Voice actually has a listening problem because it just reads the transcript. And when you fundamentally evaluate voice systems based on text, you cannot perform.
Why listen
It goes beyond the title with direct discussion of right, like, voice, including: And when you fundamentally evaluate voice systems based on text, you cannot perform.
Key takeaways
01Voice actually has a listening problem because it just reads the transcript
02And when you fundamentally evaluate voice systems based on text, you cannot perform
03Fundamentally, we just do not believe voice systems should be evaluated to start based on text
Best for
listeners looking for a practical AI episode debrief