Home → News → Malevolent Machine Learning → Abstract

Malevolent Machine Learning

By Chris Edwards

Communications of the ACM, Vol. 62 No. 12, Pages 13-15

[article image]

At the start of the decade, deep learning restored the reputation of artificial intelligence (AI) following years stuck in a technological winter. Within a few years of becoming computationally feasible, systems trained on thousands of labeled examples began to exceed the performance of humans on specific tasks. One was able to decode road signs that had been rendered almost completely unreadable by the bleaching action of the sun, for example.

It just as quickly became apparent, however, that the same systems could just as easily be misled.


John Canessa

A very interesting article. In particular the paragraphs towards the end of the paper regarding the use of the "softmax" in predictions. I am no expert in ML but have taken a couple courses and read a couple books. I am about to start work on a model for predicting orientation of a specific type of images. Will spend time experimenting with alternate techniques when the model performs predictions.

Displaying 1 comment