In today’s era, self-driving vehicles, robotics and automated home theatres are no longer a magic. This has been the result of innovation and development in computational algorithms and automation. This development has provided solutions for many medical, social issues to name a few. The use is not limited just to these sectors but have innumerous applications in every sector which is driven by data. The use of artificial intelligence is really developed in some sectors while just in initial phases in others.
According to an article Deep learning is the money tree for IT companies in Economic times, the projected revenue share by different sub sects of artificial intelligence technology will be dominated by deep learning (42%), machine learning (21%) and natural language processing (16%). (Source: http://cio.economictimes.indiatimes.com/news/strategy-and-management/deep-learning-is-the-money-tree-for-it-companies/60774469 )
What is Developmental Artificial intelligence and mechanism behind it?
Most of us know about Artificial Intelligence (AI) and its applications. Developmental AI is a new branch of AI which intends to make artificial system that can develop independently like a newborn baby. This system if integrated in a robot will experiment and adapt to surroundings like a new born. The long-term objective of Developmental AI is to create a robot capable of reaching human-level intelligence.
Human brain is the most intelligent of all if we see today and researchers want to imitate it. Developmental AI focuses on creating AI system that can learn a variety of things and develop knowledge and intelligence. Also, the knowledge keeps on changing and expanding based on experiences; that is the motivation behind developing developmental Artificial intelligence.
This process involves autonomous goal construction and playful behavior. It is based on key human cognitive behavior like sense-making, choice making and creativity. These developments will then be leading to mechanisms like brainstorming, learning languages and solving problems.
The most important feature of developmental learning algorithms is to generate self-programming which would allow the data once learned to be executed by the system itself. This feature differentiates developmental learning algorithms from traditional machine learning such as neural networks and reinforcement learning algorithms, which learn values and weights.
Some applications of Developmental Artificial Intelligence
Little AI: Instructional game
It is an instructional game based on Artificial intelligence. It can be used for teaching purposes for both in classroom as well as online learning platforms. The way it functions is so facile that even children can use it. There is a button which player needs to press to control the baby robot. Little AI assists a child in a way by helping him with the tasks that are studied in developmental psychology and constructivist artificial intelligence. The player must master the functioning of robot as well as environment to reach to the highest levels. The environment structure can be learnt by understanding the patterns in array of commands and feedback got from it. The innovative thing about this game is that it showcases the history of interactions i.e. which helps find the regularities of interaction visually. (http://www.sciencedirect.com/science/article/pii/S2352711017300213)
Of all the areas that might be refined by artificial intelligence, the smart home is not any exception. All of us have different needs when thinking about our future smart homes. This seems to be a difficult thing by using classical AI. There is a research being done by using Developmental Artificial Intelligence and Constructivism Theory for making smart homes which has been experimented on earning schemas from a simulated two-weeks home scenario. The agent in this case builds its knowledge in situ through user’s interactions. This interaction process makes it possible to customize to meet user’s personal needs.( http://ieeexplore.ieee.org/document/6529515/ )
APRIL: Educational robotics
Plymouth University’s Centre for Robotics and Neural Systems (CRNS) and A-Lab of the industry partner SoftBank Robotics (Aldebaran) have developed APRIL (Applications of Personal Robotics for Interaction and Learning) which is the first Marie Skłodowska-Curie European Industrial Doctorate (ITN-EID) to train the next generation of researchers and engineers for the emerging field of personal robotics. This will help PhD students to conduct research on developmental and social cognitive systems, along with hands-on experiments on the application of human-robot interaction and assistive systems led by industries. ( http://www.tech.plym.ac.uk/SoCCE/CRNS/APRIL/#About )
Insights into the developmental AI course by Millionlights:
This course will teach you cognitive science background and the programming base to design robots. Olivier Georgeon is the instructor of the course who is a sagacious Researcher in Developmental Artificial Intelligence and inventor of Radical Interactionism, a novel paradigm to implement artificial cognition that keeps knowledge grounded in the robot’s individual experience, which allows open-ended developmental learning driven by intrinsic motivation without ontological presuppositions.
Key points covered in the course:
- The developmental sharks
- Rudimentary Developmental Robot
- Rudimentary self-programming
- Demonstration of developmental learning
- Constructing Ontology from experience of interaction
- Enactive Cognitive Architecture
- Little AI Level 4
We are really excited to inform that the enrolment for our new course in Developmental Artificial intelligence is completely free of cost and also self-paced so you can complete the course according to your suitable time slots and speed. We are confident that with our courses taught by world-class instructors our students are guaranteed an exceptional learning experience!
To enroll for the course visit: https://www.millionlights.org/Course/AboutCourse?id=114&type=&pe=1