Namaste! Deep Learning… The Machine Learning technique that is based on Artificial Neural Networks. It is very easy to use. Get and clean up your data. Find a problem with that data. Create a Neural Network architecture that could solve the problem. And then… Train, train, train. Ten minutes, an hour, over night, for a couple of weeks. Depending on the problem. Quite straightforward. But wait… Deep Learning is a hype that is doomed to fail, isn’t it?
Talk to people about Deep Learning and discover the patterns.
Yes, Deep Learning as a field generates a lot of memes. We have been discussing for quite some time. And you read about Deep Learning everyday. Which is usually very insightful. There is one meme that is very good at replicating itself: Deep Learning is just a hype and there is going to be a huge wave of disappointment which is going to hit hard.
I talk to a lot of people with different backgrounds. That’s what I do. We exchange anecdotes and opinions. And we discuss heavily about the prospects of Artificial Intelligence in the next years. Sometimes I run into people with expectations that are a little bit too high. Like the all-knowing, universal AI that you can program in your basement. We are not there yet! Not yet, comrades! And sometimes I meet people that doubt the whole concept of Deep Learning because it is just a hype and we are on the peak of inflated expectations. Well, that is what they say. But only partially rightfully so.
There is a huge truth about Deep Learning that needs to be heard. There is an elephant in the room.
It is my job to read about Deep Learning every day. You know what I like the most? It is the huge list of use-cases that have been successfully solved with Deep Learning. Everyday there is a new success that you can learn about. It is either outperforming an already existing algorithm. That is what we saw a couple of years ago in the image processing domain. Or it is solving problems that were yet unsolved. Many problems have already been solved.
And that is exactly my point. Mark my words: There is a lot of Deep Learning enthusiasts out there that could make a living by just applying those successful use-cases. For years. I repeat. You can do a great job for quite some time just applying the already existing Deep Learning problem solvers. Find a solution in the huge pool and map it to your challenges. Simple as that.
Sometimes you do not want to be the groundbreaker or the iconoclast of Deep Learning. Sometimes you just want to get things done quickly.
You know. I want to make one point very clear: Deep Learning today is not the question of doing groundbreaking new work. Creating special layers for your Neural Networks or a promising new optimizer. Just to name two options. At least not the main part. Today Deep Learning is a little more about finding already solved problems. Once this is done you would „just“ map the solutions to your domain. Maybe add something. Or maybe change a couple of things. Finding, adapting, solving.
My mantra is: Everyone can do Deep Learning now! And the biggest challenge is not solving problems from scratch. The biggest challenge is finding an already existing problem solver. And there is a lots of solved problems already out there. Keep your eyes wide open!
Final words. A conclusion.
Yes, there is a certain hype-feeling surrounding Deep Learning. And yes, there is a huge portion of expectations that are too high. But Deep Learning is not going to fail. Simple because we are now in the situation where you could just spend years already existing problem solvers.
Stay in touch.
I hope you liked the article. Why not stay in touch? You will find me at LinkedIn, XING and Facebook. Please add me if you like and feel free to like, comment and share my humble contributions to the world of AI. Thank you!
If you want to become a part of my mission of spreading Artificial Intelligence globally, feel free to become one of my Patrons. Become a Patron!
A quick about me. I am a computer scientist with a love for art, music and yoga. I am a Artificial Intelligence expert with a focus on Deep Learning. As a freelancer I offer training, mentoring and prototyping. If you are interested in working with me, let me know. My email-address is firstname.lastname@example.org - I am looking forward to talking to you!