>>optima+h2
To be fair, deepmind, google brain, others, are are well funded and have been comparatively lackluster (commercially, they are almost certainly doing more legit academic work). And scientists like to do science, it's hard to focus them commercially. Their (openAI's) success suggests good leadership