Sorry, “”Midtown West””

Google Maps zoomable map I found: https://www.google.com/maps/d/u/0/viewer?hl=en&ie=UTF8&msa=0...

I wonder why it seems to circumvent Hells Kitchen?

This sort of tech is also useful in that situation since it can better understand and deliver vocal nuances (e.g., emphasis/tone that delivers meaning)

I'm around the age these guys were during this story. I feel the exact opposite way. I spent middle/high school feeling similarly, only pining for the 2000s ("wow, with smartphones and the internet the industry was wide open with opportunity, anything was possible. Now it seems like everything's been done and giants rule the world"). However, the GenAI boom completely changed my mind. I feel like we're the most lucky of all the generations of engineers so far considering how many crazy things are now possible with just a few determined individuals.

In addition to what Centigonal said, even if the autoencoder was trained on only speech data, an all zero vector is probably just out of distribution (decoder has never seen it before) and causes weird sounds. However, given the hallucinations we're seeing, the AE has (maybe unintentionally) likely seen a bunch of non-speech data like music and sound effects too.