What are the environmental and social costs?

What are the environmental and social costs?
What are the environmental and social costs?

Generative artificial intelligence (generative AI) is booming, transforming key sectors like healthcare, education and creativity. However, behind its promises of innovation lies a much less rosy reality: its ecological and social impact. A recent study, published in Environmental Science and Ecotechnology by researchers from Northwestern University, Harvard and the University of Texas at San Antonio, sheds light on the hidden costs of this technology. It calls for concrete actions to guide its development in an ethical and sustainable manner.

A worrying environmental cost of generative AI

The development of generative AI relies on a resource-intensive technological infrastructure. Hardware components, such as graphics processing units (GPUs) and data centers, require rare and expensive materials to mine. The mining of metals such as cobalt and tantalum has serious consequences: deforestation, water pollution and soil impoverishment.

The energy impact of data centers

Data centers, at the heart of AI systems, consume considerable amounts of electricity. According to forecasts, they could represent more than 8% of total U.S. energy consumption by 2030. This phenomenon increases pressure on electricity networks and worsens greenhouse gas emissions, hampering global efforts for the energy transition.

The explosion of electronic waste

As AI systems evolve, older hardware becomes obsolete, generating an increasing amount of e-waste. This e-waste, often poorly recycled, aggravates global pollution problems, contaminating soil and water with toxic substances.

Social inequalities reinforced with generative AI

The social impact of generative AI is not limited to its use. Its production and deployment process is based on practices that perpetuate injustices and inequalities.

Precarious working conditions

The production of components necessary for AI relies on the exploitation of workers, particularly in cobalt mines. Reports indicate the presence of children in these mines, exposed to dangerous conditions for paltry wages. Moreover, workers responsible for training AI systems – often located in developing countries – are also poorly paid and subject to precarious contracts.

An increased digital divide

Access to generative AI remains uneven. Industrialized nations and speakers of majority languages, such as English, enjoy a considerable advantagewhile marginalized communities struggle to access these technologies. This imbalance risks widening the global digital divide, leaving millions of people behind.

Generative Artificial Intelligence: Why are French VSEs and SMEs hesitant to take the plunge?

Solutions for a responsible future

Faced with these challenges, researchers are calling for immediate measures to limit the negative impacts of generative AI. Among their recommendations are:

  • More energy efficient workouts to reduce the ecological footprint of AI models.
  • Sustainable hardware design aimed at limiting dependence on rare metals and extending the lifespan of equipment.
  • Improved working conditions in supply chains, with enhanced protections for workers.
  • Inclusive governance frameworksguaranteeing equitable accessibility and a fair distribution of benefits.

The authors also emphasize the need for transparency on the part of developers and decision-makers. They recommend the establishment of mandatory reporting of environmental and social impacts related to generative AI.

A collective responsibility

As Mohammad Hosseini, lead author of the study, noted, “this research reveals the hidden costs of generative AI and calls for collective action.” The challenge now consists of reconciling technological innovation, respect for human rights and preservation of the environment. Artificial intelligence can only truly serve the common good by adopting ethical and sustainable practices.

Article source: http://dx.doi.org/10.1016/j.ese.2024.100520

-

-

PREV Obesity: Lonza is involved in the development of a nasal spray
NEXT Pension funds: a colossal gain, double AVS pensions!