Hacker Newsnew | past | comments | ask | show | jobs | submitlogin

I can see what Google tried to do. But it's obvious that the technology failed miserably, on a problem that likely should not have been "solved" with automation. What if there had been a terrorist attack? Clearly these messages would be not just incorrect, but unimaginably inappropriate. It's clearly not ready to be used now, if it's even possible to teach a machine what 9/11 conspiracy theories are.

The team that built this should be dissolved. This was a bad call on the engineering side, the product side, and the management side. Nothing about this was a success.



It’s a success except when it’s not. We only see headlines when it fails. Calling for the dissolving of this particular integrity team is ridiculous.

Mistakes will happen, the team will learn from them, and their product will improve.


It should never have been built in the first place. It's an attempt to put a band-aid on the conspiracy theories on the platform, but completely lost sight of the fact that bad things do happen and that bad things often are visually similar.

Even if it set out to achieve its original goals flawlessly, it fails to actually solve the problem Google faces in the first place (being a platform for misinformation). If you got all the way to a conspiracy theory video, is Encyclopedia Britannica really going to change your mind? And being a not-flawless computer system, it will _always_ have false positives, which means as long as it exists it will flag events like this as misinformation. The latter, in my opinion, is inexcusable. Any engineer building any sort of classification system knows that there will be false positives, and nobody at Google thought that this would be problematic and stopped the system from being built.


I can't think of any place where this sort of "feature" would've been appropriate. An article from Wikipedia, which anyone can edit, is like the least convincing thing they could've linked to.


Of all the things to get angry at Google about, falling short every once in a while at solving an incredibly difficult technical challenge is just ludicrous.


Or maybe (because they might have some great programmers, but that's not necessarily the same as having smart people) they are "solving" something that isn't a technical problem in the first place with code?


What is the problem that they are solving? I don't want Google to be my lens on the world. I never paid them for that. It is for this reason that I unsubscribed from YouTube Red. They stopped being a neutral platform.




Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: