We trust you and your ability to police each other….. But still ;-) bring on the thought police.


As we develop our product and others that are related, like Amazon Echo, we want to be able to allow people to create public information that is shared for all. As an example, someone might say “cats are small mammals etc."  They would decide to make that public so that everybody can benefit from the definition of “cat”. Obviously, there are a lot of things individuals are intrigued that might be only known by a few people, but there are also many that would be publically shared. In the case of the “cat”, it's also possible someone purposely erroneous says “cats are small dogs with pointy ears etc...”.  This would not ;-) be generally agreed across many people.


So we would need the ability, like Wikipedia, to correct these or have people provide alternatives to resolve a particular definition. We want to make this open so that we don't police it as there are too many terms to manage.


In general, we will trust people to define things correctly, nicely in a non-offensive way.  Most people will.  I find 99.9% of the people are good.  We can and will scan for words and other phrases before publishing a particularly new term. This prevents many ‘bad’ apples but not all.  As we want the public to dynamically contribute language, people could also define bad phrases or words.  This allows for fast adaption that a single company could not accomplish easily.  Someone could say “supercat is a bad word”, supercat being some new lingo we’ve never heard.  We would then track that and try to get others to validate.  Like Gmail spam buttons.  There are many machine learning mechanisms to help us in the process in addition to people directly.  


We will also allow people to flag definitions to help reduce bad information.  Since we want people's voices to be heard vs computer voice for definitions where possible, we think this will help self-police “evil” definitions.   Like in any other situation, there will be a rating list of people and blacklist of people who provide bogus or offensive information etc.…


So, build trust, and self-police, but allow for all to contribute to the collective knowledge.  


Human enablement!

Leave a comment

Please note, comments must be approved before they are published