In the previous chapter, we discussed how artificial intelligence is no longer confined to a single outstanding model but is becoming a complex ecosystem where various tools and functions cooperate.
Once, we simply thought of artificial intelligence as 'the technology that makes machines smarter.' Then chess-playing computers appeared, and voice-understanding AI emerged.
AI that handles language must understand well how words appear in order in sentences. 'I ate rice' and 'Rice I ate' have different word orders but the same meaning.
We sometimes see AI recognizing cats in photos, understanding human speech, or creating naturally flowing sentences as if written by humans.
Recent AI technology has advanced remarkably. Now AI can understand human speech, draw wonderful pictures, and even write sentences naturally on its own.
Scaling in deep learning refers to the process of adjusting the size and complexity of AI models to improve performance — in other words, making AI smarter.
So far, we've examined three major ways artificial intelligence learns and develops: supervised learning, which teaches it correct answers in advance; and more advanced methods.
The AI technologies we frequently hear about in the news and internet today all started from one special structure. That structure is the artificial neural network.
In Chapter 1, we explained that machine learning and deep learning are technologies that learn by themselves by looking at data. Here, a curious question arises.
The word artificial intelligence is no longer unfamiliar today. Smartphone facial recognition, email spam filters, YouTube's video recommendation system — all powered by AI.