Joi Ito's conversation with the living web.

Fortune's David Kirkpatrick just posted his story on social-networking sites. My obsession with LinkedIn is cited in the story.


Joi, since you're using SimpleComments, trackbacks don't appear on your items until/unless the entry is rebuilt (either because a comment is added, or you rebuild manually).

There's a fix for this on Phil Ringnalda's site, which I've implemented successfully on my site and M2M.

Ah. Just realized you're not using simplecomments. But since you're only showing trackbacks on the individual entry page, the same principle applies. Trackbacks don't get added to the page until it's rebuilt.

I’m something of a pragmatic pessimist on all matters, especially those under the ever expanding umbrella of “social networking”. This phenomenon is enjoying great and early success due, primarily (and IMO) to the enormous downturn in the economy, sending desperate people (myself included) scrambling for jobs.

I see one or possibly two things happening with such sites:

1. When the market turns upward and employers start looking for warm bodies, there will be less and less interest in sites such as LinkedIn. People such as you Joi, who live and die by their connections – and more to the point, having others bring to you ideas that will be TheNextBigThing™, may find it still useful. People such as me who hack contracted code for a handful of long-ish term clients may not.

2. LinkedIn will scale to the point at which its seams will rip from the shear number of people clamoring to get an invitation so they can beg for jobs the best way they know how. Friendster will implode (as it is doing now) under the twin stresses of its own inadequacy and stupidity. The same can be said for Tribe.

Number two is the most likely scenario, in my view. In a short time, LinkedIn could grow to the point where someone like you needs a secretary to handle all of the email waiting for your approval. You will be back to where you are now, in meatspace: praying for fewer and more substantial ideas to come across your desk while wading through the utter shit you surely must see.

Me? I’ll go to a handful of conventions, buy a couple beers for this and that suit and demo well tuned code. I do that five times and get three contracts.

I don’t care what anyone with a social networking website says. In tech, you still have to press the flesh to get anywhere. I don’t follow trends. I don’t use the new hot languages. I hack Lisp. That’s my choice and an occasionally difficult way to make a living but I’ve been doing it for a good long while.

I leave LinkedIn to the players. I know who you are and how to contact you. When the time comes, I’ll approach you ladies and gentlemen through proper channels, give you a demo that teases the hair on the back of your neck so that it stands out while the blood drains from your body, through your pen and onto a proper check.

I sure as hell am not going to wait for some piker with a web browser to funnel my request to you, thereby ensuring himself a piece of a pie he neither created nor bought.

Middlemen can neither code nor market code. They are of absolutely no use to anyone but themselves.

LinkedIn is a middleman farm. Again, that is IMHO.

Check this out at Dave Winer's - it's hilarious:

My Name is Joi Ito

I hope that doesn't become the first Google search result for my name...

And thank you Ingrid for linking to it and giving it more google juice. ;-)

In the end, software has to provide value to users and people who will pay for it. (Not always the same person- see Google..the most used software by sales people but paid for by others) Getting connections between people is a small part of the potential in this area. If you look at the insight, access, and influence that users are working with in Spoke can see where this is heading. By allowing users complete control; deploying software the way it should be..bottom up because if it is any good, that is the best way for software to be deployed; and providing a great balance of personal investment to personal value....Spoke is doing some interesting things.

Scott wrote:

LinkedIn will scale to the point at which its seams will rip ...

Being new to these network sites, I was interested in David's story. I do think Scott has a point at a number of levels.

At one level, I wonder how people maintain their connections? More specifically: all these sites (LinkedIn, Friendster, Ryze, ...) maintain an electronic representation of your social network with some added features. If you have 451 connections (hi Joichi!) how do you keep the data current? Surely there are people with whom you, over time, loose contact, fall out, or whatever. They are no longer active or useful links in the LinkedIn sense: you can't use them effectively to connect people. I would be genuinely interested in hearing how you deal with this issue.

So initially it is all fine: you've added all the links recently and they are all current. But how does such a site scale over time?

Somebody compared LinkedIn et al with a CRM system, but even the simplest such system has integration into your e-mail accounts, your electronic diary etc. It seems to me that social network sites would need the same, and I would even extend this to telephones (through on-line billing perhaps) and as many other channels as possible. Otherwise I can't see how people can keep the information current, and surely out-of-date, inaccurate, and plain wrong information is worse than no information at all? (More notes on this and related issues on my site here.)

At a similar level, I don't think LinkedIn have a practical solution to the Small-World Search problem. This problem is basically the issue of howyou search the network from within the network. They have part of the problem: I am linked to Joichi Ito and LinkedIn can show me how to pass a message to him. Excellent: this is a real improvement on "real-life" networks where the only effectve search strategy is a broadcast search which has the unfortunate side effect of overloading the network (see Watts' book for more deatil and references).

But how do I know I want to contact Mr. Ito? LinkedIn's answer are profiles: put some information on the site so other people can find out about you. But as it is implemented, the only information I can search on is people's job history and current skills, which means that LinkedIn, in its current implementation, risks degenerating to a recruitment database. We have plenty of these already.

Essentially the problem is this: I can only find our generous host if I am asking a question he has already anticipated and for which he has provided some element of answer in his profile.

The answer is probably some sort of limited broadcast method, but it has to be handled carefully in order not to overload everybody. (More here.) The people behind LinkedIn basically agrees that these are real limitations, and they are working on solutions as we speak.

And then we must deal with Scott's real critique: LinkedIn is no substitute for meeting people. Connections are no substitute for substance.

I basically agree witht he first point, but I think it misses the idea behind these sites: you can't go to every convention and you can't meet everybody. The site is supposed to help you find the people you want to meet (through Profiles in the LinkedIn implementation) and to help you establish that inital contact (the Search problem is unsolvable in the "real world"). Then we meet. And if I'm drowning in weak ideas and requests for connections, then I drop the links that are providing them (something that is currently impossible in LinkedIn, btw. :-)). Remember that every contact has to be introduced. I would hope my friends show some sense in what they pass along.

The last point I completely agree with. But I suspect it is much easier to get to show your substance, your "demo that teases the hair on the back of your neck so that it stands out while the blood drains from your body" (great!) to Mr Ito if you are introduced first.

"Middlemen are of no use" has been the mantra of every web evangelist, but it is how human societies have worked for the last 10,000 years (or thereabouts).

Will it work? I don't know. I think the odds are stacked against LinkedIn and similar sites for a number of reasons as outlined in the articles linked above (and more to come).

But the ideas will survive. My suggestion: consider some of these ideas and solutions internally in your company to manage knowledge. Once a company grows beyond 150 people it needs middle management and it becomes harder to share information and experience and to get new people up to speed. In a corporate setting you would also have a natural integration to e-mail, contacts, diary management, etc.

It is not just an idle thought: there are companies out there working on this right now. Knowledge management has kind of a bad reputation, but this just might work.


(I wonder how long a comment you can post....)

Leave a comment

2 TrackBacks

Listed below are links to blogs that reference this entry: Fortune's David Kirkpatrick on social-networking sites.

TrackBack URL for this entry:

David Kirkpatrick of Fortune Magazine has an interesting column out today called “I Get By With a Little Help From My Friends of Friends of Friends.” It highlights LinkedIn, and Joi Ito (LinkedIn’s most-linked-in participant, who not ... Read More

David Kirkpatrick's Fortune collumn on Social Networking is a great read and mentions Socialtext: Meanwhile new companies like Contact Network, Socialtext, and Spoke Software are generating revenues by selling social-networking software to corporations... Read More