When was the artificial intelligence? There is no specific moment. Like all groundbreaking technologies, from the phone to the Internet, it is the sum of many discoveries of many people.
One of the key moments was Alexnet’s creation in 2012: The first example of deep learning (deep learning). The birth of modern artificial intelligence is considered. And now You can download your source codeto examine or apply it wherever you want.
It is an initiative of the Computer History Museum (CHM), which has taken five years to convince Google to free it, and find the exact source code that marks The birth of modern AI.
Alexnet, the seed of deep learning
Alexnet was developed in 2012 by then postgraduate students at Toronto University, Alex Krizhevsky and Ilya Sutskever, and his advisor professor Geoffrey Hinton.
Interestingly, although Alexnet was created by these students, Who took the Nobel Prize in 2024 was Geoffrey Hintonfor its contributions to AI since the 80s of the last century. When asked about the subject, he says in a jocular tone: «Ilya thought we should do it, Alex made him work and I took the Nobel Prize.»
Why is Alexnet so important? In 2012, the artificial intelligence of the time could already do things like finding objects within photos, but demanded a training guided by an expert, which tells AI what he has to look for, and when he does it right or wrong.
Alex Krizhevsky and company got, with Alexnet, create The first AI that uses deep learning. This means that she learns alone to identify objects within photos, without anyone telling it.
It was infinitely faster and infinitely more precise, with fewer failures, than everything that existed so far. With the advantage that he did not require human intervention, so he could train much faster.
Actually, what these pioneers did is take three technologies that already existed independently, and put them to work at the same time: deep neural networks, databases with millions of images, and GPUS.
AlexNetwhich is called its programmer, Alex Krizhevsky, was formed by several layers of deep neural networks that learn increasingly complex visual characteristics.
These deep networks could automatically discover patterns at different levels of abstraction: from simple silhouettes in the first layers, to details of objects in the deepest layers.
For example, the first layers distinguish basic forms, such as the difference between a tree and a cat. For more detailed features, such as the type of hair or race, the information is passed to the deepest layers.
For all this to work, a great calculation power was required. Alex Krizhevsky decided to use a graphics card, so This is the origin that Nvidia GPUs are now the most used the chipsand Jensen Huang’s company has become one of the richest in the world.
Although Alexnet developed to recognize images, his deep learning algorithm began to be used for everything: voice recognition, language models, image generation, etc.
Intuiting the immense possibilities of deep learningAlex Krizhevsky, Ilya Sutskever and Geoffrey Hinton formed the company Dnnresearch Inc … which was bought by Google just a few months later, in 2013.
If you wonder how Google, Amazon, Meta or Microsoft have come to master everything, here are a good example: any revolutionary technology that arises, buy it instantly and prevent it from becoming competition.
But None of them are still in the company. Ilya co -founded Openai in 2015, and works in Chatgpt. Alex left Google in 2017 to work at Desa, and Geofrey did so in 2023 for a very different purpose: denounce the dangers of AI, due to the few security measures of technological ones, where the only thing that matters is to win.
Alexnet is already computer historybecause it is the birth of modern AI. And now You can download your source code for free In Github, thanks to the initiative of the Computer History Museum, which has convinced Google to free him.
Its utility today is limited, because it has been highly overcome. But Study Alexnet’s source code It is like reading the first edition of El Quijote, or examining the interior of the first computer: in those lines of code is the spark of the life of modern artificial intelligence, which is changing the world. For better or worse, it is still about to see.
Know How we work in NoticiasVE.
Tags: Culture, artificial intelligence, viral, computers, retro, curiosities, history