Skip to main content
European Commission logo
AI Watch

Evolution of AI technology

Approach

There is still much uncertainty over how to assess and monitor the state, development, uptake and impact of AI as a whole, including its future evolution, progress and benchmarking capabilities. While measuring the performance of state-of-the-art AI systems on narrow tasks is useful and fairly easy to do, in AI Watch we are trying to go one step further by mapping these performances onto more general AI. We aim to understand how these can have an impact on society in terms of benefits, risks, interactions, values, ethics, oversight into these systems, and more. 

In brief, we are trying to provide a deeper understanding of the evolution and progress of AI by: 

  • Collecting, exploring, and monitoring data about AI results, progress and capabilities
  • Analysing readiness and maturity levels of AI technologies
  • Disentangling indicators, metrics, and explanatory factors behind AI results

 

Tools

Publications

Report | | Joint Research Centre

This report provides the analysis of multiple indicators related to the development of artificial intelligence from several perspectives.

Scientific paper | | Joint Research Centre

The widespread use of experimental benchmarks in AI research has created competition and collaboration dynamics that are still poorly understood. This paper provides an innovative methodology to explore these dynamics and analyse the way different entrants in these challenges behave and react.

Report | | Joint Research Centre

This report summarizes the evolution of AI, it introduces the “seasons” of AI development (i.e. winters for the decline and springs for the growth), describes the current rise of interest in AI, and concludes with the uncertainty on the future of AI.

News

Collaborations

Universitat Politècnica de Valencia

The Universitat Politècnica de Valencia partners with AI Watch in the development of AICollaboratory, a data-driven collaborative initiative for collecting and organising AI evaluation results.