docs: add ethical context note to CIFAR-10 classifier tutorial

[stuckvgn]

Suggestion

Add a brief ethical context note to the "Training a Classifier" tutorial regarding the CIFAR-10 dataset's treatment of animals as classification objects.

Context

The "Training a Classifier" tutorial is one of PyTorch's most-visited pages and many developers' first exposure to image classification. It uses CIFAR-10, which is 60% animal classes (bird, cat, deer, dog, frog, horse).

The tutorial trains a model to classify animals purely as objects — identical to how it classifies airplanes, automobiles, ships, and trucks. While this is technically correct for the classification task, it's worth acknowledging the ethical dimensions of animal classification in AI, particularly given:

1. Precedent: The CIFAR-10 parent dataset (80 Million Tiny Images) was taken offline in 2020 after MIT researchers identified racist, sexist, and offensive labels. This established that ethical review of classification datasets is appropriate.

2. Research: Peer-reviewed papers have documented how AI classification systems perpetuate speciesist assumptions:

   • Hagendorff et al. (2023). "Speciesist bias in AI." AI and Ethics. Notes that image classification schemas "treat animals identically to inanimate objects."
   • ImageNet's ILSVRC schema has ~400 animal classes among its 1,000 classes with no ethical distinction from object classes.

Suggested Change

A brief note (2-3 sentences) in the tutorial acknowledging that:

• Datasets reflect choices about what to classify and how
• Practitioners should consider the ethical implications of classification systems, especially those involving sentient beings
• Links to resources on responsible AI and dataset documentation

This doesn't require changing the tutorial's content or code — just adding a brief awareness note consistent with PyTorch's commitment to responsible AI.

Scope

A single paragraph addition. No code changes.