Learning Outcomes:

Topics:

Introduction to AI Ethics

Course Coverage

• Overview of Ethical Challenges in AI
  • Autonomy and AI Decision-Making
  • Privacy Concerns in AI Applications
  • Accountability in AI Systems
  • Explainability in Large Language Models (LLMs) and Its Importance in Modern AI Systems
• Key Philosophical Concepts in AI Ethics
  • Utilitarianism and AI
  • Deontological Ethics in AI Decision-Making (actions vs. consequences)
  • Virtue Ethics and AI Behavior 

The Alignment Problem in AI

Course Coverage:

• Understanding the Alignment Problem
  • Definition and Importance of AI Alignment, particularly with LLMs
  • Challenges in Aligning AI with Human Values
• Approaches to Addressing the Alignment Problem
  • Value Loading in AI Systems
  • Human-in-the-Loop Approaches
  • AI Governance and Policy Considerations

Attribution of Responsibility in AI Systems

Course Coverage:

• The Problem of Responsibility in Autonomous AI
  • Defining Responsibility in AI Contexts
  • Legal and Ethical Implications
• Models for Assigning Responsibility
  • Responsibility Attribution Models
  • Case Studies: Responsibility in Autonomous Vehicles
• Debates on AI and Moral Agency
  • Can AI Be Considered a Moral Agent?
  • Philosophical Perspectives on AI Autonomy

Philosophical Debates in AI Safety

Course Coverage:

• Machine Consciousness and Moral Status
  • What is Machine Consciousness?
  • Ethical Implications of AI Consciousness
• AI Safety and the Precautionary Principle
  • Applying the Precautionary Principle to AI
  • Case Studies: AI Safety in HighStakes Applications
• Long-Term Considerations in AI Ethics
  • The Future of AI and Human Coexistence
  • Ethical Challenges in Superintelligent AI

Debates on AI Alignment

Course Coverage:

• Key Arguments in the AI Alignment Debate
  • Proponents of Strong AI Alignment
  • Criticisms and Counterarguments
• Ensuring AI Aligns with Human Values
  • Techniques for Value Alignment
  • Case Studies: Successful AI Alignment Examples

Evaluating AI Safety Frameworks

Course Coverage

• Overview of Existing AI Safety Frameworks
  • Analysis of Popular AI Safety Methodologies
  • Strengths and Weaknesses of Current
  • Frameworks
• Case Studies in AI Safety
  • Implementation
  • AI Safety in Healthcare
  • AI Safety in Autonomous Systems

Global Collaboration in AI Safety

Course Coverage:

•  The Role of International Collaboration
  • Importance of Cross-Border Cooperation
  • Case Studies: International AI Safety Initiatives
• Challenges in Global AI Safety Efforts
  • Regulatory Differences Across Countries
  • Ensuring Consistency in AI Safety Standards

Understanding Bias in AI

Course Coverage:

• Types of Bias in AI Systems
  • Data Bias and Its Sources
  • Algorithmic Bias: Causes and Effects
  • Societal Bias Reflected in AI Outcomes
• Identifying Bias in AI Applications
  • Common Techniques for Bias Detection
  • Case Studies: Bias in Facial Recognition Systems

Metrics for Measuring Bias in AI

Course Coverage:

• Overview of Bias Metrics
  • Disparate Impact Ratio
  •  Equal Opportunity Difference
  • Demographic Parity
• Applying Bias Metrics in Practice
  • Practical Examples of Metric Application
  • Case Studies: Measuring Bias in Hiring Algorithms

Impact of Bias on AI Decision-Making

Course Coverage:

• Consequences of Bias in AI Systems
  • Legal and Ethical Implications
  • Impact on Marginalized Groups
• Strategies for Mitigating Bias in AI
  • Pre-Processing Techniques
  • In-Processing Adjustments
  • Post-Processing Corrections

Developing a Bias Mitigation Strategy

Course Coverage:

• Steps for Addressing Bias in AI
  • Identifying Bias Sources
  • Selecting Appropriate Mitigation Techniques
• Implementing Bias Mitigation in AI Projects
  • Case Studies: Bias Mitigation in Financial Services
  • Best Practices for Bias Reduction

Introduction to Algorithmic Fairness

Course Coverage:

• Concepts of Fairness in AI
  • Individual Fairness vs. Group Fairness
  • Causal Fairness in AI Systems
• Approaches to Algorithmic Fairness
  • Pre-Processing, In-Processing, and Post Processing Methods
  • Trade-Offs Between Fairness and Accuracy

Accuracy vs. Fairness in AI Models

Course Coverage:

• Understanding the Trade-Off
  • Balancing Model Performance with Ethical Fairness
  • Case Studies: Accuracy vs. Fairness in Lending Algorithms
• Practical Implications of the Trade-Off
  • Impact on Decision-Making Processes
  • Managing Stakeholder Expectations

Implementing Fairness Measures in AI Models

Course Coverage:

• Using Python Libraries for Fairness
  • Introduction to Fairlearn, AIF360, and Other Libraries
  • Practical Exercises: Applying Fairness Libraries to Real Data
• Evaluating the Effectiveness of Fairness Interventions
  • Comparing Fairness Metrics Preand Post-Intervention
  • Case Studies: Fairness in Healthcare AI Systems

Real-World Applications of Algorithmic Fairness

Course Coverage:

• Case Studies in Fairness Implementation
  • Examples from Different Industries: Finance, Healthcare, Law Enforcement
• Lessons Learned from Fairness Implementation
  •  Challenges and Solutions in Real-World Applications
  • Best Practices for Ensuring Fairness in AI

The Need for Explainable AI (XAI)

Course Coverage:

• Importance of Transparency in AI
  • Why Explainability Matters: Trust and Accountability
  • Regulatory Requirements for Explainability
• Key Concepts in Explainable AI
  • Transparency, Interpretability, and Justification
  • Challenges in Achieving Explainability

Comparing XAI Techniques

Course Coverage:

• Overview of Popular XAI Methods
  • Local Interpretable ModelAgnostic Explanations (LIME)
  • SHapley Additive exPlanations (SHAP)
  • Model-Specific vs. ModelAgnostic Approaches
• Strengths and Weaknesses of XAI Techniques
  • Performance Across Different AI Tasks
  • Case Studies: XAI in Medical Diagnosis

Evaluating Robustness in XAI

Course Coverage:

• Robustness of Explanations
  • Assessing the Reliability of Explanations in Various Contexts
  • Case Studies: Robustness in HighStakes AI Applications
• Challenges in XAI Robustness
  • Adversarial Attacks on Explainability
  • Maintaining Explainability in Complex Models

Practical Implementation of XAI Techniques

Course Coverage:

• Implementing XAI in Python
  • Using LIME, SHAP, and ELI5 for Explanation Tasks
  • Practical Exercises: Applying XAI to Classification Models
• Evaluating XAI Implementations
  • Comparing the Quality of Explanations Generated by Different Methods
  • Case Studies: XAI in Financial Services

Introduction to XAI Libraries

Course Coverage:

• Overview of Python Libraries for XAI
  • LIME, SHAP, ELI5, and Other Tools
  • Installation and Setup Guides
• Choosing the Right Library for the Task
  • Comparing Capabilities and Use Cases of XAI Libraries
  • Best Practices for Library Selection

Building AI Models and Applying XAI Techniques

Course Coverage:

• Developing a Simple AI Model
  • Building Classification and Regression Models in Python
  • Practical Exercises: Model Development
• Applying XAI Techniques
  • Using LIME for Local Explanations
  • Using SHAP for Global Interpretability

Evaluating the Quality of XAI Explanations

Course Coverage:

• Metrics for Assessing Explanation Quality
  • Fidelity, Interpretability, and Stability Metrics
  • Case Studies: Evaluating Explanations in AI Applications
• Practical Evaluation of XAI Implementations
  • Hands-On Assessment of XAI Libraries
  • Reporting Findings and Insights

Presenting XAI Findings

Course Coverage:

• Preparing Reports and Presentations
  • Structuring XAI Reports for Stakeholders
  • Visualizing Explanations and Their Implications
• Case Studies in XAI Presentation
  • Examples of Effective Communication of XAI Results
  • Lessons Learned from Industry Applications

Indicative reading list 

• Floridi, L. (2014). The Ethics of Information. Oxford University Press.

• Russell, S., & Norvig, P. (2020). Artificial Intelligence: A Modern Approach. Pearson.

• O’Neil, C. (2016). Weapons of Math Destruction: How Big Data Increases Inequality and Threatens Democracy. Crown.

• Binns, R. (2018). Fairness in Machine Learning: Lessons from Political Philosophy. arXiv:1712.03586

• Doshi-Velez, F., & Kim, B. (2017). Towards A Rigorous Science of Interpretable Machine Learning. arXiv:1702.08608

• Miller, T. (2019). Explanation in Artificial Intelligence: Insights from the Social Sciences. Artificial Intelligence.

• Lepri, B., Oliver, N., Letouze, E., Pentland, A., & Vinck, P. (2018). Fair, Transparent, and Accountable Algorithmic Decision-making Processes. Philosophy & Technology.

• Barocas, S., Hardt, M., & Narayanan, A. (2019). Fairness and Machine Learning. MIT Press.

• Lipton, Z. C. (2018). The Mythos of Model Interpretability. Queue.

• Mitchell, M. (2019). Artificial Intelligence: A Guide for Thinking Humans. Farrar, Straus and Giroux

• Fairness, Accountability, and Transparency in Machine Learning (FAT/ML): www.fatml.org

Entry requirements

To enroll The Master Award, the learner must possess:

• Graduated with a Bachelor’s degree from an accredited university or achieved a Level 6 Diploma according to the European Qualifications
• For a degree from non-recognized universities; The learner should have followed Accreditation of Prior Experiential Learning for Qualifications (APEL.Q) policy of SIMI and/or University Partners.
• Learners must be over 21 years old.

The SIMI Swiss reserves the highest decision-making power for admission whether to accept or not accept after a specific review of each candidate’s profile to ensure that they can comprehend and gain benefits when participating. For the fake university or diploma mills, University Partners shall not be accepted.

English requirements

If a learner is not from a predominantly English-speaking country, proof of English language proficiency must be provided.

• Common European Framework of Reference (CEFR) level B2 or equivalent
• Or A minimum TOEFL score of 101 or IELTS 6.5; Reading and Writing must be at 6.5 or equivalent

After graduating with Master Award, students receive all certified documents from the SIMI Swiss.

Certified Documents:

• e-Certificate from the Swiss Information and Management Institute (SIMI Swiss).
• Hard copy certificate from the Swiss Information and Management Institute (SIMI Swiss) – Optional.
• Accreditation of Prior Experiential Learning for Qualifications (APEL.Q) certified from SIMI Swiss for credit and tuition fee transfer.

Because the program is accredited and recognized, students can easily use certified in the working environment and have many opportunities for career advancement. In addition, in case if you want to study for a SIMI degree or university partner degree, students can convert all credits and the full paid tuition fee when participating in the program University Partners.

The SIMI Swiss’ Master Award means:

SIMI Swiss Master Award is the award at the master level and is equivalent to:

• Level 7 certificate of Regulated Qualification Framework (RQF) of UK
• Level 10 certificate of Scottish Credit and Qualifications Framework (SCQF)
• Level 7 certificate of Credit and Qualifications Framework (CQFW)
• Level 7 certificate of European Qualifications Framework (EQF)
• Level 9 certificates of the Australian Qualifications Framework (AQF)
• Level 7 certificate of ASEAN Qualifications Reference Framework (AQRF)
• Level 9 certificate of the African Continental Qualifications Framework (ACQF)

Students can convert all credits and the full tuition fee when participating in the SIMI Swiss and/or University Partners academic programs if they want to study for an academic degree.

Credits transfer:

Learners can accumulate 20 credits from the Master Award program when participating in the Master of Artificial Intelligence. Please see the credit transfer policy HERE

Tuition fee transfer:

When participating in the Master of Artificial Intelligence program, students who have graduated 1 Master Award will receive a discount of full tuition fee which you paid. Please see the tuition fee transfer HERE

The SIMI Swiss micro-credential program allows for the transfer of credits and tuition fees into full degree programs from SIMI Swiss and/or its university partners. SIMI Swiss reserves the right to limit admissions once the number of students exceeds the quotas.

Apply Policy:

• To participate in the SIMI Swiss micro-credential program, students need to meet the entry criteria corresponding to each level. Please see the “Entry” tab for more details.
• SIMI Swiss will not accept applicants if their entry qualifications are from diploma mill universities or schools/universities that are not accredited.
• For Master Award programs, if an entry bachelor is unavailable, students must demonstrate a minimum of 5 years of work experience in the relevant field. Please note that a bachelor’s degree is required for the Master’s program at SIMI Swiss and University Partners so that you could study Master Award but could not move to the Master’s program of SIMI and University Partners.
• English is not a mandatory entry requirement for short course programs, but candidates need to ensure that English is used in reading documents, listening to lectures, and doing assignments. Candidates should note that English is a mandatory requirement when switching to an academic program at SIMI Swiss and University Partners.

Apply Process:

1.  Choose the program that suits your requirements. Note that applicants without a university degree will not be able to participate in the program at Master’s level, and applicants without a Master’s degree will not be able to participate in the program at the Doctoral level.
2. Email your application to support@simiswiss.ch with all the required documents. You could download the application form here.
3. Our admission department will contact you and guide you through further processes if the registration documents need to be supplemented.
4. SIMI Swiss will issue the Letter of Acceptant (LOA). You wil proceed to the next steps according to the instructions and pay tuition fee.
5. SIMI Swiss will issue a student confirmation letter, login account to the e-learning system and related documents.
6. You have become an official SIMI Swiss student and enjoy your study journey.

The SIMI Swiss micro-credential program is fully online, allowing you to study anytime, anywhere. You have the option to attend live classes with SIMI Swiss. The final exam will be uploaded to the system and evaluated by the academic panel of SIMI Swiss. Students must submit assignments on time; failure to do so will result in the student being considered to have discontinued the program.