From a technical standpoint, voice generation is already solved (recall Google's recent Omni releases). However, the UX (user experience) remains disastrous: bots don't understand intonation, sarcasm, and don't know when to "shut up" when a person pauses to take a breath. FinVolution, which actively uses AI in credit scoring and customer service, is willing to pay developers to create algorithms capable of reading non-verbal context. Whoever solves the "interrupting" problem will definitively bury the traditional call center industry, replacing human operators with absolutely lifelike avatars.
Source: FinVolution Group / TechNode
Voice AIFinVolutionUXFinTechB2B