BRIO – Bias, Risk and Opacity in AI
National Research Project (PRIN MUR)
Emanuele Bottazzi gave a cycle of dissemination talks titled “Anarchia Artificiale, Osservazioni filosofiche sul ruolo dell’AI” at the cultural association “La corte dei Miracoli” in Milan. The cycle consisted of the following three talks:
Emanuele Bottazzi will give a talk titled “State as a Risk in the Worst to Come. Towards Artificial Anarchy?” at the XXIX conference of the Sociedad C.L. de Filosofía “LA FILOSOFÍA ANTE LOS RETOS DEL FUTURO”. The talk will also concern the connections between AI, risk and sociotechnical systems considered from the perspective of social ontology. The congress will take place at the Campus Unamuno of the Universidad de Salamanca on the 25–27 October 2023.
Emanuele Bottazzi will participate, with a talk titled “Adversarial Writing”, to the debate “Dall’ispirazione alla sostituzione: la scrittura letteraria di fronte all’IA”, which will take place on October 21 at 17–18:30 in the context of the 20th anniversary of the literary journal Nazione Indiana. The talk will also discuss Large Language Models from a Wittgensteinian perspective and the anthropomorphic biases that they introduce.
https://www.nazioneindiana.com/2023/10/10/fragilicity-festa-per-il-ventennale-di-nazione-indiana/
Mattia Petrolo is giving a talk with the title “An epistemic route to AI” at the Permanent Seminar of Philosophy of Sciences of the Centre for Philosophy of Sciences of the University of Lisbon to be held in Lisbon (both in presence and on Zoom) on December 5th.
Mattia Petrolo is giving a talk with the title “A logical approach to algorithmic opacity” at the workshop BEWARE-23 – Bias, Risk, Explainability and the role of Logic and Logic Programming, to be held in Rome on November 6th as part of the AIxIA 2023 conference.
The in-house artist Guildor will deliver a performing lecture on November 9th at Politecnico di Milano (room 26.15) 12:30–14.
Greta Coraglia talks on “Categorical models for subtyping” at the 108th Peripatetic Seminar on Sheaves and Logic, organized in Terrasini (Palermo) on September 16-17.
We present the innovative tool capable of detecting possible bias in Machine Learning models, developed in collaboration with Alkemy. The software is opensource.
Click here to try the tool or contribute to its development yourself: https://github.com/DLBD-Department/BRIO_x_Alkemy
For a preview of the functionality, watch our YouTube video: https://www.youtube.com/watch?v=9W8_DTmXcts
Members of BRIO will present the first output of the collaboration BRIOxAlkemy in a workshop with title Algoritmo e Pregiudizio at the 10th edition of the festival Il Tempo delle Donne. The BRIO team will introduce and show the bias detection application developed in collaboration with Alkemy, the industrial partner of the project.
The workshop will take place at Triennale Lab of Triennale Milano (Palazzo dell’Arte, Viale Alemagna 6, 20121 – Milano) on
More information on how to participate at
Giuseppe Primiero presented “Trustworthiness Ranking with Uncertain Information” (joint work with Davide Ceolin (CWI), Francesca Doneda and Francesco Genco (UNIMI)) in an invited lecture at the Joint Workshop GRID/ENIGMA at the 20th International Conference on Principles of Knowledge Representation and Reasoning (KR2023), held in Rhodes (Greece) on 2-3 September 2023.