Steve Easterbrook

Harshit Gujral, Eshta Bhardwaj, Dushani Perera et al.

Datacenter expansion under generative AI is increasingly framed as compatible with sustainability because of efficiency gains, cleaner electricity procurement, and improved facility design. Yet these claims often do not show that absolute electricity, water, material, waste, and community-facing burdens are falling. This Perspective addresses that evidentiary gap. Rather than asking whether efficiency gains are real, we ask when such gains are being enlarged into claims of system-wide sustainability to justify continued expansion. We develop a rebound-informed diagnostic framework for evaluating AI and datacenter sustainability narratives across five tests: metric, boundary, reinvestment, burden shifting, and governance. Applied to major AI industry sustainability reporting, the framework shows that firms largely justify continued expansion through efficiency improvements and clean-energy procurement, rather than by demonstrating reductions in absolute resource use. Applied to plateau claims in the literature, we show that many claims establish local or relative improvements while leaving energy rebound, lifecycle burdens, and enforceable limits unresolved. We argue that these sustainable-growth narratives begin to function as greenwashing when they use efficiency improvements to claim sustainability even as absolute energy, water, material, and public health burdens continue to increase. We conclude by positioning digital sufficiency as a burden-of-proof framework for governance: those advocating further datacenter expansion must show that it reduces, rather than merely redistributes or defers, absolute burdens across the full system.

Wei Yu, Wenxin Chen, Songhenh Yin et al.

Recent works in video prediction have mainly focused on passive forecasting and low-level action-conditional prediction, which sidesteps the learning of interaction between agents and objects. We introduce the task of semantic action-conditional video prediction, which uses semantic action labels to describe those interactions and can be regarded as an inverse problem of action recognition. The challenge of this new task primarily lies in how to effectively inform the model of semantic action information. Inspired by the idea of Mixture of Experts, we embody each abstract label by a structured combination of various visual concept learners and propose a novel video prediction model, Modular Action Concept Network (MAC). Our method is evaluated on two newly designed synthetic datasets, CLEVR-Building-Blocks and Sapien-Kitchen, and one real-world dataset called Tower-Creation. Extensive experiments demonstrate that MAC can correctly condition on given instructions and generate corresponding future frames without need of bounding boxes. We further show that the trained model can make out-of-distribution generalization, be quickly adapted to new object categories and exploit its learnt features for object detection, showing the progression towards higher-level cognitive abilities. More visualizations can be found at http://www.pair.toronto.edu/mac/.

Wei Yu, Yichao Lu, Steve Easterbrook et al.

Applying resolution-preserving blocks is a common practice to maximize information preservation in video prediction, yet their high memory consumption greatly limits their application scenarios. We propose CrevNet, a Conditionally Reversible Network that uses reversible architectures to build a bijective two-way autoencoder and its complementary recurrent predictor. Our model enjoys the theoretically guaranteed property of no information loss during the feature extraction, much lower memory consumption and computational efficiency.

Soukayna Mouatadid, Pierre Gentine, Wei Yu et al.

Climate projections suffer from uncertain equilibrium climate sensitivity. The reason behind this uncertainty is the resolution of global climate models, which is too coarse to resolve key processes such as clouds and convection. These processes are approximated using heuristics in a process called parameterization. The selection of these parameters can be subjective, leading to significant uncertainties in the way clouds are represented in global climate models. Here, we explore three deep network algorithms to infer these parameters in an objective and data-driven way. We compare the performance of a fully-connected network, a one-dimensional and, a two-dimensional convolutional networks to recover the underlying parameters of the Lorenz-96 model, a non-linear dynamical system that has similar behavior to the climate system.

Christoph Becker, Ruzanna Chitchyan, Leticia Duboc et al.

Sustainability is a central concern for our society, and software systems increasingly play a central role in it. As designers of software technology, we cause change and are responsible for the effects of our design choices. We recognize that there is a rapidly increasing awareness of the fundamental need and desire for a more sustainable world, and there is a lot of genuine goodwill. However, this alone will be ineffective unless we come to understand and address our persistent misperceptions. The Karlskrona Manifesto for Sustainability Design aims to initiate a much needed conversation in and beyond the software community by highlighting such perceptions and proposing a set of fundamental principles for sustainability design.