While I agree AI lowers the barrier to entry, current generative AI is just an amalgamation of existing art. It doesn’t create original works, regardless of how good a Prompt Writer you are. LLM’s have been trained on millions of pieces of literature, music, videos and pieces of art. When you say it won’t be homogeneous - that’s exactly what AI does. It takes what we have already produced and parrots it back at us. Generative AI cannot, in its current form create originality.
Absolutely true. This vision is probably about 10 years out.
With AI-generated tracks actually ranking on charts, we are not too far from the day when at least music generated specifically for us streams exactly the way we want it.
I’ll fire up “Chris’s station", and it will seamlessly create music to match my moods, location, and situation. It will know I’m at my writers’ meetup and generate writing music. It will know when I’m driving and create driving music. It will know when I’m reading and create reading music.
I see what you are saying. And Spotify is already doing that with existing music. The problem is, even with purely AI generated music, it still needs a prompt writer and it still only repeats back what it has learned. The personalization piece exists in Ad retargeting and why that widget you spoke about with a friend suddenly shows up in your search results and your Social Media feed (coming soon to substack) Location sensitive ads have existed for around 10 yrs and they are getting more sophisticated. Bringing all this together is certainly on the cards and I think you’re underselling your argument at 10 yrs. I’d say maybe 5 at the outside.
Thirty years in sales taught me audiences aren’t real. There’s only the person in front of you, and content is finally catching up. The problem was never ideas, it was getting them out of your head.
While I agree AI lowers the barrier to entry, current generative AI is just an amalgamation of existing art. It doesn’t create original works, regardless of how good a Prompt Writer you are. LLM’s have been trained on millions of pieces of literature, music, videos and pieces of art. When you say it won’t be homogeneous - that’s exactly what AI does. It takes what we have already produced and parrots it back at us. Generative AI cannot, in its current form create originality.
Absolutely true. This vision is probably about 10 years out.
With AI-generated tracks actually ranking on charts, we are not too far from the day when at least music generated specifically for us streams exactly the way we want it.
I’ll fire up “Chris’s station", and it will seamlessly create music to match my moods, location, and situation. It will know I’m at my writers’ meetup and generate writing music. It will know when I’m driving and create driving music. It will know when I’m reading and create reading music.
Ain’t that the truth!
I see what you are saying. And Spotify is already doing that with existing music. The problem is, even with purely AI generated music, it still needs a prompt writer and it still only repeats back what it has learned. The personalization piece exists in Ad retargeting and why that widget you spoke about with a friend suddenly shows up in your search results and your Social Media feed (coming soon to substack) Location sensitive ads have existed for around 10 yrs and they are getting more sophisticated. Bringing all this together is certainly on the cards and I think you’re underselling your argument at 10 yrs. I’d say maybe 5 at the outside.
It is probably closer to 5. We futurists can never get the timing right. explains why we are poor.
Thirty years in sales taught me audiences aren’t real. There’s only the person in front of you, and content is finally catching up. The problem was never ideas, it was getting them out of your head.