Claim: Algorithmic intelligence can be creative and inventive.
What do you mean, “algorithmic”? The reference you give is PR nonsense of little substance and doesn’t demonstrate creativity in any relevant sense. That intelligence can be sufficiently creative is demonstrated by humans.
What wedrifid said. In this particular case, the interpretation of the results as supporting possibility of “creativity” in any interesting sense is nonsense, even if the results themselves are genuine. Also, in the vast majority of cases, abstracts are adequate to tell you whether you want to read the paper, not to communicate the (intended meaning of) results.
Frequently. Most notably in medical science (where the most money is involved) but to a lesser extent elsewhere. This particular instance is borderline. I would have to read the rest of the paper to judge just how far the spin has taken the description.
If composing music, deriving scientific laws and generating functional hypotheses doesn’t count, what else?
Actually doing those things.
It is, of course, impressive that the researchers were able to solve the problems in question in a particularly general way.
What do you mean, “algorithmic”? The reference you give is PR nonsense of little substance and doesn’t demonstrate creativity in any relevant sense. That intelligence can be sufficiently creative is demonstrated by humans.
The abstract of scientific papers as PR nonsense? If such studies are not enough, what else do you suggest is valuable evidence?
If composing music, deriving scientific laws and generating functional hypotheses doesn’t count, what else?
This kind of creativity may arise from our imperfect nature. It is important to demonstrate that this can be formalized, intelligently designed.
What wedrifid said. In this particular case, the interpretation of the results as supporting possibility of “creativity” in any interesting sense is nonsense, even if the results themselves are genuine. Also, in the vast majority of cases, abstracts are adequate to tell you whether you want to read the paper, not to communicate the (intended meaning of) results.
Frequently. Most notably in medical science (where the most money is involved) but to a lesser extent elsewhere. This particular instance is borderline. I would have to read the rest of the paper to judge just how far the spin has taken the description.
Actually doing those things.
It is, of course, impressive that the researchers were able to solve the problems in question in a particularly general way.