28 ms·

These discussions of "unintended consequences of AI" always seem to ignore that those behaviours emerge during the training phase of the models. But then genera

by cynusx 7y ago

These discussions of "unintended consequences of AI" always seem to ignore that those behaviours emerge during the training phase of the models. But then generally you take a trained model and deploy it into production where it doesn't learn any new behaviour anymore. You train them for a specific ability in a controlled environment that reflects deployment environment and after that the deployed system no longer has an ability to learn and will just fail if the environment is different. Journalist in particular seem to be blissfully unaware of this distinction and always cast "artificial intelligence" as following similar mechanics then human brains.

booleandilemma 7y ago
Kinda sounds like people going through trade school.
serioussecurity 7y ago
NLP systems don't "just fail". https://www.npr.org/sections/thetwo-way/2017/01/10/508363607/what-happened-when-dylann-roof-asked-google-for-information-about-race https://www.npr.org/sections/thetwo-way/2017/01/10/508363607...
- Isinlor 7y ago
  If you look for terms that are used by bad people and not discussed otherwise, you will find what these bad people write. I don't see how someone would expect Google to police for example topics related to "The Bell Curve". Recently Sam Harris had trouble dealing with that topic. Should Google suggest "polish concentration camps" when you type "polish concent"? I have no idea.
  - AstralStorm 7y ago
    It will look for what you typed, and find top hit of "German concentration camps in occupied Poland" and second is about naming controversy. Predictable behaviour of a simple statistical debiased algorithm. PageRank is not really an AI.
ganeshkrishnan 7y ago
True but lots of the learned answers are not visible to humans until they start solving them in real life. An example is the "pawn to h6" strategy that reinforcement network chess engines play which even the world champion Magnus copied. Or the infamous Leela bug where it did not learn that you could underpromote to knight in the b(?) File. The devs only noticed it months later when Leela lost a game. Training phase is billions of repetitions. It's not possible for humans to study it.

by cynusx 7y ago

booleandilemma 7y ago

Kinda sounds like people going through trade school.

serioussecurity 7y ago

NLP systems don't "just fail". https://www.npr.org/sections/thetwo-way/2017/01/10/508363607/what-happened-when-dylann-roof-asked-google-for-information-about-race https://www.npr.org/sections/thetwo-way/2017/01/10/508363607...

Isinlor 7y ago
If you look for terms that are used by bad people and not discussed otherwise, you will find what these bad people write. I don't see how someone would expect Google to police for example topics related to "The Bell Curve". Recently Sam Harris had trouble dealing with that topic. Should Google suggest "polish concentration camps" when you type "polish concent"? I have no idea.
- AstralStorm 7y ago
  It will look for what you typed, and find top hit of "German concentration camps in occupied Poland" and second is about naming controversy. Predictable behaviour of a simple statistical debiased algorithm. PageRank is not really an AI.

ganeshkrishnan 7y ago

True but lots of the learned answers are not visible to humans until they start solving them in real life. An example is the "pawn to h6" strategy that reinforcement network chess engines play which even the world champion Magnus copied. Or the infamous Leela bug where it did not learn that you could underpromote to knight in the b(?) File. The devs only noticed it months later when Leela lost a game. Training phase is billions of repetitions. It's not possible for humans to study it.