I've been experimenting with machine learning models for voice synthesis recently and have made a demo you might find amusing if you know who Jordan Peterson is. You can check it out at www.notjordanpeterson.com. I'd love to hear your feedback!