Artificial Intelligence (AI) programmes have far surpassed humans when it comes to games such as chess. However, when collaborating with AI, Can the same technology get along with people?
In a new study, MIT Lincoln Laboratory researchers sought to find out how well humans could play the cooperative card game Hanabi with an advanced AI model trained to excel at playing with teammates it has never met before. In single-blind experiments, participants played two series of the game: one with the AI agent as their teammate, and the other with a rule-based agent, a bot manually programmed to play in a predefined way.
The results revealed that the scores were no better with the AI teammate than with the rule-based agent. However, humans consistently hated playing with their AI teammate as they found it to be unpredictable, unreliable, and untrustworthy and felt negative even when the team scored well.
The study highlights the nuanced distinction between creating AI that performs objectively well and creating AI that is subjectively trusted or preferred. It may seem those things are so close that there is no daylight between them, but this study showed that those are two separate problems. We need to work on disentangling those.
– Ross Allen, Co-author of the paper
Humans hating their AI teammates could be of concern for researchers designing this technology to one day work with humans on real challenges — like defending from missiles or performing complex surgery. This dynamic, called teaming intelligence, is the next frontier in AI research, and it uses a particular kind of AI called reinforcement learning.
The researchers did not develop either the AI or rule-based agents used in this experiment. Both agents represent the best in their fields for Hanabi performance. In fact, when the AI model was previously paired with an AI teammate it had never played with before, the team achieved the highest-ever score for Hanabi played between two unknown AI agents.
Objectively, there was no statistical difference in the scores between the AI and the rule-based agent. Subjectively, all 29 participants reported in surveys a clear preference toward the rule-based teammate. The participants were not informed which agent they were playing with for which games.
The researchers note that the AI used in this study was not developed for human preference. But the problem is that not many AIs are designed for this purpose. Like most collaborative AI models, this model was designed to score as high as possible, and its success has been benchmarked by its objective performance.
If researchers do not focus on the question of subjective human preference, then they will not create AI that humans actually want to use. It is easier to work on AI that improves clean, objective parameters. It is much harder to work on AI that works in this ambiguous world of human preferences.
Mastering a game such as Hanabi between AI and humans could open up a universe of possibilities for teaming intelligence in the future. But until researchers can close the gap between how well an AI performs and how much a human likes it, the technology may well remain at machine versus human.
As reported by OpenGov Asia, U.S. Scientists have developed AI for various purposes, including a new, automated, AI-based algorithm that can learn to read patient data from Electronic Health Records (EHR). The scientists, in a side-by-side comparison, showed that their method accurately identified patients with certain diseases as well as the traditional, “gold-standard” method, which requires much more manual labour to develop and perform.
