Santa Experiment
The ‘Ask Santa’ experiment was intended as a quick exploration into creating a database for question and response, and to work out some of the technical side of the project, particularly embedding a database into a website, with a talking flash avatar, and running Filemaker Server.

As happens, due to a combination of technical failings and complexity, it took much of a week, didn’t succeed as planned, and I am heartily sick of the site.

But this throws up interesting questions for me about what I am doing generally, walking a very tight rope between being the fool and being a fool - can the outcome of this work be taken seriously in fun, or will my students see it as naff. The comic dies if the audience does not respond as anticipated, much of the art lies in the space and interaction between presenter, presented and audience. The talking santa was inane, fitting perhaps with much of the commercial face of Christams. Some of the people to whom I sent this talking christmas card took it too seriously. Most of the questions asked maintained the mimetic ‘suspension of belief’ created in the context of a figure dressed in the guise of a Santa Claus. Many found the random answer not good enough, expecting some form of artificial intelligence. Sadly this test would not be any contender for the Turin test, the answers being based only on a simple word count. But even with a very limited number of responses, chance threw up a few enjoyably surrealistic juxtapositions of question and answer; to the question ‘Please can I have world peace? the database gave the answer ‘no, actually, I don’t think so.’ To the question ‘Why didn't you bring me a bicycle last year?’ the answer was ‘If not for all this snow.’ And in keeping with the inanity of this particular experiment; to the statement ‘Really Ross, this is rubbish! the database responded ‘Difficult to say, might be the time of year or even the state of mind.’

In looking at examples of ‘chatbots’ on the internet, many match responses to key words in the questions that are asked. With my Santa experiment I found that ‘players’ were generally expecting it to be much more intelligent that it was, and actually found the simplicity of a random answer annoying. The younger and perhaps the younger at heart were more able to take ‘Santa’ at face value.

The site had a simple introduction, created using Reallusion’s ‘Crazy Talk’ application. This was a simple test at embedding a flash file into an online database. On pressing the button the database generates a new record and jumps to a field into which a question can be typed. The second script simply counts the number of words in the question, and displays an answer based on this count. If the question is longer than 25 words, the default setting shows the first record. I attempted to devise answers that would fit a wide range of questions, but I did not spend a great deal of time over this.

Initially this experiment was simply to put a database on line, to see how it might work. But even so it enabled me to collect responses, with no indication on this particular site that I would be doing this.

Click on the image to link though to the crazytalk/flash introduction, and here for the very simple flash game, no more than experiment in scripting drag and drop in Flash.

The Santa experiment has generated some interesting thoughts. Bar staff, super market check out assistants, groups on a night out, they can all happily wear silly hats, indeed it has become almost obligatory in the run up to Christmas. And we wear the paper hat that comes in the cracker for the duration of the Christmas meal, but would look pretty silly wearing it out of that very particular context. The aim of ‘the story teller’s apprentice’ would be to get work from students in a fun way, to get them to participate in a game. This may only work if the content of the game is set at the right level for the age group.

Early Christmas morning I was responding to a Christmas greeting on ‘Facebook’, when I noticed that an Adobe employee that I am ‘friends’ with was looking at a site with a young nephew, a site that tracks the progress of Santa Claus in ‘real time’ albeit with a note explaining that Santa is not bound by the normal space/time continuum. The site also linked to a counter that calculates global population, to inform visitors of the number of children Santa has to visit. The aspect of this site that scared me the most was the fact that it is published by the North American and Canadian early warning defence system, Norad. A reminder that so much of the technology that we are using, from the internet to GPS, have been developed by the military for military ends.