2 ms·
If Google (very many smart people, massive amounts of data, massive amounts of processing) can't do it then it seems like it could be a hard problem. The abuse
by DanBC 11y ago
If Google (very many smart people, massive amounts of data, massive amounts of processing) can't do it then it seems like it could be a hard problem.
The abuse of ContentID on YouTube, or auto-generated DMCA requests, show that letting content owners report stuff doesn't work particularly well at scale.
- jarcane 11y ago"Hard"? Or just "expensive" and "not easily automated away"? Because there's a difference. In this case though, I struggle even to see how that applies. The examples given are literally cases of identical text content between pirated and real works. We've had text figured out for a while now, it's not like it'd take image recognition software to see that the diff between the two books literally amounts to two words for the bogus author's name.