The Deep Learning Hype At The Peak Of Gartner’s Hype Cycle

There is currently a huge hype around deep learning. You can also see that on Gartner’s hype cycle.

What’s a hype cycle and how does it work? Continue reading to learn more.

Gartner is creating for some time now hype cycles for a lot of areas, like machine learning, big data and so on. I personally use them to find out how far has a technology has come.

How far tech like deep learning for instance has come. Getting from idea, development, the first implementations, into a state of productivity.

The state where people are actually getting a benefit from it, where companies are making money with it. The hype cycles from Gartner are a very nice tool for that.

How A Hype Cycle Looks

A hype cycle is curve that first ramps up to a peak, then falls down into a low and gets back up into a plateau.

Innovation

The first stage where stuff is ramping up its when ideas are created when new technologies emerge. Its the innovation stage.

How the hype cycle works is let’s look for instance at deep learning (DL). It started way on the left and made its way way up the hill of the curve.

People have realized that DL is a topic for interest and people are working on it. They are conceptualizing and trying out ideas.

It’s the innovation phase.

The more and more people are doing this DL is going up the curve.

Peak of inflated expectations

Until it reaches the second stage: The top. Gartner calls this peak of inflated expectations.

It is the time when the hype is really big. Everybody is talking about deep learning, everybody is working on it. People think that DL is a technology that is getting everywhere and it influences everything in our lives.

It’s basically the time when a lot of of people think you can implement something everywhere and make a lot of money with it.

It’s also where deep larning is in the current hype cycle of 2017.

Trough of Disillusionment

What happens next is, as more and more companies are working on it, they realize that it’s harder then they thought.

DL is harder to implement, harder to develop and harder to sell as a product. This is when a lot of people are frustrated, it is where a lot of people give up and companies go out of business.

At that point what happens with with DL is that it’s going down the hill. It’s going away from the peak of inflated expectations into a low.

At this point people are disillusioned and wake up from the hype.

Plateau of productivity

From that point on DL will be going a little bit further up the next hill into a plateau. Gartner calls this the plateau of productivity.

When a tech goes through the low point of disillusionment and reaches the plateau of productivity, then it is an it is accepted by the public. In the case of DL that would be when we interact with deep learning algorithms on a day-to-day basis. Often without even knowing it.

It’s the plateau of productivity where people have realized what you can actually do with it.

The tech or the tool is a mainstream application.

Time to Plateau Prediction

What’s also very interesting is that Gartner tries to foresee a time frame. How long it will take for one of those items on the hype cycle to reach the plateau of productivity.

So, in case of human augmentation, one of my favorite topics, Gartner predicts more then 10 years. This hasn’t changed for years now 😀

When you look at other tech you see that it’s 5 years, 2 to 5 years away. So, you can actually make a prediction how fast will this take go through the whole hype cycle.

What I totally recommend is not only look look at this year’s hype cycle. Look at the hype cycle of the past years as well. This way you get a feeling how fast stuff really happens.

Those predictions of how long it takes until it reaches the plateau of productivity on the right side change. Year by year, sometimes even tech completely disappears from cycle.

One year it says to five years and in the next year it’s disappeared completely. That is because it’s it’s gone through the hype cycle it’s reached its plateau of productivity already.

Because this is not exact science, it’s a estimation by Gartner.

A Awesome Tool You Have To Use

The hype cycle is a very interesting tool that helps you get a feeling how fast is stuff emerging. It tries to predict how long it will take something to be mainstream.

So, if you want to keep track of trends I highly recommend the Gartner hype cycle.

There are many hype cycles not only for emerging tech. Just google it, you will find a lot of them.

They also make an analysis, but this analysis costs a lot of money. But the hype cycles usually you can google and find.

What are the trends you currently watch closely?

Do you use the hype cycle? Or are you using something else?

Please leave a comment that would be super awesome ☺

 

Say Hello On: Instagram | LinkedIn | Facebook | Twitter | YouTube | Medium | Snapchat

Don’t miss my Daily Data Science and Big Data Podcast on SoundCloud and Anchor

Subscribe to my newsletter: Here

Leave a Reply

Your email address will not be published. Required fields are marked *