How to innovate like a robot

Over on Innovation in Practice it's suggested that machines will most likely never be able to innovate - Automated Innovation.

Granted such discourse can be quite dull - computers/machines cannot think/love/walk properly/chew gum/...../invent seems to be for my generation the equivalent of the ancient "how many angels can dance on a pin?".

What saved this post for me was the description of what innovation is -

For a machine to innovate, it would need to:

1. Take a product or service and break it into its component parts
2. Take a product or service and identify its attributes (color, weight, etc)
3. Apply a template of innovation to manipulate the product or service and change it into some abstract form
4. Take the abstract form and find a way for humans to benefit from it

Surely not. This is the blogosphere, at a minimum 4 should be "... find a way for cats to benefit from it."

More seriously if the process of innovation were to be simulated in a worthwhile way, some benefit would probably need to be delivered to the innovator. What I'm saying here is that feedback is required. I'd also argue that for a machine 2 and 3 are probably not required either. After all if a machine was capable of generating prototypes at a fantastic rate - consider synthetic drugs - and then test them for good and bad effects, it is quite likely that useful new drugs could be developed. For all I know this is happening right now.