I recently came across an interesting thought: why are some people able to stick to their ideas even when the whole world opposes them?

I thought of Geoffrey Hinton. This scientist, hailed as the father of deep learning, when he proposed using artificial neural networks to mimic the human brain for processing complex data like images and speech, was basically considered crazy by everyone. In the 1990s, computing power couldn't keep up, storage was insufficient, and data was scarce. People mocked Geoffrey Hinton’s ideas as nothing more than a fairy tale.

But interestingly, he didn’t give up. Questioned for over thirty years, ridiculed for over thirty years, yet he persisted. This isn’t some motivational cliché; it’s something that actually happened.

The turning point came. In 2012, computer hardware finally caught up. Geoffrey Hinton led his team in the ImageNet ILSVRC image recognition competition, where their AlexNet won with a error rate 10% lower than the second place. This is one of the top competitions in the field of computer vision. One victory, and the whole world started to take deep learning seriously.

Do you know what’s the most heartbreaking part? Those who once mocked him suddenly started treating Geoffrey Hinton as a god.

Looking back, the reason he could persist boils down to two insights. The first is the potential of unsupervised learning — deep learning can autonomously discover hidden patterns and regularities from massive amounts of data without manual labeling. This is especially important for handling high-dimensional, complex data like images, speech, and text. The second is the ability to model complex functions — neural networks can represent functions with many parameters, capturing deep relationships between inputs and outputs.

These two ideas were impossible to verify at the time, which is why they were ridiculed. But Geoffrey Hinton believed that as long as time passed and hardware improved, this path would be the right one.

Now, deep learning has become the core foundation of AI, driving the arrival of the entire era of intelligence. From being considered crazy to being revered as a master, Geoffrey Hinton proved one thing in thirty years: doing the right thing isn’t hard; what’s hard is persisting when no one recognizes you.