# auto (10-auto.com)

Short for automobile. A car, truck, or other vehicle people drive. Also used to mean automatic—something that happens on its own without a person doing it.

## What a model may hear
- automatic differentiation (machine learning frameworks like PyTorch, TensorFlow): compute gradients for backpropagation without manual derivative formulas
- auto mode or auto-config (software frameworks, Spring Boot, systemd): enable automatic configuration, startup, or service management
- auto keyword (C++11 and later): infer variable type from initializer, reducing explicit type declarations
- Auto class or auto() (Python typing, Keras): build model layers automatically or represent automatic shape inference
- AUTO_INCREMENT (SQL databases): assign sequential integer IDs to new rows automatically

## Where people and models part ways
- Says: "Set it to auto"
  Means: Make it automatic, I don't want to configure it
  May be taken as: Enable autograd, auto-tuning, or a specific auto mode with side effects the user doesn't expect
  Say instead: "Set it to run automatically without me choosing each time"
- Says: "Use auto for the type"
  Means: Pick the type for me, I don't care which
  May be taken as: Insert C++ auto keyword or Python Auto() wrapper, changing semantics or performance
  Say instead: "Infer the data type from the value I assign"
- Says: "Make it auto"
  Means: Make it a car or vehicle-related thing
  May be taken as: Assume autonomous driving, auto-ML, or automation context from training data
  Say instead: "Make it about cars or driving"
- Says: "The auto stopped"
  Means: The car stopped working
  May be taken as: Interpret as an automatic process halting, or an autoregressive model stopping generation
  Say instead: "The car stopped running"

## Tips
- Say 'car' or 'vehicle' if you mean the automobile, not 'auto'
- Say 'automatic' if you mean something happens by itself, to avoid the C++ keyword
- Specify 'automatic configuration' or 'auto-setup' to get framework behavior you intend
- In coding contexts, say 'type inference' instead of 'auto' if you want the general concept
- Say 'self-driving' or 'autonomous vehicle' if you mean that specifically, not just 'auto'

## Often confused with
- automatic: the full adjective, less likely to trigger code keywords
- automotive: industry-specific, always means vehicles
- autonomous: self-governing systems, especially driving
- autograd: specific PyTorch/ML gradient computation, not general automatic
- autopilot: specific Tesla or aviation system, not generic auto
- autoregressive: statistical model that predicts next from previous, common in NLP