Conclusion

Key takeaways

We’ve seen how some problems are easy for humans to solve, but hard for traditional computer approaches. Image recognition is one of these so-called artificial intelligence challenges.

Neural networks have enabled huge progress in image recognition, and a wide range of other kinds of hard problems too. A key part of scientists’ early motivations was the puzzle that biological brains—like pigeon or insect brains—appeared to be simpler and slower than today’s huge supercomputers, yet they could carry out complex tasks like flight, feeding, and building homes. Those biological brains also seemed extremely resilient to damage or imperfect signals. Digital computers and traditional computing weren’t either of these things.

Today, neural networks are a key part of some of the successes in artificial intelligence. There is continued interest in neural networks and machine learning, especially deep learning, where a hierarchy of machine learning methods are used. In early 2016, Google’s DeepMind beat a world master at the ancient game of Go. This is a massive milestone for artificial intelligence because Go requires much deeper strategy and nuance than chess. Researchers had thought that a computer playing that well was years away. Neural networks played a key role in that success.

We hope you’ve seen how the core ideas behind neural networks are quite simple. And we hope you’ve also had fun experimenting with neural networks. Perhaps we’ve sparked your interest in exploring other kinds of machine learning and artificial intelligence.

If we’ve done any of these things, we have succeeded.