Deep learning
Machine learning with neural networks that have many layers: the approach behind today's chatbots, image generators and voice assistants.
Deep learning is machine learning done with neural networks that have many layers between input and output. Each layer builds on the one before: in a network that reads photos, early layers respond to edges and colours, later ones to shapes, and later still to whole objects or faces.
The ideas are decades old. They took off in the 2010s, when 3 things arrived together: enough data, graphics chips fast enough to train big networks, and better training methods. Its pioneers won computing's top prize, the Turing Award, for 2018. Nearly every AI feature you use now, from voice typing to chatbots, is deep learning.
For example
Your phone transcribing a voice note on the train, with no signal, is deep learning running on the device itself.
Put it to use
Where you'll meet deep learning in ChatGPT, Gemini and Claude, how to use it there, and the mix-up most people make.
Daily Bytes Plus is in the Android app.