3 ms·
Thanks to AI can't we also do the opposite? E.g. detect conspiracy theory speech and automatically provide relevant links/warnings to relevant curated sources?
by crazygringo 8y ago
Thanks to AI can't we also do the opposite?
E.g. detect conspiracy theory speech and automatically provide relevant links/warnings to relevant curated sources?
- thfuran 8y agoThat's way, way harder.
- jdietrich 8y agoThe list of things that are true is orders of magnitude smaller than the list of things that appear true. Randomly generating plausible-looking bullshit is a thing we can do now, but ascertaining the validity of a factual statement is tantamount to solving the AGI problem.
- pjc50 8y ago"We" the public can't, but Google might. But that would require them acknowledging the problem. And once they've done that the easier solution seems to be to not recommend (which implies endorsement!) these things.