With this paper, we aim to put an issue on the agenda of AI ethics that in our view is overlooked in the current discourse. The current discussions are dominated by topics such as trustworthiness and bias, whereas the issue we like to focus on is counter to the debate on trustworthiness. We fear that the overuse of currently dominant AI systems that are driven by short-term objectives and optimized for avoiding error leads to a society that loses its diversity and flexibility needed for true progress. We couch our concerns in the discourse around the term anti-fragility and show with some examples what threats current methods used for decision making pose for society.

In this paper, we analyze whether regulation reduced risk during the credit crisis and the sovereign debt crisis for a cross section of global banks. In this regard, we examine distance to default (Laeven and Levine, 2008), systemic risk (Acharya et al., 2010), idiosyncratic risk, and systematic risk. We employ World Bank survey data on regulations to test our conjectures. We find that regulatory restrictions, official supervisory power, capital stringency, along with private monitoring can explain bank risk in both crises. Additionally, we find that deposit insurance schemes enhance moral hazard, as this encouraged banks to take on more risk and perform poorly during the sovereign debt crisis. Finally, official supervision and private monitoring explains the returns during both crisis periods.

No abstract

International audienceFor a Brownian motion $B=(B_t)_{t\le 1}$ with $B_0=0$, {\bf E}$B_t=0$, {\bf E}$B_t^2=t$ problems of probability distributions and their characteristics are considered for the variables $$ \begin{array}{c} {\mathbb D} =\displaystyle\sup_{0\le t\le t'\le 1}(B_t-B_{t'}),\qquad {\mathbb D}_1=B_\sigma-\inf_{\sigma\le t'\le 1}B_{t'}, \\ {\mathbb D}_2=\displaystyle\sup_{0\le t\le\sigma'}B_{t}-B_{\sigma'}, \end{array} $$ where $\sigma$ and $\sigma'$ are times (non-Markov) of the absolute maximum and absolute minimum of the Brownian motion on $[0,1]$ (i.e., $B_\sigma=\sup_{0\le t\le 1}B_t$, $B_{\sigma'}=\inf_{0\le t'\le 1}B_{t'}$)