(The slides are on available from [Reginald Braithwaite's blog http://www.braithwaite-lee.com/weblog/2005/11/powerpoint-web-20-beyond-ajax.html]) 

* Applications "Do stuff with/to stuff"
* (Stuff includes People)
* Surnames suck (why do we have to put names into two fields just for the computer?)
* Don't make me plan up front (works for space shuttle software, not for organizing my data)
** This applies to users not to web developers who definitely need to plan in advance
** Workflows especially suck
** Don't trap me in your unreality
** "Everything fits into a system" isn't true
*** Think of music having different scales around the world - there's no one scale that best represents it
*** Usually a manager assumption
* Taxonomies are static
** They need to be fluid and change over time (your definition of who is "hot" will change)
* Suckage cannot be fixed in CSS
** Outlook on the web still sucks even though it's been well replicated on the web
* Plan ahead to not suck
** BBEdit: It still doesn't suck.
** One of Microsoft's releases of Word for Mac "Sucked" according to reviewers, so MS fixed it and then outfitted their booth weenies with "Word 3.1 Sucks Less"
* Messy and flexible is good
** Del.icio.us is great even though it's messy and flexible
** But, why do we have to tell it that something with the word "Rails" in the title is about Rails and not about "realestate" or "rehab"
** JonLax says: check lazysheep for a fix to this problem 
** DavidJones: the value of tags is that you put effort into it. 
** Reg: I should put effort in so other people can benefit from it? 
** DavidJones: You're putting it in for your own use that other people can benefit on.
** Reg: Basically I want it to suggest what other people said already
** (Which it may already do but it still sucks if you can't see it)
* Attempts to not suck that suck
** Toronto Star's registration page is way to long 
*** They want us to do the work
** Yahoo! knows that Reg likes Rock Climbing and is married but they never show him ads for rock climbing or anniversary gifts
*** JonLax: That's a business development problem within Yahoo! not a tech problem
*** Reg: Give the man a potato! That's the point.
*** DavidJones: They should know that you've never clicked an ad and sthat they need to show you different ads.
*** Reg: Their advertisers need to have higher demands and standards that will force them to do things to optimize click throughs.
* Lessons learned
** You know things about me and about us so watch, don't ask
** Google does this right - type in "what is ajax?" and you'll get a useful set of responses
*** Ask Jeeves gives answers about the Greek hero, not the technology
*** Within the context of the Reg's search, why does Google beat Jeeves?
*** [Sutha Kamal http://suthakamal.blogspot.com]: Google works better for Geeks because of the way PageRank works
**** Forget the word "geeks" for a moment: My understanding of the point is that if links drive relevance, Google favours the subset of the user base that compose web pages. since people who blog/make web pages are more interested in AJAX the technology than Ajax the mythical person, Google gives us the technology. This is very interesting.
*** DavidJones: Google has filed patents on personalizing search so that it will learn from your results
*** JayGoldman: Google got the context "what is AJAX?" not "who was AJAX?"
*** John: Google is becoming an ISP so that they can measure how long people stay on pages after they leave Google
*** Reg: Google beats Jeeves because they have more data and they keep building and working on it.
*** Jon Lax: It's not just collecting data - Google has the experience to do something with it
* Broad is better than deep
** Better to do something simple (broad) with data than deep (sophisticated)
** If I need a mathematician to describe what I'm doing with the data, that's deep
** John: Yahoo! started showing ads about cribs, etc. after he started looking for books about expecting and birth
*** Was it Yahoo or Amazon?
** DavidJones: This is called "attention" and is one of the biggest web trends of the year
** KenSchaffer: Ken Gilmour is behind all that
** Note: I can't find a link - someone want to add it?
** Think this is what is being referred to above: http://attentiontrust.org/
* More users means breadth
** Enterprise = fewer users = work harder to make data work
* Naivete works en masse
** You don't need PhDs to make this work - you just have to plan for it to happen
* Long tails imply fat heads
** Creates a power law distribution whcih is useful in some cases but very problematic in others
** Popular bloggers become more popular and you loose the less popular ones who still have something to say
** This is NOT how it works in nature - the alpha male gets all the chicks
*** Actually I said this is EXACTLY how it works in nature when chicks (literally in the case of birds) can see which males are popular, such as in a Lek.
* Things do not change; we change
** Folder organization on a computer gets outdated quickly
** How much of the archived data is relevant?
** Tags are at the time they're tagged but not over time
* New data cleans up old business
** JonLax: Watching the past is a useful predictor of the future in some situations but not in others. There may be no amount of data or deep programming that might get who I decide is 'hot' later.
** Reg: You spend time on the things that are important in your life. If you mark things as 'urgent' but you never actually do it, then it's not important. That's predictable.
** JonLax: Do we change in predictable ways?
** Reg: Absolutely. For a very large number of people, you can make important predictions out of small and shallow information. Someone will come along and serve the niche of leftover users later.
** Someone (name?): NOW magazine does a Top 100 in the city list every year, but they say the top coffee spots and they don't mention Jet Fuel. It's important that you know who you're targeting so you know which market segment you need to hit.
** Reg: How do we know something is popular? It's because we see so many of them. Why is Starbucks better than Jet Fuel? Because of Jet Fuel was better they would be everywhere. It's the tyranny of the Long Tail.
** John: What if you can't see the feedback effect? What if I can historically supply context? Google's been watching me do searches so they know what I like. This gets me to one of my big pet peeves about people. The Tyranny of Self-Referential Groups. 
** Reg: That's just the Tyranny of the Long Tail scaled down. Regardless of the scale, if you lack the feedback loop then you get a Long Tail. For example: I like positive feedback so most of my friends like me.
* 1. Social is more than skin deep
** social apps give more users and more data
* 4. Manage the fat haead
* 5. Handle change