What do you think when you keep hearing about artificial intelligence ?
There’s been a lot confusion among the public about the term. Not helped by often dramatic news stories about how Artificial Intelligence, or“AI”, will destroy jobs, companies or even the world in the near future.
A lot of that confusion comes from the misuse of such terms like “AI” or “machine learning”.
So here’s a short text-guide to explain AI:
First of all: what is the difference between “#AI” and “#machinelearning”?
Think of it like the #difference between “#economics” and “#accounting”.
Economics is a field of study, but I´m sure you wouldn’t hire a Nobel Prize-winning economist to do your taxes. So, Artificial intelligence is the field of science. Covering how computers can make difficult decisions as well as humans can. But machine-learning refers to the modern technique for creating the software that learns from big data.
The difference becomes much more important when money is involved. Investors from Venture capital companies often dismiss Artificial Intelligence as hype because they have got skin in the big game.
So, VC´s often prefer startups that make machine-learning software with a focus on clear, commercial applications, like platforms that can filter emails from companies with natural language processing or which are able to track customers in a retail store with facial recognition. To mention some real business cases.
On the other hand, many universities and some of the largest technology companies like Facebook, Google or Apple have large laboratories to carry out research to drive the wider field of Artificial Intelligance forward. A lot of the tools these companies have invent are freely available online. So TensorFlow from Google or Pytorch from Facebook to mention two of them.
Ok, so far so good. But why does the term “deep learning” come up everywhere?
This is easy. It´s because the most exciting applications of “AI” today give computers the ability to “learn” hoe to carry out tasks from data. And this without being programmed to do those tasks.
But the terminology used is confusing. It involves a mix of different technologies and techniques. Many have also have the word “learning” included in their names.
For instance, there are three core types of machine learning which can all be carried out in different ways: supervised, unsupervised, and reinforcement. And most important, they can also be used with statistical machine learning, Baeysean machine learning or symbolic machine learning to be accurate.
Most popular applications of machine learning use a neural network so you don’t really need to be clued up on these thoughts.
And what is “deep learning”?
Deep Learning a specific approach to use a neural network—essentially, a deep neural network with lots of layers in total. The techniques have led to popular services that are used today, including speech-recognition technologies and Google’s automatic translation that can be used by everybody on the internet.
Each of those layers can represent increasingly abstract features. e.g. a social media company might use a “deep neural network” to recognize faces from users and customers. One of the first layers describe the dark edges around someone’s face, another describes the edges of a mouth and nose, and another describes other parts of the human body. These layers become increasingly abstract, but put together they can represent an entire face and so a human. This is deep leaning.
I´ve heard the phrase neural network above. What is a neural network?
Neural Network are a computer system inspired by the #human #brain. It´s been been going in and out of fashion since the last world war.
And what does a neural network look like on a screen?
Well, it´s like a code. Engineers from Google’s Artificial Intelligence subsidiary “DeepMind” write nearly all their code in the general purpose programming language “Python”. First released in 1991.
It has been used to develop all sorts of programs, basic, highly as well as complex, including some of the most popular services on the Web of modern times: So it´s been used for YouTube, Instagram as well as for Google itself.
If you want to learn more about AI, look here …..