You can add location information to your Tweets, such as your city or precise location, from the web and via third-party applications. You always have the option to delete your Tweet location history. Learn more
Which makes the concept of “understanding” not very interesting and overused. Better to refer back to tasks and benchmarks and not extrapolate from that as @tdietterich suggested in his recent medium piece.
it is worse than just “they make silly coherence mistakes” when i forget to carry a 1 in a math problem, that is a silly mistake. i have the overall framework & goof. these models don’t. all they do is dredge statistics; the reading models NEVER build real models of the world.
Dredging statistics would be fine if they were the right statistics. But GPT-2 &co are not capturing statistics about agents, goals, intentions, actions, etc. So unsurprisingly, they don't learn about them
I agree that "silly mistakes" understates it. It has no clue about many things that are obvious for humans. But it captures some meaningful statistics and some limited model of the world (e.g., Obama is a human). Again I would refrain from using "not real" as a criticism.
Twitter may be over capacity or experiencing a momentary hiccup. Try again or visit Twitter Status for more information.
