00:00 Ben Brumfield introduces the webinar
00:22 Jon Ippolito introduces IMPACT RISK framework
05:08 Quiz: what uses more energy and water?
08:38 AI champions versus critics
09:18 Misleading claim #1: LLMs are parrots
11:54 Misleading claim #2: AI's impact is a known quantity
13:22 Misleading claim #3: Massive water per prompt
15:15 Misleading claim #4: Negligible water per prompt
19:21 Misleading claim #5: Negligible electricity per prompt
20:15 Misleading claim #6: Low per-query footprint = low impact
24:38 Misleading claim #7: Focusing AI AI distracts from serious impacts
26:10 What's the difference between consumptive and non-consumptive water use? (Elizabeth Russey Roke)
29:49 What is a token? (Veronica Nargi)
33:33 How does AI's footprint compare to other work-related digital tasks?
36:35 How can we minimize our AI footprint?
39:23 Tools to assess your AI footprint
40:51 Integrating footprints and recommendations into chatbot interfaces (Greg Nelson)
46:31 "What Uses More" challenge
47:46 Will CollaborAIte be available to the public?
50:12 Student motivations (Ethan Morin)
50:58 Less impact running on local machines? (Anuj Gupta)
53:00 Can we track sources and timeframes for energy efficiency? (Ben Brumfield)
54:57 Have these tools been used in the classroom? (Brumfield)
56:34 Data centers powered by nuclear for in space? (Gupta)
58:06 Could the excess heat from data centers be used for useful heat, eg water? (Gupta)
59:07 Which researchers with this specialty do you recommend? (Gupta)
This teleconference is a project of the University of Maine's Digital Curation program. For more information, contact ude.eniam@otiloppij.
Timecodes are in minutes: seconds
Is it fair to compare AI’s footprint to Netflix binges or burgers?
In this webinar organized by Sara and Ben Brumfield of FromThePage in conjunction with the University of Maine's Digital Curation program on 2 April 2026, New Media professor and Digital Curation program director Jon Ippolito reviews the science behind these comparisons and applies it to determine which activities and techniques use the most or least electricity and water. The webinar also reviews the critical difference between global and local impacts, and why numbers alone don’t always tell the full story.
AI’s footprint is often compared to frivolous or leisure activities like watching movies or scrolling TikTok. This time we’ll also compare it to tasks relevant for people who work with media like archivists and educators, from digitizing images to meeting via Zoom to uploading videos.
Computer scientist and special guest Greg Nelson joins the webinar with student Ethan Morin to demo a new chatbot that integrates AI footprints to help individuals and organizations understand their environmental impact, based the What Uses More app. Nelson's app, CollaborAIte, also features recommendations for ways users can offset AI use by abstaining from other carbon-intensive activities.
Watch the entire video, choose an excerpt from the menu on this page, or view the slides from the talk.
Or view more teleconferences from the Digital Curation program.