narrow AI

narrow AI

Narrow AI is artificial intelligence applied to a single task such as a chatbot, virtual assistant or self-driving car. Also called "weak AI." Contrast with strong AI. See AGI.
Mentioned in ?
References in periodicals archive ?
Admittedly there are different levels of AI, some still falling into the automation category, but, usually categorised into Narrow AI, Strong AI and Super AI.
What we have today is not general AI, but narrow AI, which is to say AI that is very good at a specific task.
But what happens if domain-specific narrow AI is supplanted by artificial general intelligence that can function, like a human, in any circumstance?
Hammond draws another distinction, as well, between AI designed to accomplish specific tasks ("narrow AI") and systems that can engage in general reasoning ("general AI").
On the other hand, narrow AI was being used to diagnose disease, he said, which was incredibly beneficial.
Robinson writes, "Narrow AI attempts to solve specific problems." He cites Siri, Alexa, and Watson as models, noting, "They are not any form of general AI." He continues his education of the misinformed by clarifying that narrow AI systems are considered "brittle" because they can fail if asked a question that's outside their area of expertise.
Such an approach would take into account concepts around cybernetics, distributed cognition, the limits of narrow AI, and the complexity of human creativity.
"The AI technology we are working with is Narrow AI, which can do one or two tasks better than humans," Zoghbi told a crowd of 40 or so people at a recent AI startup competition.
Narrow AI is important, as it replicates human capability, not behavior.
The lack of a common language among all parties remains a barrier to scalability, as do specific and narrow AI skillsets.
Now, as in today, we are in an era of narrow AI (NAI).