While artificial intelligence is the buzzword of the moment with everyone talking about its attributes and a future powered by AI, new research from digital transformation firm Orion Business Innovation aimed at testing the industry’s understanding of AI’s context and history reveals that enthusiasm for AI exceeds knowledge of its origins and fundamental mathematical constructs.
The AI quiz, which was completed by 618 IT and data analytics professionals, found that only 2 percent of respondents—a total of just 12 people—got nine or 10 questions correct. The implication is that there’s great interest in AI—but without great knowledge.
“It’s not too surprising that the attention to AI outweighs broad understanding of its origins and principles,” said Raj Patil, Orion CEO, in a news release. “AI appears to be a shiny new object. The reality is that AI is a body of mathematical and operational principles that have evolved over more than 60 years. AI has become more powerful by the confluence of inexpensive big data storage, large data sets, accelerated computing speeds and machine learning algorithms. We hope that the results of this simple quiz will inspire more people to learn more about AI.”
Answers reveal knowledge gaps
The results further demonstrated the gap in background knowledge with almost half of respondents (49.6 percent) answering five or less questions correctly. A few key highlights of the results include:
What does the Church Turning thesis say?
Forty-three percent answered correctly with computers can simulate formal reasoning, while 20 percent believed it was that computers can mirror human behavior.
An intelligent agent ‘perceives its environment and takes action’
Thirty-seven percent of respondents answered correctly, while 29 percent believed that it “finds its own pathways.”
Which element is not an AI learner?
Forty-three percent answered correctly with “fitness function,” while “decision tree” and “nearest neighbor” both received 15 percent of the votes.
Morovec’s Paradox
Thisis defined as the notion that simple human skills are the most difficult to program, which 46 percent of respondents answered correctly. Twenty-four percent believed it was “the simplest answer is usually the right one.”