Exciting news! We at SPATIAL are proud to announce that our two online courses – Trustworthy AI and Advanced Trustworthy AI – ARE AVAILABLE AT THIS LINK.
If you are a researcher, student, industry expert or professional working on different types of businesses, from engineering to law, this course package is for you!
To cover different needs and types of knowledge about AI, from more experts to people with basic knowledge, we designed two courses: a basic course for knowledge workers in AI and a technical one dedicated to experts in the field.
In addition, whilst the modules work as standalone courses, they are designed to be a series worth 2 ECTS credits. This means universities can include these courses in their curricula!
Are you curious to discover more about these courses? Read the description provided by our colleagues from MinnaLearn, the leading team of this task.
Closing a “technical-societal gap” requires a product for each audience
The task of building a series of educational modules around the topic of trustworthy AI was set out to reach three main objectives:
- To disseminate the findings from the SPATIAL project,
- Help close the technical-societal gap by introducing socio-technical skills and ethical socio-legal awareness to the readers
- Teaching the fundamentals to future engineers and AI developers for building trustworthy AI solutions
To close a gap, you need to consider the two edges of a gap. Whilst the original idea was to create one modular course for engineering students, interviews with industry experts revealed the increasingly pressing need to get a basic understanding of trustworthy AI to other professionals, like business and team leaders, and lawyers: it’s hard to make teams working on AI solution diverse without a common language between professionals from differing backgrounds. Moreover, the upcoming legislation around AI in the EU will demand a wider audience to be updated on the regulation in the market.
We noted that one module would hardly serve both the technical and the more general audience: for the other, it would end up too technical, whilst the other audience would likely find the content too general and high-level. Thus, we decided to create two separate courses: an intro course into trustworthy AI and an advanced course for developers more familiar with machine learning. Both of the courses introduce the same topics in the same order but at different difficulty levels. Both of the courses consist of five chapters, with 2–4 sections in each chapter. The sections have 1–10 exercises for the reader to test their knowledge. The text-based content is balanced with inviting illustrations, and the course platform works on desktop and mobile devices.
What are the two courses about?
As described, Trustworthy AI is mostly about understanding the basics. The value proposition of this module is to allow its takers to engage in discussion about trustworthy AI both in their organizations and the society at large.
The course starts with the definitions and main concepts: what do we talk about when we talk about trustworthy AI? The perspectives of trustworthy AI are introduced via a machine learning pipeline to give the reader an overview of how trustworthiness is applied at the different stages of AI model development. Over the chapters, the reader is walked through the different steps of the trustworthy AI-building process more closely: detecting and mitigating bias, adding transparency to the AI models, and preparing them for disturbances and attacks.
Advanced Trustworthy AI continues from where the first course left off, but this time, it explains the technical details behind these steps in more detail. The reader will be given many practical examples with code previews on assessing and documenting bias in AI models, applying and analysing different explainability methods, and describing potential real-life security threat scenarios.
Trustworthy AI
Understand the importance, considerations, and impacts of trustworthy AI.
- Level: Intermediate
- Certificate: Available for purchase upon completion
- Target audience: Everyone working with AI products and services
- Scope: 1 ECTS credit
Advanced trustworthy AI
Learn the principles, methods, and tools for implementing trustworthy AI in practice.
- Level: Advanced
- Certificate: Available for purchase upon completion
- Target audience: Software developers, data scientists, engineering leads
- Scope: 1 ECTS credit
Stay tuned to our channels (Twitter, LinkedIn & YouTube) and subscribe to our to make sure to catch the release of our course next month!