Few thoughts about Standardization & Artificial Intelligence a new discussion table promoted by ISO

Few days ago I was honed to act as Head of the Italian delegation to the first meeting of the newly established ISO/IEC JTC1 SC42 technical committee, in Beijing – China, which has taken on the standardization of Artificial Intelligence (AI).

The interest in the theme is high, at the conference we already have a full presence and contribute form 15 countries: Canada, China, Denmark, Germany, Finland, France, India, Israel, Ireland, Italy, Japan, Korea, Russia, UK, US and 2 observing countries such as Australia, Luxembourg.

I enjoyed the experience specially for someone that always looked at standards from “outside-in” perspective. So, I reverted the perspective enjoying also to contribute to the discussion sitting between Israel and Japan, in a rigorous alphabetic order 🙂

Artificial Intelligence is attracting multiple forms of interest not only from the IT technical community, obviously, but increasingly from the business world that is looking on it as one of the basic ingredient of the Digital Transformation for every industry.

AI is shaping itself as a powerful B2B, Business-to-Business, concept. We are actually not only observing AI systems rather, more and more, a large set of familiar systems are stating to appear to us smarter and in many cases we ignore completely the reason behind. It is like when we take a picture with our digital camera and look into it and see not only the scene we expected to see but something little bit different and rich like when the camera is highlighting main faces and it is trying to focusing on it. So, the conclusion is that the camera is becoming smarter.

Recently someone argued that AI is like the last century Electricity, another B2B concept, where companies, at beginning of 1900, appointed executive positons to leverage it in their environments and support Electro Transformation strategy.

All of this is heading us to naturally thinking that there is the need and it is the right time to look at the AI phenomena from a standardization point of view, at global level, as we did for a number of other B2B concepts in the past to create better opportunities to growth all together and go ahead in an orchestrated way in this area.

The spirit of Beijing meeting was exactly this, form my perspective. We started to discuss the foundation elements of the need of such orchestration activity and define possible new areas to be investigated in the next future.

The discussion and the work on Standardization and AI is just started with Beijing’s resolutions.

In the immediate future we are going to work on two Foundational Standards projects trying also to create links with other standard tables and contribute to build a common set of AI concepts and terminology, that I personally think is a very important first step in this area, as well explore Framework for Artificial Intelligence Systems Using Machine Learning.

In addition, we are going to study and go deep in 3 major areas to develop new possible work items proposals for the next steps such as:

1. Computational approaches and characteristics of artificial intelligence systems. The angle here is to study different technologies used by the AI systems including their properties and characteristics as well as industry practices, processes and methods for the application of AI systems.
2. Trustworthiness of AI systems. In this area we are going to investigate approaches to establish trust in AI systems through transparency, verifiability, explainability, controllability, etc. as well as engineering pitfalls and assess typical associated threats and risks to AI systems with their mitigation techniques and methods. The goal is also to investigate types of sources of bias in AI systems with aim of minimization.
3. Use cases and applications. The aim here is to identify different AI application domains and the different context of their use and collect representative use cases and describe the leveraging the concept and terminology and framework.

We have 6 months now in front of us to develop this work before next plenary meeting in San Francisco – US to build technical work and discuss it in new resolutions.