AI helps deliver JFK’s words from beyond the grave…

On a fateful day in November of 1963, JFK never got to make his “Trade Mart” speech in Dallas. But thanks to the UK’s The Times and friends, we now have a glimpse at what that speech would’ve sounded like that day. Using the speech’s text and AI, The Times:

“Collected 831 analog recordings of the president’s previous speeches and interviews, removing noise and crosstalk through audio processing as well as using spectrum analysis tools to enhance the acoustic environment. Each audio file was then transferred into the AI system, where they used methods such as deep learning to understand the president’s unique tone and quirks expressed in his speech. In the end, the sound engineers took 116,777 sound units from the 831 clips to create the final audio.”

Why It’s Hot:

It seems we’re creating a world where anyone could be imitated scientifically. While in an instance like this, it’s great – to hear JFK’s words spoken, especially the sentiment in the clip above, was a joy for someone who cares about history and this country, especially given its current climate. But what if the likeness wasn’t recreated to deliver a speech written by him during his time, but rather something he never actually said or intended to say? Brings a whole new meaning to “fake news”.

[Listen to the full 22 minute version straight from the Source]