Impact of Computing: Understanding Tradeoffs, Equity, and Fairness

Computing (in AP CSP context)

A set of technologies and processes (hardware, software, data, networks, and algorithms) that shape how information is created, stored, analyzed, and shared.

Beneficial Effect (of a computing innovation)

An outcome that improves well-being, increases capability, saves time/resources, expands access, or creates new opportunities.

Harmful Effect (of a computing innovation)

An outcome that causes damage, increases risk, reduces autonomy, or creates unfairness or exclusion.

Tradeoff (in computing impacts)

The idea that a computing innovation can create both benefits and harms at the same time, often affecting different groups differently or changing by context.

Big Idea 5 (Impact of Computing)

An AP CSP focus area that evaluates how computing innovations affect people and society, beyond whether the technology works.

Stakeholder

Any person or group affected by a computing innovation (e.g., individuals, communities, organizations, governments/society).

Mechanism (impact explanation chain)

A step-by-step causal link from a feature (like data collection or automation) to behavior/system changes and then to a real-world benefit or harm for stakeholders.

Data Collection

A capability where systems gather information about users or environments (e.g., location, clicks, contacts, biometrics), which can enable services but also create risks.

Automation

Using computing to perform tasks with minimal human intervention, often improving speed/consistency but potentially causing job displacement or unfair automated decisions.

Personalization

Tailoring content or services to an individual using data, which can improve relevance but can also enable manipulation or create filter bubbles.

Connectivity

The ability of devices and systems to communicate over networks, enabling collaboration and access but increasing exposure to attacks and rapid spread of information.

Data-Driven Insights

Patterns or conclusions derived from analyzing large datasets (e.g., using sensor data to improve traffic safety), which can be useful but are not automatically objective.

Computing Bias

Systematically unfair outcomes produced by a computing system for certain individuals or groups, often arising from data and human design choices rather than intentional discrimination.

Privacy

Control over personal information—who collects it, what is stored, and how it is used or shared.

Security

Protection of systems and data from unauthorized access or damage (e.g., via patching, training, backups); related to privacy but not the same.

Attack Surface

The number of possible points where a system can be attacked or fail; often increases as connectivity and complexity increase.

Misinformation

False or misleading information that can spread rapidly through computing systems, especially when platforms prioritize engagement.

Recommendation Algorithm

A system that suggests content to users; can improve discovery but can also amplify sensational or extreme content if optimized for engagement.

Engagement Optimization

Designing systems to maximize metrics like clicks or watch time, which can unintentionally promote sensational content and manipulation.

Economic and Labor Impacts (of computing)

Changes to jobs and working conditions caused by automation and platforms (e.g., efficiency gains for businesses vs unstable hours or displacement for workers).

Environmental Impacts (of computing)

Costs from manufacturing devices, powering data centers, and e-waste; small individual actions can add up to significant aggregate energy and hardware impacts.

Digital Divide

The gap between people who have effective access to computing/internet and those who do not; includes more than just owning a device.

Layers of Access (Digital Divide)

Dimensions of access: physical (devices/electricity), connection (speed/stability), economic (affordability), skill (digital literacy), and accessible design (works for disabilities).

Proxy Variable (in computing bias)

A measured variable used to stand in for something else (e.g., using attendance as a proxy for job performance), which can unfairly penalize groups due to correlated factors.

Feedback Loop (in computing systems)

When a system’s outputs change user behavior and the environment, shaping future data so the system increasingly reinforces the same patterns (e.g., clicks → more recommendations → more clicks).