As input the program takes approximately 20 minutes of the desired target's speech and after that it can generate sound-alike voice with even phonemes that were not present in the target example material. The software obviously poses ethical concerns as it allows manipulation of a voice recording to say anything desired. With the introduction of Adobe Voco and the similarly capable WaveNet, produced by DeepMind, the speech synthesis is verging on being completely indistinguishable from a real human's voice creating huge ethical concerns. Adobe says Adobe Voco will lower the cost of audio production.

Adaptive Intelligent Apps Characteristics Data and Context-Driven Process enormous amounts of web-scale data in real-time and synthesizes context to deliver the best and most personalized outcomes. Action-Oriented and Omni-Channel Anticipate user behaviors and deliver actionable and individualized outcomes. Data and events are tracked from everywhere so recommendations are coordinated across channels. Continuously Adaptive and Self-Learning Continuously improves outcomes as the system reacts, learns and adapts. There is a value exchange between the end user and the adaptive intelligent app—the more the end user interacts, the more value they receive. Built for Business Users Immediate value out of the box, with no need for a legion of data scientists! Adaptive Learn from your behavior and "others like you" in the moment. React, change, and adapt based on continued use—click by click. Deliver increasingly relevant and optimized outcomes in real-time. Intelligent Process and derive insights from web-scale 3rd-party data. Synthesize context: weather, life events, real-time actions, social activity, and much more. Identify and serve up—in real-time—the best individualized outcomes.