Évaluation des impacts environnementaux de l'IA générative
Dans ce converstation, je vous présenterai GenAI Impact, qui se positionne comme une association dédiée au développement de communs numériques pour mesurer et réduire l'empreinte environnementale de l'IA générative. Elle agit sur trois volets : élaboration de méthodologies d'estimation des impacts, développement d'outils open source (notamment Ecologits), et sensibilisation des parties prenantes autour des enjeux environnementaux de l'IA générative.
Biographie: Valentin Defour is data scientist at AVISA and focus on the following topics
As a data scientist within the AI, DS & MLE team of a fintech company :
- Developing and industrializing (API/DSS) machine learning models to answer business requirements ;
- Building an ethic & trustworthy AI framework to be used by the team ;
- Building a ML carbon footprint estimator to provide environemental insights about ML tasks ;
- Leveraging GenAI to answer business issues (real time text data classification, self service BI)
As a consultant,
- Writing articles about data, AI and ML ;
- Building training about specific subjects : greener AI, ethics (fairness) of ML
- Developing and industrializing (API/DSS) machine learning models to answer business requirements ;
- Building an ethic & trustworthy AI framework to be used by the team ;
- Building a ML carbon footprint estimator to provide environemental insights about ML tasks ;
- Leveraging GenAI to answer business issues (real time text data classification, self service BI)
As a consultant,
- Writing articles about data, AI and ML ;
- Building training about specific subjects : greener AI, ethics (fairness) of ML

PDF version