3 ms·
Or essentially that deep learning is unable to discriminate between beliefs and facts. And we can't read the models well enough to tell which is which. So it w
by ethbro 9y ago
Or essentially that deep learning is unable to discriminate between beliefs and facts. And we can't read the models well enough to tell which is which.
So it would be like being handed a list of high level bridge building rules with no context, then being asked to build a bridge in a geological environment where one has never been built before.
Will it work? Who knows! How would we say?
- jacquesm 9y agoWell, it's early days. We already know it is possible to generate pathological inputs to elicit a false response from a network much like optical illusions can fool humans. For instance, in the context of adaptability of self driving cars: The retraining could theoretically be fixed using an OTA update that periodically pools new training data from in-service vehicles to further improve and generalize the models. You'd have to build that in from day one but presumably anybody busy with self driving cars would have this at the forefront of their thinking. Just so that when and if black clothing with white stripes becomes fashionable we don't end up with a large amount of roadkill because self driving cars interpret these as empty lanes.