I recall with fondness doing a lot of advertising on AOL in 2004-2005. AOL placed a lot of additional creative and technical restrictions on advertisers. One of the most annoying of these not just on AOL but on a variety of platforms was with respect to the version of Flash that Flash ads were published in. It caused endless strain because while everywhere else pretty much accepted Flash 6, AOL still required ads to be in Flash 5 format. Pain. Some things just weren’t possible in earlier versions, or when you had painstakingly had your designers make sure the ad was under 30kb in size, doing it in the older version now made it > 30kb etc. etc. You get the idea.

 I was reminded of this when I saw the message below on Facebook’s ad platform. This is just for the display of the dashboard graphic, mind you, that shows you clicks and fans and so on (in general I find 90% of these types of dashboards useless, but that’s a topic for a different, much longer conversation). This is certainly not requiring me to change the Flash version of all of my ads and is really a minor annoyance, but seriously, you’re making me upgrade from Flash 9.0.45 to 9.0.47? Sheesh, what’s up with that?!

interesting when you can see first hand (literally) how small things can have a big impact! I was waiting for the san francisco train this morning around 8am at fruitvale station. The richmond train arrived and picked up passengers. But then when it was set to leave, the door directly in front of where I was standing would not close. The operator kept trying to jam it closed remotely, for several minutes. They did not attempt to investigate the issue. Eventually they announced they would be taking the train out of service and that everyone had to get off!!

at this point they opened all the doors, and before any of the people in the car ahead of me could get out I reached down and saw there was a very small stone lodged in the door groove, and pulled it out!

Many of the people on this car were incredulous that the train was supposedly going out of service. Nobody had come to check on the door. In the meantime I was walking up the platform to see if I could find the operator and tell them. I realize perhaps I or someone in that car should have grabbed the operator phone and called them… But surprise surprise they closed the doors (all did) and with some people still in that car, the mostly empty train headed off to wherever!

with now resulting delays and overcrowded trains, not just hundred(s) but thousands of people inconvenienced - and who knows what else (like that gwyneth paltrow movie) - all by a small stone! (written on the bart train)

Looks like mmmzr.com which I mentioned before had to shut down (”closed for repairs” it says) since it is difficult to manage your pyramid when some of the pyramid building blocks start to want refunds and when the amounts get large.

Another interesting way to think about something like this is the Martingale betting system (basically doubling your bet when you lose) - good results possible in the short run but more of a loss avoidance strategy in the long run — but “noise”/transaction costs/rules (imposed by the casino for example - table maximums) render the strategy useless. A variant of this is used in Taleb’s Fooled by Randomness to entertainingly discuss the myth of money manager outperformance.

Interesting idea (well interesting area really) — Matt Marshall on the other Mr. Hoffman’s network/reputation startup, Rapleaf. The first Silicon Valley Hoffman is Reid, of course.

After I wrote the entry below I typed in “Starbucks Map” into Google and came across this blog entry, where someone decides to use Starbucks’ market research (partially) as a way to plan a sightseeing trip in Shanghai, China. Now imagine if everyone had this approach to sightseeing …. (!)

I haven’t blogged about this until now, but theFacebook raised $12 million dollars from Accel at what seems like an outrageous valuation. Jeff Clavier’s blog has some numbers from their VP of corp dev, Matt Cohler who is a friend and former colleague. The user numbers aside (which are quite amazing), it seems crazy to take that much VC money for a business that is already profitable. Then again perhaps not that crazy if they really did get the kind of valuation suggested by Always-On. This comes on the heels of me hearing stories about lawyers starting to not only accept but ask for (!) equity in exchange for legal services… shades of 98-99 all over again perhaps, with too much money chasing too few good deals? Sure there’s a real online advertising market out there and companies like theFacebook are making money with decent scale of users, but what kind of a return is possible even with a fabulous ad market on that kind of investment? I’d be happy to be proven wrong

Thanks to Gary Stein for posting this referring to this fascinating implementation of Google Maps along with housing listings from craigslist. Very very cool!

Mediapost’s Online Spin discussed college networking sites - and this led me to the CampusNetwork.com website. There are subsites for a number of different colleges and each one clearly has that school’s look and feel built into it. What’s surprising to me about some of these sites is how open they are and how much information there is even to people who have not joined, but then I guess that isn’t really surprising to me anymore. Check it out - you can find interest groups like “Columbia College Drinkers Against Bush” for “politically active weekend alcoholic[s]“.

I’m reading (the audiobook) of “The Wisdom of Crowds” by James Surowiecki. Quite enjoying some of his anecdotes. He also references Taleb’s Fooled by Randomness a few times. It’s pretty wide ranging, one feels like he doesn’t really drill down deep enough to really make his point. Nonetheless a good, entertaining collection of anecdotes to bolster his point. His overall thesis is decent and worth repeating here, about what makes a group collectively clever:

” ‘Wise crowds’ need (1) diversity of opinion; (2) independence of members from one another; (3) decentralization; and (4) a good method for aggregating opinions.”

Worth a read, or a listen as the case may be.

First blogged on March 21, 2003 (Jupiter analyst blogs), these were some of my themes for 2003, I think they’re even more relevant now —

Contextualization: The distribution of context. Context is becoming better distributed. Take weblogs for example ? context comes from the repeated assessment and application of subject-matter relevance to geographical space or opportunity (like a space on a website or a moment of a person?s attention). Humans are pretty good at determining context, but machine tools can help by creating a framework for categorizing and assigning attributes to objects, a framework for relevance and a means for pushing the subject matter into the geography.

Discontinuity: If access is almost- but not everywhere, there will be gaps in the network, in the ongoing access to information. On the one hand these gaps appear to be diminishing as access propagates but on the other hand the very same technologies create gaps themselves. Multitasking means people devote less attention to each individual task, and tasks are broken into smaller bits and pieces. Email and instant messaging are a form of turn-based communications (thanks to Tim Smith) that have largely replaced continuous discussion/interaction in many business and personal contexts. The low cost of sending these messages, of generating news and publishing data creates further discontinuities and noise in the system as half-baked thoughts and ideas are shared without the benefit of reflection and digestion. Sensational but false rumors propagate with amazing speed, corrections and truth not always as fast. That is not to say that this discontinuity is always negative ? as there are great potential benefits to rapid prototyping, fast failure, testing concepts out before fully investing time or effort in developing a full version of a product or service.

However, discontinuity becomes a problem when context is not maintained through a process or an endeavor. Without some way to unify disparate experiences, without some type of glue to loosely couple people, conversations or data elements, chaos rules (for example if a news story is exposed to be erroneous, should all digital instances of the story reflect the new information, reflect the correction as an appendix? is “loose coupling” even possible? are our systems designed for fuzzy connections between things?).

“I already gave my best. I have no regrets at all.” — so said Berkeley junior William Hung on American Idol and received his 15 minutes of fame. And a website to match - though very interesting, this overnight success does not have a Google pagerank to match. Interesting that many Amazon affiliate links have sprung up on the page in the last day or so… (he should have done this from the start). And of course GoDaddy very very clever gets a nice mention on the site in return for providing a dedicated web server. GoDaddy does rock, it is true. The same cannot be said for William Hung, but then you already knew that.

Good phishing example in Seth Godin’s blog where he laments how “distrust is changing our culture.” In a sense, it’s a struggle between the proliferation of information (which should lead to accountability) versus the power of anonymous suggestion. What I mean by this is that it should be fairly easy to figure out when someone (esp a public figure/politician) says something whether this thing is true or not, or is inconsistent with past statements. However, at the same time we all have to defend ourselves from anonymous persons who can easily assume the veneer of being authentic.. as in Seth’s example, can look like someone or something we know and trust. It’s the downside of the network effects of brands and of groups of trusted individuals and companies. The shortcuts we use to figure out if we should buy a new product from an existing entity are the same ones that bad actors can exploit to fool us. Unfortunately, this works in politics, as well as in online commerce…

So-called “Flash Mobs” have quickly risen to mass-media prominence this summer. Nice blog entry (bear left on on unnamed road:) about it for those who’ve been under a rock for the past few months. It provides a small answer to a part of this question: will technology bring people closer together or will it move us further apart? Flash mobs are the ultimate fake plastic community. It is merely one step above the type of community interaction most of us living in America enjoy on a daily basis with one another as we speed by one another at 60-80 miles per hour on the freeway.

For example celebrities and popular culture icons are unnaturally revered in the US. Perhaps partly because for those of us living in the ‘burbs, we have very few shared experiences, and yearn for a way to connect with fellow human beings. Email replaces the phone in many interactions today; answering a ringing phone is a commitment — it’s not smart to make a commitment when you’re not sure who’s on the other end… or what that person might need. Reality TV is part voyeurism, part community and shared experience, and vicarious celebrity-creation. The confluence of these is the start of a kind of rediscovery of community. These temporary fleeting bonds and actual shared experiences (instead of the cave shadow of popular culture) powered by technology will evolve and harden into real social and political communities, often harnessing existing bonds/affinities (cultural/racial). In the midst of it, perhaps we will rediscover who we are and what possible things we have in common…

“(They) prove people can still form ad hoc communities and make things happen that are beyond the reach of the gigantic, corrupt corporate and governmental powers that seem to dominate so much of modern life. But maybe I’m reading too much into it.” says Sean Savage [from Wired article on E-mail mobs]

Swarms, “the ant model”, behavior not controlled from a central point but forming a larger system by the interaction between agents, these are the cool topics of the 21st century, and much energy will be expended in trying to harness the power of the mob, the swarm, the network (for good, evil or both i.e. politics). I can’t wait for the combination phone-voice and phone-camera picture blogging software that must be about 6-12 months away (the pieces exist, are they being tied together yet?).

I religiously use the Google toolbar, and thus Google, to search for and research just about anything. The problem is, that’s often all I do — I often fail to consult other sources or dig any deeper. Just as I’m sure many students default to using the Internet instead of trekking down to the library if they can avoid it, the power of a decent (the best available, but still pretty inferior in the grand scheme of things) search lulls me into accepting its results as the final word on a matter.

Intuitively, “Googling” something seems like the 80 of the 80-20 rule, at high speed and low cost. And thus we have the myth of searchability: a powerful search that can deliver a wide enough range of good-enough results will sometimes have the effect of convincing people that a lack of results implies a lack of information - that the data do not exist. The limitations of the search engine become confounded with the limitations of the information. Lack of searchability = lack of content.

There are many implications that I’d love to explore with people interested in this topic - but one we’re just starting to learn about is how the (to date very successful) branding of an informational process feeds back upon the quality (and the organization itself) of that information…. search engine optimization is a fine euphemism for this, but that is only but the beginning.

I was digging through some old emails from December 1999 and came across a presentation by Mohan Sawhney where he discusses the “Myth of Substitution”. Quoted from his presentation:

- When a new technology is created, it is seen as a perfect substitute for the existing technology
- The substitution argument is based on an incomplete understanding of the new potentialities, and an extrapolation of old mental models to the new technology
- In fact, the new technology often provides a different set of benefits for a different set of end-users and a different set of end-use applications
- Therefore the new technology and the old technology often end up co-existing as complements, not as substitutes
- However, the old technology often ends up serving a different benefit space and different customers in the hybrid world

Mr. Sawhney goes to give some old and new/current (at the time) examples of the myth of substitution in action:
- VCRs vs. movie theaters
- TV vs. Radio
- Ballpoint pens vs. fountain pens
- Wireless phones vs. wired phones
- Digital watches vs. analog watches
- Solid state components vs. vacuum tubes
- Distance learning vs. universities
- PDAs vs. laptop computers
- Cyberbrokers vs. brokers
- CareerCentral.com vs. headhunters
- Etailers vs. malls
- Online advertising vs. offline advertising
- Online grocers vs. supermarkets

Some great examples - worth thinking about as we look at how technology changes the rules or more often, makes new rules!

Enjoyed Bradford DeLong’s Wired 11.05 piece about the equivalent of Moore’s Law as applied to increasing storage power…. it brings up an important point. We are at a point in time, moving forward, that any information that is captured or stored electronically (like this blog entry for example) will probably never be lost in an absolute sense. Bandwidth has not become free, but storage now effectively is. This increases the value of search technology and any other schemas or infrastructure that can add structure to data or help us organize and retrieve information.

It will be hard to lose information in an absolute sense, so just as we already see the strange and unpredictable power that search’s network effects can have in elevating the unknown into the mainstream (e.g. via Google or Yahoo!’s Buzz Index that then becomes self-reinforcing), there will be just as much hard work in making unwanted information unrelevant. Digitally “papering over” one’s electronic tracks will become commonplace — in a world that Remembers Everything, one of the keys to success will be encouraging Selective Memory.