Many companies are trying to make artificial intelligence more accessible, but few are doing it as well as Lobe. The startup, which launched in early 2018, provides users with a simple and easy-to-use drag-and-drop environment that makes building deep learning programs a breeze. So far, it’s based on the visual part. This means that if you want to build a program that recognizes how many birds are on a tree or automatically identify all types of plants, you can do it with Lobe, without writing a single line of code!

Mike Matas, co-founder of the company, says that Lobe is not a full-fledged program production tool like PyTorch or TensorFlow. Instead, it is a tool designed for the less familiar, which will give them the right foundations for how an artificial intelligence program works. “People have ideas that they could implement through this technology, but they don’t know how,” says Matas. Lobe allows you to take the first step without prior experience, and helps develop programming logic.
A simple example that can be implemented through the platform is a deep learning algorithm that recognizes letters from Scrabble pieces via a webcam. The order in which such a process is performed is divided into 3 steps. First, a list of samples (datasets) must be created. Then the user must classify the samples according to their content, into folders, that is, in this specific example, to state that an image showing the letter A, contains the letter A. Finally, he enters the classified folders into the program, which in turn studies the images, and learns each letter.
Running the process through Lobe couldn't be easier. The user doesn't need to download anything as everything is done through the browser. There are also templates with ready-made structures from which you can choose depending on what you want to create (more will be added soon).
Of course, as the application is still in beta, there are some flaws and bugs that need to be fixed. Also, some experts are against such applications as they argue that creating programs using only drag and drop tools and not writing code does not provide any substantial knowledge.
Finally, if artificial intelligence is going to change the world, then it is obvious that the more people get involved with this technology, the better. There is a great potential for development and through Lobe anyone can make a start.
