PICNIC07: Future Technology Trends, Arnold Smeulders

From MultimediaN. Dreadful demo of 3D museum based on semantic web, with extremely rousing music. Not quite sure what it does, mind.

Second search engine: crawls LiveJournal looking for moods (a bit like We Feel Fine then).

Third one: P2P video exchange search engine. "Windows for your TV screen". Connects media centres in living rooms to provide "enormous computing power" to do video analysis.

Fourth one: video browsing product.

Have to say, in the limited time here I didn't really get much of a feel for any of these products.

PICNIC07: Future Technology Trends, Alexander Straub

Pixsta: interesting way of looking at images without looking at images themselves: looking at similarities in and closeness between images, and building web applications around these.

Shows demo of shopping site showing handbags similar to the bag you're currently looking at - in an interestingly messy display. Quite pretty actually, a nice way to browse items. A very quick way for human beings to browser 4-500 products in a short time, using the brain for what it's best at - recognising images, finding similarities and matching patterns. Simple but neat. Cool for something like Etsy.

No tagging behind the shopping demo: it's all done via the image data itself.

PICNIC07: Future Technology Trends, Nikolaj Nyholm

This is an introduction to work at Polar Rose, inspired by growth of photo sharing sites online. Flickr doubles in size every 8 months - how do we manage this data?

Polar Rose does facial recognition. Traditionally this needed a very controlled environment. Extracts 3D data from 2D image and computes lighting sources (much like the PhotoSynth guys yesterday, I'm guessing). There's a lot of data in faces: we can work out whether a face is male or female quite accurately.

It's a plugin for Firefox, highlighting faces and asking them to be recognised. During the beta phase they're gathering face data. Getting into indexing video too.

PICNIC07: Everything is miscellaneous, David Weinberger

We're good at organising things, coming up with organisational schemes. But there's always a box for things that don't fit elsewhere: Miscellaneous.

Digitally, allowing this box to grow and take over the organisation is the right thing to do. The web could only scale because it didn't ask us for permission to post - it was easy to chuck stuff in. The real worlds main purpose seems to be to keep things apart and in one place - you cannot put two things in the same place. But this is exactly the wrong principle for the digital world; the rules we apply to the physical just aren't appropriate.

Human editorial experts, for instance, make decisions about importance - but in the digital world they don't have to organise things into discrete boxes. Since Greek culture, we've assumed ideas are organised the way the real world is. "The wise person carves nature at its joints" - Plato - implies ideas have physical structures (joints).

"To think that there is a single order to the universe is to consider the universe a shoe store"

There are 3 orders of organisation:

1. The things themselves - e.g. a physical photographic archive;
2. Physically separate the metadata - e.g. a card index gives you maybe 3 ways of organising (author, subject, title). Relies on shrinking information down to card-side;
3. Digitise contents and metadata;

4 principles of change:

1. Filing things in multiple places at once;
2. This causes messy in the real world, but digitally messiness is good - every link adds value and information. Data mess can be analysed and sorted
3. No difference between data and metadata online where there is in the real world;
4. Unowned order; in the real world, owners of items also organise those items; in the digital world ordering can be divorced from what's being ordered: users control and own the organisation of data

We're not letting experts prune the world for us; nowadays we're letting everyone bring this order, in a messy way. We can't predict what other people will be interested in - but now we don't have to decide that for others.

It's also better for businesses to share information than lock it up. e.g. the airline industry sharing fares allows sites like Travelocity to add value, Farecast to analyse the data... all of which sends business back to the airlines.

Explicit vs implicit: Facebook gathers explicit information (how you know someone) and surrounds it with messy fauna of implicit information.

Wikipedia: quote from Jimmy Wales about realising an independent viewpoint by having people edit an article until the arguments stop. "The expert who is unwilling to engage in conversation is losing relevancy". Knowledge has always been social; in a good conversation, we drive the bugs out of knowledge.

Britannica can be trusted; Wikipedia can't necessarily, but it has a great deal of credibility. Britannica gets its credibility from the credentials of its authors; Wikipedia reminds us that facts aren't as black and white as we tend to think.

We've succeeded as a species because we've externalised functions of knowledge: writing externalises memory, calculators externalise maths, etc. We're now externalising meaning. The great step forward we're taking is to

PICNIC07: Designing Killer Experiences, Peter Frankfurt

Every company in entertainment wants to be Pixar when they grow up. Started off with title sequence for Seven, attracted clients and young designers. Putting something onto a cinema screen is quite attractive for these folks :)

Started working on Transformers 1+ year before the film came out. Teaser created with no footage of the film. Michael Bay saw it: "Fucking cool". Shows different versions of the trailer for different markets, with logotype transforming into various scripts. Doing 20+ versions was financially pretty good for a graphics company ;)

Amusing trailer for Spike TV - like 24hrs of Bravo TV squeezed into 15 seconds with volume and contrast up to maximum.

Very visual presentation, difficult to get notes done...

Building virtual worlds is about taking something familiar and doing something to it that upsets, delights, whatever - just gets your attention.

Talks about World Trade Centre redesign: the original proposals were deeply unpopular so it was put out to the design community. Small agencies worldwide got together. "We literally had to create a world in lower Manhattan in 6 weeks". Decided to treat the presentation and process of the design as a film-making process: storyboarding informed a lot of the design, and a film was created as part of the presentation.