First of all not only does Odlyzko not discuss public vs private, he also doesn't really discuss commercial vs non-commercial. For instance in the point to point category he includes phones. But telephone companies generate substantial fractions (the majority?) of their revenue from calls involving companies, most of whose traffic is likely to be commercial in nature. What he really discusses is whether broadcast commercial content can pay for a peer to peer network. (No.)

Furthermore the public vs private distinction and the broadcast vs peer to peer one both seem to me to have a hazy boundary that is fuzzy and fading fast. At one point to engage in public broadcasting of information was fairly difficult. It needed substantial facilities for production and distribution. This barrier to entry created businesses whose natural job was to be a bottle-neck between would-be content creators and potential content consumers. Both natural bottlenecks are fading because of improvements in technology, but organizations of those companies are attempting to create new barriers to protect their businesses. (Think RIAA and MPAA.)

However there is a growing group of semi-public, semi-broadcast models. Part of it is in online forums like this one. Publically mirrored email lists. Blogging. Online comic strips. And so on and so forth.

These are publically accessible. But the vast majority of the content produced is consumed by small groups of people who mostly know each other. This is public in the same way that my conversation with friends on a street is public - someone walking behind us can listen in. In reality it is basically peer (me) to peer (small circle of friends). However straightforward power laws indicate that you will see an entire spectrum from personal discussion up to circulation numbers that professional columnists can respect. With no division between them.

Of course this is communication without commercial aspirations behind it. From Odlyzko's point of view, it makes broadcast payments even less able to pay for content. (A trend which doesn't make companies based on generating profits from broadcast bottlenecks very happy...) But I find it an interesting blurring of boundaries as what had been a private activity (talking among friends) shades into clearly public consequences.

All true - what interesting times we live in!

People will carry on making money from content production - quite probably in new and interesting ways (witness the way PayPal and Amazon's donation system have become a poor-mans micro-payment system). Some middle men will bite the dust, but I won't shed too many tears about that ;-)

Things like blogs, Wiki's, FOAF, etc. are fascinating teasers of the way that communication and information sharing strategies are changing. Especially now that the net is becoming ubiquitous in many areas. It's odd to think that in less than ten years everybody in my social circle has gone online - techie or not. That's going to change things in interesting ways.

Haven't we gone a full circle? The polarized world of mass media is gone we are back in the situation that everybody has the abillity to speek.

The natural state is that everybody speaks and everybody listens. The problem is that we can't listen to everything. There must be mechanisms for filtering the information. One of them can be a marking by the creator indicating the intended audience. This mechanism is more fine grained then others used since the adressed audience might be a single person. And there are messages that are interesting only to a single person.

The paper doesn't compare public and private communication - it compares commercially produced content and everything else.

I must admit I don't have any experiences in developing online communities (other then participating in a few of them). So all of my arguments are just speculations.

Perhaps public/private is wrong distinction, perhaps it indeed should all be public in the sense that it would be available to everybody, but just some part of the information, addressed individually, would be by default filtered out by all others then the individuall (althogh by a change in the settings they could view it).

What they compare is mass media (which is clearly public) and communication i.e. phone and post (which I would describe as nearly entirely private).

I think you might be surprised. The postal service transports a lot of things that don't fit under "private" communications (e.g. newspapers and magazines). As for the phone system - I hear that this "World Wide Web" thingy is taking off and using up quite a bit of bandwidth :-)

Bulk wise, public wins. Profit wise, private wins.

Perhaps public/private is wrong distinction

Possibly. For example, look at the blogsphere where you're getting conversations built around authors rather than topics. It's all interesting stuff.

The newspaper/magazines part was specially treated in the article and their argumentation about subsidizing should be applicable here as well. They present much data from time that WWW had not too much impact on the economy - the most detailed data table covers two years 1994 and 1997, at least at 1994 there were not that much WWW trafic.

Even if the proportion is in favor of public it is not so much as here. I can't say how much private messages other users receive, but mine sum up to much less than a 1% of the information I read from PM.

Following tilly the distinction between public and private becomes more fuzzy with usage of internet - there are forums in the whole spectrum between private and public. But that's not all, there are fora that are more private in one aspect of being private while in some other they are more public. So the two value system first becomes a linear system, then it becomes a multidimentional space. I feel we need some new distinctions to describe and analyze that. A good example is the chatter box - it is public in the sense that it is evailable to everyone, while it is a private because at each period of time the message is addressed to a closed audience.