I have seen that text, and there were a bunch of very useful things – but I found myself disagreeing with him on at least two occasions which I will cut down to one in the interests of brevity:
The UA system doesn't so much work with the individuation of the user base, because as it stands only a single view of the users exist within the UA space.
(Edited later to cut clear insanity and say, "What the Hell was I thinking?" Always go to bed before exhaustion curdles your mind, kids.)
Well, you could think of it as an 80% solution in that it would help to eliminate noise and provide a far saner starting point than the current view of the system.
I will absolutely grant that it certainly would provide a far saner starting point than the current view. No question about that.
It still provides a top down impositional valuation rather than asking me what I like and building something bottom-up, but that's a whole separate fleet of issues.
Asking what the user likes is a great way to get a biased and incorrect answer... As a user, the one thing that I'm certain about, is that I don't really know what it is that I want from the platform with regard to finding other people's work. I know what I want myself to create, but I don't know what I want to find of others.
But making up what the user likes and pretending it's true is also a great way to get a biased and incorrect answer.
The only way to get an idea of what the user actually wants is to pay attention to what the user actually says he wants.
The alternative is, philosophically, to fail to care what the individual wants and only care about what we want to give them. That's not a system I want to be a part of. That's the exact opposite of a system that I want to be a part of.
Part of the problem is that we can never know all of what is being created, all we can do is provide ourselves the best opportunity to be exposed to things that we are likely to want. That's the key problem. We need to maximize that probability. We can't make it perfect, but we can make it better.
Think of it this way... man I wish I had a link, I'll just have to paraphrase.
Question: How do you define pornograph?
Answer: I don't know, but I'll know it when I see it!
I was just cautioning that the user may think they like something or that they ought to like something based on some idea of their self image or whatever, but it might be more useful to put stuff in front of the user that you think they'll want, and they can choose to axe it and you can make a better guess next time.
Meanwhile, the user is still free to peruse content to locate stuff they want without forcing them to surrender a preference a priori.
The downside is the user may not get a chance to say they like a thing because it was never put in front of them, especially if the raw signal is such a firehose that they drown whenever they go off on their own.
Part of the problem is that we can never know all of what is being created, all we can do is provide ourselves the best opportunity to be exposed to things that we are likely to want.
And I think we're now starting to converge on the same idea.
I have seen that text, and there were a bunch of very useful things – but I found myself disagreeing with him on at least two occasions which I will cut down to one in the interests of brevity:
(Edited later to cut clear insanity and say, "What the Hell was I thinking?" Always go to bed before exhaustion curdles your mind, kids.)
Well, you could think of it as an 80% solution in that it would help to eliminate noise and provide a far saner starting point than the current view of the system.
I will absolutely grant that it certainly would provide a far saner starting point than the current view. No question about that.
It still provides a top down impositional valuation rather than asking me what I like and building something bottom-up, but that's a whole separate fleet of issues.
Asking what the user likes is a great way to get a biased and incorrect answer... As a user, the one thing that I'm certain about, is that I don't really know what it is that I want from the platform with regard to finding other people's work. I know what I want myself to create, but I don't know what I want to find of others.
But making up what the user likes and pretending it's true is also a great way to get a biased and incorrect answer.
The only way to get an idea of what the user actually wants is to pay attention to what the user actually says he wants.
The alternative is, philosophically, to fail to care what the individual wants and only care about what we want to give them. That's not a system I want to be a part of. That's the exact opposite of a system that I want to be a part of.
Part of the problem is that we can never know all of what is being created, all we can do is provide ourselves the best opportunity to be exposed to things that we are likely to want. That's the key problem. We need to maximize that probability. We can't make it perfect, but we can make it better.
If we can make it continuously better – we win.
Think of it this way... man I wish I had a link, I'll just have to paraphrase.
Question: How do you define pornograph?
Answer: I don't know, but I'll know it when I see it!
I was just cautioning that the user may think they like something or that they ought to like something based on some idea of their self image or whatever, but it might be more useful to put stuff in front of the user that you think they'll want, and they can choose to axe it and you can make a better guess next time.
Meanwhile, the user is still free to peruse content to locate stuff they want without forcing them to surrender a preference a priori.
The downside is the user may not get a chance to say they like a thing because it was never put in front of them, especially if the raw signal is such a firehose that they drown whenever they go off on their own.
And I think we're now starting to converge on the same idea.