This TED talk has a great example where the same sound sounds different to you based on what you expect to hear, if you need to prove this to yourself [1]. His thesis is also that the brain is a prediction machine.
From memory, part of the trial was the defence taking a song, suggesting what words the court would hear in it played backwards, and then playing it backwards. e.g. this clip https://youtu.be/uyIsu93zAoM?t=1123 and showing that you hear the suggested words out of the noise.
The trial was thrown out, after deciding there were words and they were subliminal, but were not deliberate. The whole situation included: "In a pre-trial motion, the judge ruled that subliminal messages were incapable of being protected speech under the First Amendment to the United States Constitution, since they were by definition not noticeable and thus could not form part of a dialogue."
Documentary - https://www.imdb.com/title/tt0104140/