I don't remember whether I ever stated it, but I place a lot of value on new ideas. In this sense of the word, Accelerando did fairly well. However, there were problems that I don't feel comfortable going into unless you have already read it. I don't like to ruin books for people, so I will just say read it with the knowledge that every section is probably around 90% as interesting as the section preceding it, but that it is worthwhile book if you enjoy being presented with new ideas.
***** SLIGHT PLOT DETAILS FOLLOW *******
As I remember, the plot was roughly divided divided into 9 sections, 3 groupings of 3 sections each. The first grouping is our immediate future, the second grouping is our protagonists daughter's story, and the third is their further descendants (and his reincarnation... twice... at least [don't ask]). As you might have guessed, the plot is a accelerating mess. Towards the end, I just didn't know what to do with it anymore. In terms of narrative, flow, character, and story, the first third is pretty good, the second is okay, and the third was frankly dull. I actually thought this was kind of funny, as I my progress in this book was slowing down as the singularity was speeding up.
But enough about the book itself, lets talk about the ideas! <Work in progress, I am filling these out as I have the time>
1) I though the scene where manx looses his goggles and literally has no idea who he is or what he is doing was very interesting.
2) Reason we don't observe, and don't deal with alien races is that singularity just means building smart matter, not conquest or anything like that.
3) Economy 2.0 will be about novelty, not power, matter, influence, or anything else.
4) Self aware contracts and corporations.
5) External brains allow concurrent though.
6) External brains allow accelerated though.
7) Sea slug is born in a digital realm, does not understand that there is actually a physical one.
8) Concept of being able to fully model a human being