Navigating the Digital World

• Explores navigating the digital world through computational thinking, quantitative reasoning, cybersecurity, digital misinformation, ethics, intellectual property, blockchain, and artificial intelligence

Computational Thinking

• Thought processes for problem-solving in a way that computers/humans/machines can execute.
• Involves decomposition, pattern recognition, abstraction, and algorithms.

Decomposition

• Breaking down complex problems/data/processes into smaller, manageable parts.
• Biology example: Decomposition of organic material into fragmentation for fungi and bacteria.
• Computer science examples: Functions and factorials.
• Real-world example: Moving houses which breaks down the moving process into a list of smaller tasks.
• Organisational example: Companies break down responsibilities by departments

Pattern Recognition

• Involves identifying patterns or trends within a problem.
• Biology example: Protein synthesis, where proteins are coded by mRNA.
• Computer science examples: Machine learning and AI.
• Real-world examples: Games (Uno, tic-tac-toe etc.)

Abstraction

• Identifying similarities/differences among similar problems to simplify solutions or identifying the structure of concepts and ideas to simplify.
• Biology example: DNA double helix simplified to assigning ATCG names.
• Real-world examples: MRT maps showing only essential information

Algorithm

• Developing step-by-step instructions for solving a problem applicable to similar problems.
• Biology example: Transcription and translation where DNA sequences transcribe mRNA into amino acids.
• Real-world examples: Recipes, instruction manuals, and legislation.

Quantitative Reasoning

• Applying mathematical skills to analyse real-world information within a discipline.
• Involves framing numerical questions, identifying tools/data, building models, and analysing results.

Drug Data Distribution & Comparisons

• Line plots connect data points.
• Histograms count frequency across specific bins for continuous data.
• Comparing distributions involves idealising them with mean and standard deviation.
• Mean represents the average behaviour.
• Standard Deviation (SD) represents data point deviation from the average.
• Hypothesis testing compares actual vs hypothesised means using a Z-test.

House Data Distribution & Comparisons

• Inaccurate price estimates result from naive (generic) estimation.
• Feature-dependent estimation relies on joint plots to check distribution strength via linear proportionality.
• Wider scatter plots indicate larger standard error.
• Correlation (1 to -1) indicates the linear relationship between two variables; values closer to 1 or -1 have smaller margins of error.
• Linear models estimate one variable using another (y = mx + c).
• Assuming a normal distribution, 68% of data points fall within one standard deviation from the mean.

Cybersecurity

• Technologies, processes, and practices protecting devices, programs, and networks from damage, attacks, and unauthorized access.
• Phishing prevention: check sender/hyperlinks, use HTTPS, and delete suspicious emails.
• Protect personal information
• Protect website integrity with good cyber hygiene.
• Prevent device lockouts.

Strong Passwords

• At least 8 characters.
• Use a codified phrase with mixed upper/lowercase and symbols.
• Avoid personal information.
• Enable two-factor authentication (OTP via phone/email).
• Use different passwords across accounts.

Data Security

• Data exists in digital and physical forms belonging to individuals or corporations.
• Prevention: lock terminals, avoid leaving files on desks, do not use personal accounts for organizational data, and secure data storage devices.
• CSA is the Cyber Security Agency.
• Classifications: Open (publicly distributable), Restricted (accessible to organization members), Confidential (contractually defined or sensitive), and Classified (under Official Secrets Act).
• Unauthorized disclosure of confidential/classified data may result in penalties/damage to national security.

Digital Misinformation

• Types of falsehood: misinformation (inadvertent), disinformation (intentional), and fake news (deceptive news-like content).

Motivation for Creating Fake News

• Financial gain through clicks and advertising revenue.
• Pushing a personal agenda or ideology.
• Fake news can be a weapon of mass information.

Consequences of Fake News

• Short term: Affects political decisions, business, and peace and order.
• Long term: Devalues information, erodes trust, and increases social divisions.

Vulnerabilities to Fake News

• Sender/source credibility based on trust with original, invisible, trusted, or disregarded sources.
• Message plausibility, expert references, conversational tone, emotional appeal, and calls to action.
• Channel factors: popularity cues, reliance, lack of gatekeeping, and information overload.
• Receiver characteristics: information behaviour, confirmation biases, and motivations for sharing sensational news.
• Higher social media news use increases belief in fake news; those avoiding news are also susceptible.
• Informational apathy can lead to further spread.

Fighting Fake News

Individual Authentication of Information

• Internal: Self-gut feeling, source evaluation, message analysis.
• External: Interpersonal (family/friends), Institutional (official websites, mainstream news).
• Authentication is a social process driven by self-image and group cohesion.

Actions Individuals Can Take

• Reflect on information behaviour (sources, reliability, and motivations for sharing).
• Engage with content critically.
• Strive to understand others' perspectives.
• Support reliable information sources.
• Maximise available resources.
• Equip ourselves with knowledge

Government Prevention of Fake News

• Legislation like POFMA allows ministers to issue directions regarding online falsehoods within the public interest (even posts outside Singapore).

Tech Companies Helping Prevent Fake News

• Gamifying interventions.
• Supporting third-party fact-checkers and journalists.
• Promoting media literacy.
• Reducing financial incentives for content producers.
• Implementing features to flag content.
• Deleting posts and content.

Journalists & Fact Checkers

• Verification (pre-publication) vs Fact-checking (post-publication).
• Types of fact-checkers: affiliated with news orgs, government-owned, independent, volunteer groups, or individuals.
• Using fact-checking tools to monitor trends and verify images and sites.
• Video and other interesting mediums
• Truth sandwich is effective.

Ethics

• Study of morality, vital for developing and using digital/online technologies.
• Aim to avoid harms while maximizing benefits.

Normative Ethics

• Moral theories that explain right and wrong actions while offering practical guidance.
• Utilitarianism: Actions are morally right when they produce the most welfare.
• Virtue ethics: Actions are morally right when a virtuous person would do them.
• Kant’s deontological ethics: Actions are morally right when treating individuals as ends, not merely as means.
• Moral theories aim to provide general explanations and guidance.
• SCS is the Singapore Computer Society.

Cyberbullying

• Using the internet/digital devices to inflict psychological harm.
• Examples: hurtful messages, spreading lies, revealing personal information, posting embarrassing photos, impersonation, threats.
• Effects: Depression, anxiety, low self-esteem, sleep/eating disorders, suicidal thoughts.

Cyberbullying Through Moral Theories

• Utilitarianism: Morally wrong as it makes society worse off.
• Virtue Ethics: Cruel, and a contrast to virtue thus morally wrong actions.
• Kant’s Ethics: Lacks respect for victims and treats them like objects.

Informational Privacy

• Digital/online technologies impact confidentiality, anonymity, data protection, and secrecy of facts.
• A right to privacy is recognised in international human rights instruments.

Whistle Blowing

• Unauthorized disclosure of harmful information after internal attempts fail.
• Morally permissible when the firm will do serious harm, employees report concerns, and internal procedures are exhausted.
• Morally obligatory when documented evidence exists and public disclosure is likely to bring necessary changes.
• Objections: Criteria can be too stringent or not demanding enough.

Intellectual Property (IP)

• Creations from human intellect protected by legal rights.
• Motivates creators, encourages innovation, allows commercial exploitation, and defends against infringement.

Different Types of IP

• Patents, Copyright, Trademarks, Confidential Information
• Copyright protects the expression of an idea, not the idea itself.
• Copyright arises automatically if creator is Singapore citizen or PR
• Artistic merit don't matter, Independent creation does matter
• Copyright NOT for discoveries, procedures, tangible forms

Exclusive Rights in Copyright

• Literary, dramatic, musical, sound, film, artistic, broadcasts, and cable programmes.
• Purchasing a product does not grant underlying copyright.
• The author initially owns copyright except in cases of employment (employer owns) or by agreement.

Joint Ownership

• Where more than one author creates inseparable or interdependent parts.

Contract

• Agreement creating legally enforceable obligations.
• Identifies rights and obligations, allocates risk, and sets performance standards.

Elements of Contract

• Offer, Acceptance, Consideration, Intention to create legal relations, and Capacity.

Contractual Terms and Performance

• Terms set out rights and obligations regarding performance and risk allocation.
• Common terms: purpose, payment, rights, duration, warranties, and dispute resolution.

Breach

• Breach occurs when a term is not performed, entitling the wronged party to demand cure/compensation.
• IP can be bought, sold, given away, licensed, or assigned.

License

• License grants permission to perform an act that would otherwise infringe upon IP, commonly requires a fee.
• Non-exclusive license which is granted to more than one person. Exclusive license is granted to one person only. Used on social media, SaaS, media aggregation platforms.

Assignment

• Transfers all ownership rights to the assignee.
• Grants someone else the right to use the IP Transfers the entire title and interest in someone’s IP to another

Blockchain

• Public database using technologies for data exchange, processing, and storage.
• Tamper proof
• Types: Public (permissionless), Consortium (different access rights), and Private (permissioned for closed networks).
• Bitcoin blockchain records Merkle tree root, nonce, timestamp, and hash value of previous block.
• Opportunities: Improving transparency, lowering costs, and adding security.
• Challenges: Technology risk, scalability issues, and regulatory concerns.

Artificial Intelligence (AI)

• Machines performing tasks requiring human intelligence.
• Renaissance due to powerful computers, big data, and software algorithms.

AI Implementations

• Rule-based expert systems/fuzzy logic.
• Machine learning: Supervised, unsupervised, and reinforcement learning.
• Deep learning: Artificial Neural, Convolution Neural, Recurrent Neural, Deep Reinforcement, Transfer Networks
• Deep learning is based on artificial neurons which needs data to develop its algorithm.

Neural Network

• Neural Network mimics the human brain and recognises patterns (ability to learn) and includes
• Input layer
• Hidden layer consists of learnable parameters, the 'algorithm' that can learn and improve by itself
• Output layer
• Deep Neural Network has multiple hidden layers for sophisticated algorithms

Applications of AI

• Robots, Automotive, Social Media, Consumer Electronics, Business & E-commerce, Banking & Finance, Healthcare, Farming & Precision Agriculture, and Education.

AI Concerns

• Job Loss, Misuse, Explainability, Bias, and Ethics in Decision-Making.
• Current AI is considered Artificial Narrow Intelligence; the goal is Artificial General Intelligence.