RichardT wrote:There are three main things I’m looking forward to.
The first is software that understands complex spoken instructions - what a time-saver that will be.
The second is software that really understands music - so for example, sampled instruments that play back phrases as a real musician would, a DAW that can add decent backings automatically, or automatically correct a poor performance.
The third is direct neural stimulation of the brain by playback systems - that way people will be able to hear high quality music wherever they are.
Not wanting to rain on your parade, but those 3 are a big ask
"The first is software that understands complex spoken instructions - what a time-saver that will be"
The issue here is the ambiguity of natural language. Natural language is by definition objective. And ambiguity is part of it's attraction, What yo have to do first is decide what is 'complex' ? - any utterance that is more than a single word/phrase? But i linguistics, something that is complex might include ambiguity (Time flies like an arrow but fruitflies like a banana), adjectival phrases are difficult too - consider the word 'big' what exactly is 'big' a big mouse will never be as big as a small elephant, a lot of language is contextually dependent. Natural language is dynamic - today is nice and sunny, and in days of yore, one might say "Oh what a gay day" but in today's world one might get a queer look if you said that. I am not saying all attempts at natural language processing be abandoned, but making the point that at first glance some of these language conundrums might seem trivial, but a little further consideration will show that natural language processing is a very perplexing endeavour. That said, there have been quite impressive developments shown with products like Alexa, which essentially does complex pattern matching with ever more sophisticated machine learning. If you ever find yourself out of collar, learn computational linguistics, the demand for people with those skills increases day by day.
"The second is software that really understands music - so for example, sampled instruments that play back phrases as a real musician would, a DAW that can add decent backings automatically, or automatically correct a poor performance"
Again, there is the issue of ambiguity. Although music is an artificial language, there is no ambiguity - Middle C is Middle C, whether is is played on a penny whistle or a cathedral organ. However, there is of course a large degree of nuance, interpretation etc So how do we decide what is a poor performance. OK if something for example is written in the key of C Major and is written in such a way that the performer will not stray off the path, then how do we deal with 'accidentals', 'dissonance', 'passing notes' music that might change time signature, or speed up/slow down - Schoenberg wouldn't even get out the traps. Again I am not denying the desireability of correction software, goodness me, Melodyne is my best friend but that said, by the time I have finished faffing about, I would have been better learning to sing. Additionally thee are quite a lot of 'chorders' that look at a piece of music, examine the chord and melody and then suggest or even substitute a triad with a second inversion 7th+9th chord. But I am reminded of how much of the world's most memorable music comes about by happy accident, if music were formulaic, then it might become predictable and mundane. That said, yes there are times when we want the predictability, the 'drop' in EDM, a catchy chorus etc
"The third is direct neural stimulation of the brain by playback systems - that way people will be able to hear high quality music wherever they are"
Don't we have that already? The neural stimulation of the brain comes by way of the ears, doesn't it? Or do you mean, one hears some music, played on whatever machine, but the brain enhances the audio to the extent that even if listening on a cheap pair of cans/speakers, the brain emulates the best hifi ever. I am minded to think that will come eventually. After all, any experience we have comes by way of the 5 senses. So in theory, if we could artificially stimulate those parts of the brain that process the data delivered by each of those senses, then for example we could sit on the sofa, close our eyes and for all intents and purpose feel we are loafing about on some exotic beach, sipping at a rum baba while serenaded by the sensuous sound of a distant samba carried on the gentle breeze scented with the fragrance of the frangipanni. After all, when we are dreaming, we do experience the most bizarre and fabulous events, so it is clear the brain can 'take us where ever' so it is not beyond all reason that at some point in the future these things will be possible. If I am about then I want some software that will turn me (in a musical sense) into someone as prolific as Bach, catchy as Mozart, quirky as the Bonzo Dog DooDah Band, cool as the Rolling Stones, barmy as Rip Rig and Panic, as revered as Pink Floyd, poetic as Bob Dylan, adored as Take That and as influential as Jimi Hendrix and....well that'll do for now!
