Collaborating Authors

Revisiting the Importance of Individual Units in CNNs via Ablation Artificial Intelligence

We revisit the importance of the individual units in Convolutional Neural Networks (CNNs) for visual recognition. By conducting unit ablation experiments on CNNs trained on large scale image datasets, we demonstrate that, though ablating any individual unit does not hurt overall classification accuracy, it does lead to significant damage on the accuracy of specific classes. This result shows that an individual unit is specialized to encode information relevant to a subset of classes. We compute the correlation between the accuracy drop under unit ablation and various attributes of an individual unit such as class selectivity and weight L1 norm. We confirm that unit attributes such as class selectivity are a poor predictor for impact on overall accuracy as found previously in recent work \cite{morcos2018importance}. However, our results show that class selectivity along with other attributes are good predictors of the importance of one unit to individual classes. We evaluate the impact of random rotation, batch normalization, and dropout to the importance of units to specific classes. Our results show that units with high selectivity play an important role in network classification power at the individual class level. Understanding and interpreting the behavior of these units is necessary and meaningful.

Towards a Framework for Certification of Reliable Autonomous Systems Artificial Intelligence

The capability and spread of such systems have reached the point where they are beginning to touch much of everyday life. However, regulators grapple with how to deal with autonomous systems, for example how could we certify an Unmanned Aerial System for autonomous use in civilian airspace? We here analyse what is needed in order to provide verified reliable behaviour of an autonomous system, analyse what can be done as the state-of-the-art in automated verification, and propose a roadmap towards developing regulatory guidelines, including articulating challenges to researchers, to engineers, and to regulators. Case studies in seven distinct domains illustrate the article. Keywords: autonomous systems; certification; verification; Artificial Intelligence 1 Introduction Since the dawn of human history, humans have designed, implemented and adopted tools to make it easier to perform tasks, often improving efficiency, safety, or security.

Ten Ways the Precautionary Principle Undermines Progress in Artificial Intelligence


Artificial intelligence (AI) has the potential to deliver significant social and economic benefits, including reducing accidental deaths and injuries, making new scientific discoveries, and increasing productivity.[1] However, an increasing number of activists, scholars, and pundits see AI as inherently risky, creating substantial negative impacts such as eliminating jobs, eroding personal liberties, and reducing human intelligence.[2] Some even see AI as dehumanizing, dystopian, and a threat to humanity.[3] As such, the world is dividing into two camps regarding AI: those who support the technology and those who oppose it. Unfortunately, the latter camp is increasingly dominating AI discussions, not just in the United States, but in many nations around the world. There should be no doubt that nations that tilt toward fear rather than optimism are more likely to put in place policies and practices that limit AI development and adoption, which will hurt their economic growth, social ...

The rise of AI


From virtual assistants to driverless cars, technology imitating human intelligence is on the rise. But at what ethical cost and how do boards future-proof their organisations in the face of rapid change? Earlier this year, a Japanese insurance company made headlines for doing something that company executives and directors around the world have been anticipating - and fearing - for years. Fukoku Mutual Life Insurance made 34 of its staff redundant and replaced them with artificial intelligence (AI) system IBM Watson. Japanese newspaper The Mainichi reported the company will be using Watson to determine payout amounts and check customer cases against their insurance contracts. Evidently, the future of AI is already here and technology has been changing the world at a dramatic pace.