Hmm. I'm for open code broadly but good testing of output is cheaper & much more doable than catching cheating in millions-of-line codebase.
-
-
Replying to @zeynep
For VW, most likely to work and scalable solution was proper testing in the wild. DMCA shutting of research is bad but... this is also true.
1 reply 1 retweet 1 like -
Replying to @zeynep
As more objects get code, I don't see how we get a scaleable army of people figuring out all the messy code. But we can always test outputs.
2 replies 1 retweet 2 likes -
Replying to @zeynep
Also, umm, people who write original code can't sometimes figure out what that subroutine exactly does.

Reverse engineering isn't easy.3 replies 1 retweet 3 likes -
Replying to @zeynep
Playing to the test (like studying or teaching to the test) isn't new. More will come. http://www.theicct.org/news/press-release-vehicle-tests-show-dramatic-variation-nox-emissions-modern-diesel-cars-new-eu …pic.twitter.com/IZGS3w8M2i
1 reply 2 retweets 3 likes
Whenever there is a consequential assessment, there will be attempts at gaming. Whatever is measured with consequence will impact behavior.
Loading seems to be taking a while.
Twitter may be over capacity or experiencing a momentary hiccup. Try again or visit Twitter Status for more information.