# ADALINE
**Entity class:** Hardware neural network / adaptive linear neuron
## Definition and history
ADALINE is an adaptive linear neuron built and trained in hardware by Bernard Widrow and Marcian Hoff around 1959–60. [[articles/Intelligence Inside|Intelligence Inside]] uses it to anchor a documented pre-microprocessor machine-learning lineage. Inputs and adjustable weights produce an output whose error can guide later weight changes. It was a trainable physical element, rather than a metaphor applied to a conventional computer after the fact.
## Mechanism and limits
The historical connection to the 4004 passes through Hoff's career, while the mathematical connection to later communications hardware passes through the LMS rule. Neither connection turns every subsequent processor into an ADALINE. The value of the lineage is that machine learning had hardware, update rules, and commercial pathways decades before the public conversational interface.
## Relationships
**Built by:** [[wiki/Bernard Widrow|Bernard Widrow]] and [[wiki/Marcian Hoff|Marcian Hoff]]. **Trained by:** [[wiki/Least Mean Squares Algorithm|Least Mean Squares Algorithm]]. **Later silicon comparison:** [[wiki/Intel 80170NX ETANN|Intel 80170NX ETANN]]. **Article route:** [[articles/Intelligence Inside|Intelligence Inside]].
## Sources
[Widrow, *Thinking About Thinking*](https://isl.stanford.edu/~widrow/papers/j2005thinkingabout.pdf); [[articles/Intelligence Inside|Intelligence Inside]].