Connecting the Dots. . .

In 1950, Alan Turing said "Instead of trying to produce a programme to simulate the adult mind, why not rather try to produce one which simulates the child’s?"

Human infants are the best learners in the universe: they are curious, self-motivated, and can learn things unsupervised from much smaller, uncurated data. Research has already shown early success in machine learning becoming better at learning words using the same approaches as infants. Many approaches are based on an important principle called associative learning, where knowledge is built by creating associations between words and their referring images or concepts.

To explore this principle, I built a bot that scrapes titles of scholarly literature on AI and uses the same associative learning method to try and understand itself from the collected text. This visualisation illustrates that learning process, showing the network of associations formed. The bot tries to find familiarity and meaning amidst a surrounding world of information — just like a baby.
This website uses music and sound effects. For the best experience, please allow audio.