artificial voice that sounded more human than ever. With enough modelling data, AI voice and expressive art seem to be closer than ever and perhaps the next generation of social media content might feature AI powered creations.
We already see generated voice overs on platforms such as TikTok, you can pick a narration style and input whatever text. It works very well and perhaps in just a few years we will see AI speech synthesis take another leap forward. Providing expressive voices for those who cannot speak, synthetic voice acting and so much more could be possible. It’s a little scary yet exciting at the same time.