What Skeptics Say About Deep Learning
Every hyped up technology needs a healthy dose of skeptics. Deep Learning is not an exception. To keep us all level set, let’s see what a few of these skeptics are saying. Here are three recent articles from three different data science/machine learning experts.
This AI Boom Will Also Bust
The bottom line here is that while some see this new prediction tech as like a new pipe tech that could improve all pipes, no matter their size, it is actually more like a tech only useful on very large pipes. Just as it would be a waste to force a pipe tech only useful for big pipes onto all pipes, it can be a waste to push advanced prediction tech onto typical prediction tasks. And the fact that this new tech is mainly only useful on rare big problems suggests that its total impact will be limited. It just isn’t the sort of thing that can remake the world economy in two decades. To the extend that the current boom is based on such grand homes, this boom must soon bust.
What’s New In Machine Learning
Does it mean I needed to learn Machine Learning from scratch? Fortunately for me, the answer to that question was no.
Why?
Because of two reasons:
1. Machine Learning technology hasn’t changed that much since my PhD
2. And when it has changed, it is to become closer to Mathematical Optimization, something I am quite familiar with.
Has Deep Learning Made Traditional Machine Learning Irrelevant:
Even though Deep Learning ANNs could be directed toward these same class of problems based on largely structured data with some unstructured now being introduced, they are not suitable from an efficiency standpoint.
So decidedly no, Deep Learning has not and will not make traditional Machine Learning techniques obsolete.
All these criticism have come in the past month. They are all pretty tame as compared to this blog post in 2014 “Get off the deep learning bandwagon and get some perspective”:
The takeaway is this: machine learning isn’t a tool. It’s a methodology with a rational thought process that is entirely dependent on the problem we are trying to solve. We shouldn’t blindly apply algorithms and see what sticks. We need to sit down, explore the feature space (both empirically and in terms of real-world implications), and then consider our best mode of action.
Sit down, take a deep breath. And invest the time to think it through.
And most importantly, avoid the hype.
only to come back later and write in his disclaimer:
And sometimes we even explore methods decades old, applying only a slightly different twist, yielding significantly different results — and thus a new research area is born.
Interesting arguments all of them, but are any of them valid?
To understand Deep Learning better, please see my 2017 predictions for Deep Learning or sign up with Intuition Machine to get the conversation going.