Online learning may be a growing trend, but just what is it? A recent study explored how neural networks would teach a webmaster (whose task was to write blog posts).
What Is Online Learning In Neural Networks
Nanophages, Cancer Metastasis and Neural Networks in the Classroom
In 2017, top executives from Johnson & Johnson, JPMorgan Chase, General Electric, Microsoft, and many other businesses met at Stanford University. Also present were Stanford students. Many of these students had gone to a four-week bootcamp that Stanford has given in the past two years. To reach what is best termed a peak, Stanford developed a new technology called Neural Networks for Learning.
Neural Networks for Learning exists as an online course that has one thing in common with college courses – it is open to all. Using this online course is entirely different from a typical online course. The program is open source. Anyone can download the course’s code and modify it. Learning was extended when the program was published online. Unlike Stanford, the program is maintained by students. The educational content within the neural networks is reviewed every week. Several universities have already sought Stanford’s approval to make the course available for their students.
In the program, students work on challenging AI projects. There are many funding options available, most typically by donors. In the course, students are expected to make meaningful contributions that keep a positive impact on the public good. This is why the students have to receive a B or higher grade in order to be permitted to complete the program.
Creating the most appropriate neural network model for a whole network problem
Students learn an incredible amount of content in their four weeks of the class, despite being very busy. But once the material is exhausted, they have another four weeks of classwork that expands their knowledge of this level of learning. They use model-driven problems to tackle even larger problems. Their thinking is more sophisticated, deeper, and challenging. They become team players.
This online learning program has shown significant growth and strength. It is currently supported by more than 30,000 students who have taken more than 130,000 hours of this online class alone.
In 2014, Joseph Weizenbaum, the first author of the book Artificial Intelligence, was asked if he knew what artificial intelligence was. His response was this: “I thought it was a hip thing to try to be. It was completely pointless because I knew there was no place for it except as a very small part of finance or education.”
Are online learning tools like neural networks the future of education? Not for me. In an industry that has been effectively managed, that incorporates a unique combination of human subject matter knowledge, subjective teaching, and machine-learning based solutions, we need all of these. In 2019, we will need all of them. We want to create entirely new skills while teaching all the things that we know today. This is called holacracy. But we still have many major trends from the last two years that challenge this future.
With AI grown up and from off the shelf solutions becoming more available, does it matter what happens to this or any other education model? Is it better or worse? After all, companies like Netflix or Salesforce use AI to better serve consumers. Does this mean that learning in colleges and universities will become obsolete? Maybe.
Eventually, I believe neural networks in the classrooms will become part of the overall curriculum. Over time, AI will have more of an impact on education. But it will not create a new class of traditional credential holders. Instead, a new type of curriculum could emerge, i.e. a peer-reviewed academic journal that teaches fully automated solutions in the classroom. A professor could use the whole book for that purpose.