Using usefulness to prove non-obviousness

Question

Is that possible? Or must usefulness and non-obvious be completely independent arguments?

For example, if an invention was never invented because everybody thought it would be a useless invention, would it be non-obvious if someone invented it and discovered its usefulness perhaps in some obscure scenario?

Also, referencing the first answer to: When is combining prior art to invent something new obvious and when is it not obvious?

If A produced output X and B produced output X, but the combination of A + B produced 10X, then it is a non-obvious combination. Would this be an example of using usefulness to prove non-obviousness? We can assume that the combination A + B is mechanically obvious, etc.

Answer 1

Yes - but do not use the word usefulness - "unexpected results" and "teaching away" are both patent law factors in determining obviousness. Teaching away means repeated and unchallenged prior art references that say it would be ineffective to combine A and B to solve your problem.

But do not use the word "useful" in this regard. Three primary hurdles to getting a patent, and the law section stating them, are - section 101 usefulness; 102 novelty; and 103 obviousness. They are separate, although recently courts keep mentioning obviousness observations when doing a 101 analysis.

And, do not say that making A + B is mechanically obvious or any other kind of obvious. You mean it is a straightforward mechanical operation not that it is obvious to do to solve a particular problem.