The hungry ocean

A month in the life of a swordboat captain, a glimpse into the mindset needed from a leader.

permalink

Bullfinch on Pythagoras

“The first lesson Pythagoras's disciples learned was silence. For a time they were required to be only hearers. "He (Pythagoras) said so" (Ipse dixit) was to be held by them to be sufficient without proof. It was only advanced pupils, after years of patient submission, who were allowed to ask questions and to state objections."

— Thomas Bullfinch, The age of fable (1855). Do we paradoxically have less critical thought today because we are free to ask questions from day one?

permalink

Rethinking America's prisons is cheaper than rethinking incarceration

We punish people with architecture. The building is the method.

New prison construction is parceled out to a handful of large and anonymous firms, discouraging innovation.

American prisons are built in the countryside. Rural prisons need no public face. It need articulate no sense of communal pride or civic justice.

Convicts tend to come from cities; guards do not. Culture clashes inevitably arise.

Convicts tend to come from cities; their families often can't afford to travel to visit. Rehabilitation becomes difficult.

Guards serve "lifetime sentences 8 hours at a time." Guards and prisoners often want the same improvements.

Prisons that look pleasing suffer less vandalism.

— Jim Lewis

permalink

* *

“The publishing industry turns out to be a lousy place to keep stuff published."
— Jon Udell

permalink

Coping with friendfeed's time-ordered torrent

I've been increasingly using friendfeed, and the number of my subscriptions has trended steadily down. The reason: ordering by time forces me to be strict in who I let in.

Every new watering hole for conversation — facebook, google reader, twitter, friendfeed — orders my reading by time to provide immediacy. Ordering by time renders it susceptible to frequent posters. The minute I subscribe to one, the diversity of my reading goes down. Other voices become hard to find. My response to this: never subscribe to frequent posters.

But this is a blunt heuristic. High-volume sources often have great posts. As the need for other views grows, I find coping mechanisms. Sometimes I give up and leave. Sometimes I build a replacement, and sometimes I find others have done so. Once I can get around pure time ordering, I heave a sigh of relief and subscribe to the people I want to without feeling constrained by volume.

So, friendfeed, please help me navigate this stage of my reading. Find ways to keep my reading diverse, even if I subscribe to Robert Scoble.

permalink

Designing for serendipity

Lu Liu: How can we use serendipity to get out of homophily traps? I have a serendipity friend list. But if I can define my serendipity friends, then I guess they are not really serendipity friends by my expectation. In reality, I seldom read articles from that list.

Me: Yes, a serendipity list can't come from yourself. It must be an external recommendation. Automated since that's my bias :)

Time is key. What is serendipitous today is not so tomorrow. That makes it harder to 'define'. In practice, I suspect we must evaluate it like we evaluate porn: not by defining it but by categorizing examples.

Perhaps it can't be a list either, just one recommendation, with pride of place. I find I require time to appreciate something outside of my comfort zone.

Since it must take prime real estate it must be high-confidence. If nothing is good enough today, show me nothing.

Finally, it mustn't nag. Make it easy to dismiss, use the dismissal as a signal to learn from.

comments

  
Anonymous, 2009-06-01: I agree with you on the categorization part. But you need to have data to do the categorization. Where do you get those data in the first place?

permalink

The dirty secret of social media: quantity trumps quality

That's an understatement.

The one thing common to feedreaders, Twitter and Friendfeed: they order articles by time, and so incent people to post often. Keep posting, and your stuff is more likely to show up on people's screens. They become more likely to click on it, retweet it, share it on google reader, like it on friendfeed.

It can't be complete junk, of course; if you spam people they'll unsubscribe. But keep posting mediocre stuff and they won't.

In fact, people grow more tolerant of mediocrity in their feedreaders. Subscriptions have a way of getting out of control. Past half a dozen sources people lose track of what they're reading, and of who they've subscribed to. You can slip lots of crap by their eyeballs before they take the time to reorganize.

Some of your readers will give up on the medium for a time. They'll jump to the next great thing, somewhere along the facebook-twitter-friendfeed trajectory, and they'll find it works so much better! They'll think it's because of some shiny new feature in the new tool. They'll never realize it's just that they're subscribed to less crap. So they'll start subscribing to crap again, and the cycle will repeat.

A cynical strategy to game this world: write one smashing post every week or two, use it to get new readers. Interleave the smashing posts with hundreds of short, simple, unique pieces. They will keep you in the eyes of your readers once they've subscribed.

Even if you aren't this cynical, these are powerful and subtle forces. Most of us aren't pushing a brand or an agenda, and may not think we care much about clicks and links and shares. But we respond to social feedback. If more frequent posts yield more feedback, we post more frequently. It's easy to see the benefits, harder to see the ill-effects.

And ill-effects there are. When our reading is sorted by time, nobody reads. A conversational medium requires that its participants be good listeners. The alternative is monologuing, the realm of exhibitionists, clueless advertisers, and spam. When we're incented to post more frequently, the world gradually degrades to an advertising free-for-all. A garbage-in garbage-out world; fewer people saying interesting things; less diversity in what we read and who we read.

By shirking our reading we're poisoning the well for ourselves. Things must improve.

Improvements

It's amazing how settled time-order has been in Web 2.0. Twitter is entirely realtime. Facebook has lots of filtering options; it's unclear who uses them. Friendfeed's 'best of day' view is a big improvement, but it is hamstrung in two ways. First, it isn't the default, so most people never see it. Second, it changes slowly and you can't page past the first page, so there isn't as much to read. If you read a lot of stuff you will find yourself returning to the time-ordered view.

Friendfeed has a second mechanism to manage volume: it allows you to organize your subscriptions into 'lists'. Google Reader's folders are analogous. Folders help manage the volume/value tradeoff; make sure you read the low volume feeds, then dip your toes into the torrent to taste. They're still a static organization, though. Removing a feed from a folder isn't easy to do while reading, so we put it off, and our folders pile up cruft (if we ever bother cleaning them up).

The cynical blogger gaming the system to stay in his reader's feedreaders need change nothing when feedreaders get folders. For the reader, removing a feed from a folder is as static and as hard as unsubscribing.

The good news is that things are easy to improve. When I built my own feedreader, it was amazing how quickly I preferred it to Google reader. A simple policy of fairness—I never show two stories from the same source—sufficed to compensate for nifty UI features, search, and social recommendations. One can do much more.

Update May 19: Gabor Cselle's built an iPhone app with a prioritized order for email! Leaving time order behind is not just for web 2.0 social tools.

Update Jun 6: Fred Wilson's comment points out several twitter apps to sort by popularity.

Thanks to Jonathan Nelson, Simha Sethumadhavan, Adrian Perez, and Shawn O'Connor for reading drafts of this.

comments

  
fredwilson, 2009-06-06: reverse chronological order is not enough we need better ways to sort social media if you look at my twitter link filter http://fredwilson.wsmco.com/ the "popular" sort is way more useful than the "recent" sort   
  
Kartik Agaram, 2009-06-06: I want to try that. Is it a custom app?   
  
fredwilson, 2009-06-07: well yes and no. that is custom but he created a commercial version called twitrollr http://twitrollr.com/ i wrote a post about a bunch of services like that here http://www.avc.com/a_vc/2009/05/twitter-link-page.html
  
Mark Essel, 2009-06-06: I don't mind reading the mediocre posts if they have even a kernel of great information. I need a super smart virtual assistant (that filters information like me or better) and sifts through my feeds, and suggested readings finding the most pertinent and amazing web finds and allowing them to surface before my eyes. I have a fantasy about large scale emergent node/crowd sourced (check Kevin Kelly's swarm writings), combined with customized/personalized virtual versions of our web perception (yeah A.I. like superfilters). Here's a link (it's in my far out scifi like posts): http://www.victusspiritus.com/2009/05/24/do-you-perceive-a-need-for-a-virtual-web-representation-of-yourself/ Thanks to Fred Wilson for pointing this post out, very interesting read (his web presence acts as one of my superfilters in friendfeed, along with Robert Scoble, Louis Gray, Paul Buchheit and bloggers Kevin Kelly, Seth Godin, & a few others).   
Kartik Agaram, 2013-12-20: Gregory Rader comes up with an interesting alternative model: http://onthespiral.com/?p=127718179

permalink

The future of RSS: smart search engines

Over at Mashable, Ben Parr wonders, "What is the future of RSS? Is social media a better alternative?" I believe the world is groping towards a better solution than either.

In two years I have switched several times between feedreaders like Google Reader and social news sites like Reddit, Twitter and Friendfeed. Feedreaders amplify the volume of my reading. Social news helps me find the highest quality stories. The tension between quality and quantity keeps me switching. With growing volumes of news in an increasingly-online world, feed readers and social news are each incomplete.

But I am an extreme case; most people read far less. Do power users matter? They can help highlight a trend. Every reader online wants relevant news; some just go to greater lengths than others. Perhaps the migrations of power users between feedreaders and social news sites can teach us how to serve all readers.

Feedreaders assume you want to read everything by everyone you subscribe to, and nothing by anyone else. Subscribe to too many, and those assumptions start to break down. Chronological ordering starts to suck. Frequent writers drown out the rest, regardless of who you care about. Sifting through the noise becomes a challenge.

The popularity of social news is largely explained by this challenge. 10% of the users on a social news site vote on stories, and only 1% comment. The rest of us are all using social news purely to find interesting stories, often because the feedreader didn't work out.

But using a social news site has its own drawbacks. You can't find as many stories. Stay at the front page or in a small community, and filtering works. Lower down the list, quality drops. A larger community provides faster turnover, but it's also susceptible to lowest-common-denominator effects - think pictures of lolcats or youtube videos.

Even when filtering works, you only find stories your friends find interesting. Over time, you start to ignore interests that don't overlap with your network. You risk spending time reading low-quality comments or flame wars. Echo chamber effects suppress dissenting voices, though those are often the most interesting.

In a healthy community people do their reading in private, and come together to discuss what they read.

If neither works, where does that leave us? Competing incomplete interfaces create false dichotomies. Asking readers to provide favorite sources is a good start, and so is voting on stories for your friends. Accepting recommendations from friends is a part of the puzzle, but a small part, lest you risk endlessly regurgitating each other's recommendations. There's no reason these signals can't be combined.

The essential property of both feedreaders and social news sites: they aggregate content from many sources before presenting it to the reader. We need a better aggregator, a feedreader that can handle firehoses. One that can rank stories smarter than just chronologically or alphabetically, perhaps even adapt to our changing interests.

What would such an aggregator look like? It would have scale, to discover feeds quickly, and to crawl all the feeds out there. It would have smarts, to connect you up with only the stories you find interesting, and to prioritize them. These are big changes; the new tool looks nothing like its forebears. What it resembles most is a search engine. It crawls and indexes everything it can find. Rather than responding to queries, it knows you and your tastes and alerts you to interesting pages. It can be consumed in multiple layouts wherever you go - facebook, twitter, your feedreader.

This is our vision at MeeHive.

permalink

Realism in simulation

"The goal of simulation is not to simulate reality as closely as possible. With an accurate model you cannot find commonalities."

— Tom Slee

permalink

Start a startup, become a low-income household

“You look at assets and liability to judge the strength of established companies, but when you have a startup, you look at cash flow. Low-income families are the same way. When we talk about poor people living on a dollar a day, you don't get a dollar every day. It is a lump sum, and it is very irregular. The poorest households can recall their cash flows a month ago.”

— Daryl Collins

permalink