15. AI Flashcards
AI
simulation of human intelligence in
machines designed to think and learn like humans, performing tasks
such as problem-solving, decision-making, and data analysis.
Machine learning
subset of AI where
algorithms learn patterns from data and improve performance over
time without being explicitly programmed
EIA + AI challenges (6)
ethical, technical, resourced-based
- Data collection: Volume and variety
–> public involvement may be compromised (social surveys, interviews,…) - Analysis: Complexity and time constraints
- Predictive accuracy: Managing uncertainties
–> data quality issues can = inacurrate models
–>black box syndrome - Public participation: Engaging diverse stakeholders
–> ethical concerns, privacy, accountability, equity - over reliance in AI without human oversight
- infrastructure gaps in developing regions
AI applications in EIA (2)
- Automated Data Analysis
* Processing large datasets (e.g., satellite imagery, environmental sensors, drones)
* Real-time monitoring of ecosystems, biodiversity
* air/water quality analysis using ML and sensors - Predictive Modeling
* Machine learning for predicting environmental impacts (CC)
* natural disaster risk prediction and mitigation planning
Benefits of AI in EIA
- Enhanced precision and reduced human errors. accuracy
- Real-time decision-making using continuous data updates.
- Cost-effective solutions for long-term monitoring.
efficiency - Improved capacity to predict and mitigate future impacts.
predictive capabilities - Scalability to analyze large datasets efficiently.
- Inclusion.
black box syndrome
Algorithmic bias creating skewed results.
enviro and social costs of AI demanding proactive mitigation strategies
- High energy consumption to run AI.
- Significant carbon emissions from data centers.
- Resource mining for GPUs and AI hardware.
- Growing e-waste from outdated systems. disposal challenges
- Landscape change and biodiversity loss due to mining and other activities.
- Water use for cooling systems in data centers.
- Global inequality and resource-intensive AI development.
mitigation strategies
- Transition to renewable energy for data centers.
- Invest in energy-efficient AI algorithms and models.
- Develop circular economy practices for hardware reuse and recycling.
- Implement global regulations for ethical AI development.
- Foster international cooperation to share sustainable AI technologies. collaboration = essential for sustainable AI application
key things to do to leveraging AI’s full
potential while minimizing risks (going forward)
need to consider how to integrate AI responsibly
into EIA practices.
–>should focus on developing transparent AI models that
stakeholders can understand and trust.
–>prioritize data equity, ensuring that all
communities have access to high-quality datasets and tools.
–>fostering collaboration between technologists, environmental
scientists, and policymakers
AI to assess health impacts
- AI models predict exposure to pollutants and health outcomes.
- Example: Estimating respiratory risks from air quality changes due to industrial projects.
AI to assess social impacts
- ML analyzes displacement risks and community sentiment through social media data.
- Example: evaluating the impacts of a new subway line on communities along the route.
AI to assess economic impacts
- AI forecasts job creation, income changes, and local economic development.
- Example: Evaluating the economic multipliers from renewable energy installations.
