Next, feed Watson the corpus of /b/.
IBM's Watson Memorized 'Urban Dictionary,' Then His Overlords Had to Delete It
31–40 of 163 posts
Re: IBM's Watson Memorized 'Urban Dictionary,' Then His Overlords Had to Delete It
#32I like how someone commented on the main article that the time is getting close to where AI can step up to the plate of creativeness and how widespread and easy this will make our lives. Watson is a giant server farm, not a single PC, this stuff won't make a huge impact until IBM can shrink it or until computers get much much faster and smaller. Not that it won't happen, it's just not "around the corner" in any way.
"Google Search runs on a giant server farm. It won't really make a huge impact until it can run on a single PC." Can you explain why your argument is valid, but the one above isn't?
Mostly just that IBM's Watson is so very different then google search. That said you are right, there isn't much of my point that isn't invalidated when you include the ability to exchange data through the internet. Not everyone needs a Watson at home for it to be personable either, they just need IBM to save their personal settings, etc.
I suppose I was just imagining a world where everyone has their own Watson at home and not served through the web.
Re: IBM's Watson Memorized 'Urban Dictionary,' Then His Overlords Had to Delete It
#33Note to future self: we can probably neutralize indefinitely any malicious AI by directing it to start consuming tvtropes.com.
Re: IBM's Watson Memorized 'Urban Dictionary,' Then His Overlords Had to Delete It
#34I am more interested as to how they "..scraped the Urban Dictionary from its memory." - is it trivial to just delete something learned by AI?
they just overwrote every urban dictionary word instances with the string "rainbow".
So Watson still wants to call the query bullshit, but says rainbow instead.
Re: IBM's Watson Memorized 'Urban Dictionary,' Then His Overlords Had to Delete It
#35Nothing that Watson learned from the Urban Dictionary could possibly be any dirtier than what I hear from enterprise people all the time: "We use our deep subject matter expertise to deliver value through actionable advice that enables our clients to harness the power of best practices in order to shift their paradigms and achieve 10X deltas against competitive industry metrics."
That's in Urban Dictionary too ...
http://www.urbandictionary.com/define.php?term=synergasm
;)
Re: IBM's Watson Memorized 'Urban Dictionary,' Then His Overlords Had to Delete It
#36I'd really love to hear some of the 'not fit for print' things that Watson said.
Re: IBM's Watson Memorized 'Urban Dictionary,' Then His Overlords Had to Delete It
#37Nothing that Watson learned from the Urban Dictionary could possibly be any dirtier than what I hear from enterprise people all the time: "We use our deep subject matter expertise to deliver value through actionable advice that enables our clients to harness the power of best practices in order to shift their paradigms and achieve 10X deltas against competitive industry metrics."
Yeah baby, talk dirty to me.
Re: IBM's Watson Memorized 'Urban Dictionary,' Then His Overlords Had to Delete It
#38I am more interested as to how they "..scraped the Urban Dictionary from its memory." - is it trivial to just delete something learned by AI?
Watson's 'memory' is just a big database of facts, rules, and statistical models. To 'forget' a source they'd just have to rebuild any models derived from it and purge any facts it had extracted.
Re: IBM's Watson Memorized 'Urban Dictionary,' Then His Overlords Had to Delete It
#39I like how someone commented on the main article that the time is getting close to where AI can step up to the plate of creativeness and how widespread and easy this will make our lives. Watson is a giant server farm, not a single PC, this stuff won't make a huge impact until IBM can shrink it or until computers get much much faster and smaller. Not that it won't happen, it's just not "around the corner" in any way.
"Google Search runs on a giant server farm. It won't really make a huge impact until it can run on a single PC." Can you explain why your argument is valid, but the one above isn't?
I would say that Google Search is only useful to the majority of people when it's searching across an index of the entire web. This scale is likely not achievable at the desktop scale. On the other hand, the algorithm itself (though evolving) has been around for years.
I view Google Search as a less complex algorithm over a larger set of data, while Watson is a more complex algorithm over a smaller set of data. (I've been known to be wrong ;-)
Re: IBM's Watson Memorized 'Urban Dictionary,' Then His Overlords Had to Delete It
#40"Watson couldn't distinguish between polite language and profanity ...
Ultimately, Brown's 35-person team developed a filter to keep Watson from swearing ..."
Sounds just like what happens when you raise kids. "Daddy why is XXX a good word but YYY a bad word?"
"It just IS. Don't say that word again."
"Ok Daddy" (kid adds word to internal blacklist)