anticipating Google Wave

Wave: machines enter the conversation September 30, 2009

One of the more interesting features of Wave is the ability to add ‘robots’, or automated functions, to the wave.  Google have already showed off some of these:

• language correction – spelling/grammar by comparing with the aggregated text of the Internet.
• instant translation
• formatting functions – demonstrated as a robot that produces a poll/ multiple choice question.
• moving data between wave and other internet entities such as Twitter.

For developers trying to make the perfect wave, part of the imagining process has to include ‘what machines do we want around the table?’.  The limit of the term machine in this context? – anything capable of providing or processing data.  It’s quite broad, and that’s what makes trying to envisage the potential of Wave quite difficult.

Of course, a wave could include multiple machines, and the machines themselves could interact with the output of others…

These automations could:

• add relevant data into the conversation
  • real world
  • financial/ marketing data
  • location data
  • scientific data
  • search data etc
• analyse the conversation and go find answers to questions
  • respond to changes to other webpages/ datasources
• produce summaries/outcome documents from the conversation
  • publishing to web/ digital files
  • preparing for print publication
  • producing audio/presentation output

Potentially letting machines enter the conversation may help us overcome many of the things  that can make traditional conversations less than effective.  Whether we can get used to having machines around the table or not is another thing.