Despite all the published discourse on the topic, thinking about AI has not been critical enough (Lingren, Handbook of Critical Studies of artificial Intelligence. Edward Elgar, 2023). Researchers and scholars are now working to rectify this. When thinking about AI and ecology, researchers have examined the positive potential for these new technological assemblages to help deal with many environmental challenges. Unlike this important body of work, this chapter will focus not on the capacity for AI to intervene in environmental problems but on the ecological impact of AI itself. AI does not sit outside the ecological systems that it might be thought to improve. The growth of communication systems and computational power are part of the “great acceleration” that began in the aftermath of the Second World War. Following Lopez (Ecomedia literacy: integrating ecology into media education. Routledge, 2021), this chapter will frame AI as an ecomedia object within the ecomediasphere. This chapter can be framed as a contribution to the “material turn” in ecomedia research. This chapter focuses its discussion on the ecomateriality of AI and three particular disruptions: energy, raw materials, and e-waste. First, the AI assemblage is placing increasing demands on the electricity supply, demands that will only increase. Second, the AI assemblage is a product and a process dependent on supplies of raw materials that are used to construct, operate, and maintain its infrastructure and components. The expansion of AI means a growth in demand for all these materials. Third, AI explosion has implications for the afterlife of its components. As it increasingly colonizes the lifeworld, and with constant upgrades and developments, AI has the potential to accelerate the pace of obsolescence and therefore increase the already complex problem of e-waste. This chapter makes it clear why any discussion about the future of AI must include consideration of the actually existing ecology of this transformative technology.

错误:搜索内容不能为空,请输入英文关键词
错误:关键词超出字数限制,请精简
高级检索

“Limitless Computation on a Finite Planet?” The Actually Existing Ecology of Generative AI

  • Ian Collinson

摘要

Despite all the published discourse on the topic, thinking about AI has not been critical enough (Lingren, Handbook of Critical Studies of artificial Intelligence. Edward Elgar, 2023). Researchers and scholars are now working to rectify this. When thinking about AI and ecology, researchers have examined the positive potential for these new technological assemblages to help deal with many environmental challenges. Unlike this important body of work, this chapter will focus not on the capacity for AI to intervene in environmental problems but on the ecological impact of AI itself. AI does not sit outside the ecological systems that it might be thought to improve. The growth of communication systems and computational power are part of the “great acceleration” that began in the aftermath of the Second World War. Following Lopez (Ecomedia literacy: integrating ecology into media education. Routledge, 2021), this chapter will frame AI as an ecomedia object within the ecomediasphere. This chapter can be framed as a contribution to the “material turn” in ecomedia research. This chapter focuses its discussion on the ecomateriality of AI and three particular disruptions: energy, raw materials, and e-waste. First, the AI assemblage is placing increasing demands on the electricity supply, demands that will only increase. Second, the AI assemblage is a product and a process dependent on supplies of raw materials that are used to construct, operate, and maintain its infrastructure and components. The expansion of AI means a growth in demand for all these materials. Third, AI explosion has implications for the afterlife of its components. As it increasingly colonizes the lifeworld, and with constant upgrades and developments, AI has the potential to accelerate the pace of obsolescence and therefore increase the already complex problem of e-waste. This chapter makes it clear why any discussion about the future of AI must include consideration of the actually existing ecology of this transformative technology.