It all starts with Efficiency, the first level of the framework. Level-one AI can carry out specific functions through automation that make user interactions more efficient and effective. Users can employ simple and easy voice commands for various tasks like stocking up on orange juice or simply asking for the time. Currently available with many voice recognition and AI-enabled products on the market today, such as the LG ThinQ air conditioner, many people may find some scenes depicted in this video familiar.

Personalization is represented in the second level, at which AI can recognize patterns and regularities from past user interactions and use them to optimize and personalize functions for a specific person. In comparison to level one, level-two AI can order your orange juice or book a taxi before you even ask and can further identify and recommend your favorite TV channels. Through unique pattern and trait learning, the advanced AI found in this second level can improve and simplify interactions for a truly personalized user experience.

The third level is based on Reasoning and envisions AI based on causality learning that looks at the ‘why’ behind users’ actions to predict and promote better outcomes for consumers. Here, AI goes beyond surface level interactions to actively perform not only the requirements but also make recommendations such as suggesting a light breakfast before a stressful presentation or the perfect restaurant for dinner with friends. By identifying underlying motives, level-three AI not only understands but also acts on its insight to provide optimal user experiences.

The fourth and final level is Exploration. Based on experimental learning, the most advanced AI form is best explained in the video, which illustrates a future where AI generates thoughtful ideas and suggestions all on its own. Although an innovation for the far future, level-four AI is capable of original research, enough to formulate and test intelligent hypotheses to discover conclusions, such as integrating mindfulness exercises and new yoga routines into a stressed user’s everyday routine. Rather than the user having to maneuver the AI to perform a certain way, level-four AI works in tandem with the user, facilitating the experimentation of new ideas to achieve the best outcomes.

“From Efficiency to Exploration, our vision for AI innovation is meaningful for both the industry and its customers,” remarked Dr. I.P. Park, President and Chief Technology Officer of LG Electronics, “It enables and encourages us to come up with more daring, more innovative ideas for the future.”

Already shared with many partners, the framework received praise from leading figures across industry and academia. “It is our responsibility to drive the development of the field of AI towards a future in which the technology is being harnessed in a manner that is beneficial for both individuals and society as a whole,” said Yoshua Bengio, the Turing award winner, founder and scientific director of Mila, the research institute partnership between the Université de Montréal and McGill University with Polytechnique Montréal and HEC Montréal.

Further voices in support of the AIX levels included Sanjay Dhawan, Chief Executive Officer of automotive software company Cerence, who encouraged “all other participants in the AI ecosystem to adopt this framework”, as well as Sangbae Kim, Associate Professor of Mechanical Engineering at MIT leading the MIT Biomimetics Robotics Lab since 2012, who defined the framework as an effort to "recapitulate the past, present, and the future of AI."

For more information on the transformational innovations of LG ThinQ, please visit lg.com/us/lg-thinq.