I analyzed a TED talk for one of my classes. These are my thoughts. I watched Jeremy Howard’s talk about deep learning titled: The Wonderful and Terrifying Implications of Computers That Can Learn. Deep learning involves one algorithm that allows a computer to learn new things. Computers can teach themselves to select relevant images from a collection of millions based on the image and not text from a website. They can perform medical diagnostics to assist pathologists. Two of the most interesting examples was a computer learning to read and speak Chinese and a computer that could write captions for photos.
The entire talk was example driven. Photos and videos were used extensively to show how deep learning works and what it is being used for. The examples were chronologically based starting with a computer that taught itself to play chess in 1962 up to where we are at with deep learning today. I thought I might be about to watch a long winded lecture, but everything was very fast paced due to the many examples that were squeezed into the 19 minute talk.
I think the concept of deep learning is fascinating and love seeing new technologies emerge. It is amazing to see what people can create, and I look forward to seeing the new advances in this area.
Check out this TED talk at:
