This is a community-written summary of one of our CAT Salon events. This CAT Salon was inspired by XR’s #BigTechLovesBigOil campaign & Greenpeace’s Oil in the Cloud report. Find the slide deck used in the talk here.
Find the recording of this talk here
Speaker
CAT member Roelof Pieters is CTO and co-founder of overstory.ai. He has a background in Computer Science & AI, Social/Development Anthropology. Roelof also heads up Sunshine Lab, a sustainable design studio, and before that cofounded creative.ai. If he’s not at his keyboard, he can be either found with his nose in a book or hugging a tree. 💚
Twitter
Summary
- Vox chart: Global crude oil production continues to increase, even though people have predicted running out of oil multiple times
- New technologies (fracking) increased oil production. AI makes this possible.
Tech can support the fossil fuel industry in three areas:
- Upstream: 3D modeling to interpret seismic data
- Midstream: Predictive maintenance & smart pipeline monitoring
- Downstream: Facilitate refining
Big tech companies involvement in the the oil industry
- Google 8% share — more detail on contracts in the Greenpeace report. They are still active.
- Microsoft 18% share — more open about what they’re doing
- Amazon 33% share — changed page from from “AWS for Oil & Gas” to “AWS Energy”
Action
- Sign the petition — #BigTechLovesBigOil
- Spread awareness of the issue & the campaign
- If you work for Microsoft, Google, Amazon: Join the employee climate groups [Twitter list]
- If you work in AI:
- Join Microsoft & Google’s climate workshops
- Check out Google Earth Engine
It only gets you so far though as its super expensive for actual use (it will time out when using too much memory without a paid account), and it also doesn’t support more powerful ML techniques like deep learning, etc.- A simple sentinel-satellite mosaic you can use for any kind of remote sensing tasks
- Very simple landcover classification: you can basically add a couple of points just in the mapviewer, and use those to train a CART/ML model, check the results, repeat: so super quick active learning with minimal code
Resources
Some examples of good articles/campaigns around this issue:
- Oil in the could — Greenpeace report focused on Amazon, Microsoft & Google
…after which google said it will “no longer build custom ai for gas & oil” (but will still support ongoing contracts or non-custom work (still a positive change) [Article here] - autodesk.earth
- Oil is the new data (written by a Microsoft engineer)
- Microsoft writing about working with the oil & gas industry & Azure solution for the oil & gas energy sector
- How Google, Microsoft, and Big Tech Are Automating the Climate Crisis [Gizmodo 2019]
- XR’s #BigTechLovesBigOil campaign
- #BigTechLovesBigOil memes on Twitter
- XR NYC Medium post: Big Tech Loves Big Oil: A Bid For White Supremacy
- XR NYC Announcement
Big Tech Loves Big Oil
Google is carbon neutral. Microsoft is going carbon negative. Amazon’s CEO Jeff Bezos is donating 10 billion dollars to fight the climate crisis. Big Tech is on the vanguard of human progress!
Except it’s not. Google continues to use machine learning to discover oil reserves. Amazon Web Services has an Oil and Gas division. Microsoft recently hosted a conference titled “Empowering Oil and Gas with A.I.”, a relationship its General Manager of Business Applications summed up best: “…the future health of oil and gas companies is contingent upon our ability to drive transformation.”
In the midst of a global reckoning on race, Big Tech admitted they cause harm by calling for regulations on facial recognition technology used to subjugate Black People, Indigenous People, and People of Color, the same people who will bear the brunt of the climate emergency. Big Tech needs to stop profiting from the destruction of the planet and its most vulnerable citizens. We, technology’s workers and users, must claim our rights and voice the truth.