And what do I learn from this?

I mean, from reading all that slush. As a writer.

Well, mostly the do nots. Lots and lots of the *don't*s, like:

• Read the guidelines: I know this is the most-repeated advice, and I'm afraid this will not change. Especially if I see that over a third of what I receive every time is either not an RTF file, or not single-spaced (I realise that 'standard manuscript format' is blasted into every aspiring writer's brains. Still. In the 21st Century, where *all* the typesetting is done on a computer. Old habits die hard, right?). Unfortunately, this can't be repeated enough, and I was especially amused by Doug Cohen's post about getting a submission from a fifth grader who did it all right. So an eleven-year-old can, while loads of adults can't.
• Be patient. No: be very patient: I know -- I've learned the hard way -- how busy an editor is. I try to get things done as soon as I can, but never quite as fast as I estimated, or hoped. Something always gets in between, and as long as this gig doesn't pay the bills (and to the best of my knowledge only three people in genre magazine publishing get paid full time to do this: John Joseph Adams [F&SF], Brian Bienowski [Asimov's], and Trevor Quachri [Analog]. Feel free to correct me), it's never going to be my top priority. High, very high, but never number one. So responses might be delayed, and editors will always like patient submitters better than the ones who query too soon. I know we shouldn't, and for some it may not make a difference, but we are human, after all, with faults and all.
• Check your market: read at least a sample issue. It saves everybody's time -- including your own -- if you target your stories right. I suspect it's one of the reasons quite a few genre publications still don't use email submissions: the temptation to submit blindly is great (and it happens in droves).
• Conversely: don't try to write specifically for a market: write what you must write, to the best of your abilities, with passion, sweat and knowledge. Then let the evil editor work it out. Thus: write it first, and worry where to send it to later.
  
• Send your best, and only your very best: I've seen the email slushpile for Interzone grow from 300 to 500, and I know that a good story just isn't good enough: it must be superb and spectacular to make the cut. This is true for every market that gets 500 or more submissions per month, such as F&SF, Asimov's, Analog, Intergalactic Medicine Show, Baen's Universe, and others.
  

There are more, but these come to mind first.

But, I hear some of you say, reading those very good stories (rare as they are): doesn't that show you how to do it right?

Yeah, but it doesn't show you how to write a superb one.

Writing a story, a great one, is still, basically, more magic than technology. If writing a great story was a kind of trick, or something that could be duplicated, then not only would it lose all its lustre, but all slush readers would be doing it. As you already noted, we don't.

This is, I think, mainly because there is something unique, something truly personal about great stories that can't be duplicated. At least not today: maybe sometime Artificial Intelligences will develop that are multitudes more intelligent than humans, and for them writing a superb story will be a piece of cake. However, if these will come into existence (not saying that they *will*: that's a completely different discussion), then I suspect they'll be as interested in writing a story for humans as we are in -- say -- developing a solution that will make bacteria feel fantastic.

So, if anything, reading slush made me realise that I must develop my unique voice as far as I can, to make me stand out from the pack. And not exactly how to do it, but more how not to do it.