Artificial intelligence didn't appear overnight. The ideas powering today's ChatGPT, image recognition, and recommendation systems were patented decades ago — many by researchers who had no idea what they were building toward. John Hopfield's 1985 neural network patent predated the commercial internet. Early backpropagation training patents from the 1980s described mathematics that would take 30 years of hardware progress to make practical at scale. Today's large language models, self-driving cars, and protein-folding algorithms are direct descendants of this work. These are the foundational AI patents — what they actually cover, who filed them, and why they matter.
Patents in this topic
US 4914603 · 1990 · GTE Laboratories Inc
How an Early Neural Network Learned Faster with a Special Variable
US 10824959 · 2020 · Amazon Technologies Inc
How AI Models Can Explain Their Decisions with Simple Rules
US 11727263 · 2023 · Samsung Electronics Co Ltd
How AI Learns to Write Better Sentences Using Feedback
US 10607134 · 2020
AI System Learns Avatar Actions from Game Objects
US 4660166 · 1987 · California Institute of Technology
Hopfield Network for Solving Problems and Storing Memories
US 10410117 · 2019 · BrainChip Inc
How AI Learns and Shares Skills Like a Brain
US 10423875 · 2019
Using a Camera to Monitor and Control Neural Networks
US 10282665 · 2019 · Sony Corp
How a Computer Learns to Pick the Best Action
US 9361579 · 2016 · International Business Machines Corp
Large scale probabilistic ontology reasoning
More topics








