I don't think the Internet has had any detrimental effect on the quality of my ideas, but it certainly has had a detrimental effect on my ability to execute them.
When I wrote my MA dissertation, I suffered from the problem that I would write a few sentences, then check the internet or a book or paper, to make sure I wasn't asserting something false, or to see if I could find something to support my assertions or whatever.
It made for exceptionally slow progress.
To fix this, I took my laptop to a coffee shop where I would have to pay for internet access, without taking any books or papers (or only taking the specific one I knew I needed to refer to). Any time I felt I needed to look something up, I would write it in my notepad, and carry on writing.
Once I had hit my target word count for the day, I would go home and look up the things I had noted down, and make any necessary corrections or footnotes.
Basically, I physically separated the act of writing from the act of research.
I've not tried it for coding, so I'm not sure how well it would work. On the one hand, you can find example code, handy boilerplate, or whole libraries for many problems, saving you masses of time over actually coding it yourself, but on the other hand, I know I've wasted hours looking for existing solutions to problems, when it would have been faster to just do it myself.
When I wrote my MA dissertation, I suffered from the problem that I would write a few sentences, then check the internet or a book or paper, to make sure I wasn't asserting something false, or to see if I could find something to support my assertions or whatever.
It made for exceptionally slow progress.
To fix this, I took my laptop to a coffee shop where I would have to pay for internet access, without taking any books or papers (or only taking the specific one I knew I needed to refer to). Any time I felt I needed to look something up, I would write it in my notepad, and carry on writing.
Once I had hit my target word count for the day, I would go home and look up the things I had noted down, and make any necessary corrections or footnotes.
Basically, I physically separated the act of writing from the act of research.
I've not tried it for coding, so I'm not sure how well it would work. On the one hand, you can find example code, handy boilerplate, or whole libraries for many problems, saving you masses of time over actually coding it yourself, but on the other hand, I know I've wasted hours looking for existing solutions to problems, when it would have been faster to just do it myself.