• ai
• ai research

Tasks in current real-world business applications cannot yet be solved by machines alone. Humans and machines working together possess the ability to accomplish complex goals by combining human and artificial intelligence to collectively achieve superior results and continuously improve by learning from each other.

Recent research provides a perspective on this subject.

In the research paper The Future of Human-AI Collaboration the concept of a hybrid intelligence system is described: The main idea of hybrid intelligence systems is that the system consists of human and AI parts and that these ports can co-evolve to improve over time.

It is helpful to bring into mind what AI is:

By this term AI we mean systems that perform ”activities that we associate with human thinking, activities such as decision-making, problem solving, learning”

In this paper hybrid intelligence systems are defined as:

"Systems that have the ability to accomplish complex goals by combining human and artificial intelligence to collectively achieve superior results than each of the could have done in separation and continuously improve by learning from each other."

and the following meta categories (or meta-dimensions) are used to order the different relevant aspects:

• task characteristics,
• learning paradigm,
• human-AI interaction
• AI-human interaction.

This paper also mentions the term "The human in the loop".

Besides the title "The Future of Human-AI Collaboration", the paper has a focus on taxonomy and does not address the future itself in detail. But the taxonomy gives much inspiration for further research and thinking about this future!.

My thoughts on the future and on this subject are:

I think it is interesting to separate design-time from run-time; Humans can be in the loop designing the hybrid intelligence system and in running/using the hybrid intelligence system; This could be an addition to the model given in the paper;

Another thing to take into account is the position of the human in the hybrid intelligence system. Is the human a cog in the ai machine mind or is the ai machine mind a rocketship for the mind of the human?. (or framed as augmentation: is the human intelligence augmented to solve problems it cannot solve alone, is the machine intelligence augmented, or are both augmented).

This also touches "the quality of work" for the humans in such a hybrid system: One can imagine different types of work varying from routine input to highly creative work like modeling or designing based on the output.

Positioning "AI as a rocket ship for the human mind" seems to me a good principle in designing Humane hybrid intelligence systems.

Next Post Previous Post