Originally published on AI Tech Connect.
The gap is evidence, not ability The standard story about researchers moving into industry is that they are over-qualified and under-practical — brilliant at theory, hopeless at code. It is mostly wrong, and it is certainly the wrong diagnosis. Anyone who has run a three-year research programme has debugged systems nobody else understood, maintained a codebase alone, and shipped results under a deadline that could not move. That is engineering discipline. It simply does not look like engineering discipline on paper. The honest description of the gap is that the two crafts optimise for different objectives. Research optimises for novelty and defensibility: the contribution has to be new, and it has to survive an adversarial reviewer who is looking for the reason it does not hold. Product…
Top comments (0)