Impact of AI
Broad application: AI can (and already has) transformed various aspects of life, including automation, decision-making, entertainment and education.
Transformation: The broad application of AI can enhance research efficiency, accessibility, streamline tasks and personalize learning.
Considerations: The rise of AI brings ethical, environmental and societal concerns that must be evaluated.
Limitations of AI
- Inaccuracies: AI outputs can be inaccurate, biased, fabricated or contain copyright-protected material.
- Hallucinations: AI can sometimes make errors or "hallucinate" content that is incorrect.
- No original thought: AI does not "think" like humans.
This is why human review and interaction is important for accuracy.
Risks & Rewards of AI in Higher Education
Video Highlights (2:40)
- Artificial intelligence could help solve some of higher education’s most persistent problems, but AI may not always be the right solution.
- Colleges and universities would be best served using AI in low-risk situations at first.
- Inputting bad data into an AI tool will produce bad outcomes and recommendations that, if followed, could lead to bad decisions.
Source: EdTech Magazine (2024)
Explore the benefits and impact while also reducing risk and adverse outcomes
Moral considerations and consequences arising from actions, decisions or research, including issues such as bias, fairness and harm, requires careful consideration to ensure responsible and ethical conduct when using AI.
Environmental Impact & Sustainability
While AI can offer opportunities for social and economic progress, its environmental effects are complex.
(Sources: Penn State Institute of Energy & the Environment, United Nations, World Economic Forum)
Positive Aspects:
- Facilitating energy transition: AI tools can help optimize energy use and integrate renewable energy sources more effectively, aiding the transition to a more sustainable energy system.
- Efficiency gains and cost reduction: AI can improve the efficiency of various processes across industries, potentially reducing overall energy consumption in the long run.
- For example, predictive maintenance powered by AI can reduce downtime and costs in a multitude of industries, leading to more efficient use of resources.
- For example, predictive maintenance powered by AI can reduce downtime and costs in a multitude of industries, leading to more efficient use of resources.
- Optimization of energy systems: AI can optimize energy systems by predicting demand and managing supply more efficiently.
- Predictive analysis: AI algorithms can identify patterns in data, detect anomalies and anticipate and forecast future results.
Negative Aspects:
- Increased energy demand and power use: The computing power required to run AI models, especially generative AI, is significant. Machine learning and AI accounted for less than 0.2 percent of global electricity demand and less than 0.1 percent of global GHG emissions in 2021. However, the demand for AI computing is increasing rapidly.
- The electricity demand from the sector could reach 1,000 terawatt hours (TWh) by 2026, up from 460 TWh today.
- With 100 million users on ChatGPT every week, the cumulative energy demand is growing.
- Rising emissions: Companies have reported significant increases in their CO2 emissions due to the expansion of data centers needed to support AI operations.
- Water Usage: AI's water footprint is also considerable. Data centers require large amounts of water for cooling.
Environmental impacts vs benefits
Vice President and Principal Analyst at Forrester, Thomas Husson, discusses the various sides of this issue (26:32).
He emphasizes that we shouldn’t think of any specific technology as good or bad for the environment, but rather consider the impacts of specific use cases for the technology.
The conversation covers the negative impacts of AI on the environment but also highlights the positive role AI can play in improving sustainability and tackling climate challenges (26:33).
Source: Forrester (2024)