IPA IP 2026 Question 27
A question IPA Japan published in 令和8年度 ITパスポート試験 公開問題 問27, translated into English. ITPEC is based on the same exams.
When AI is classified into “strong AI,” which can make judgments on various problems, and “weak AI,” which can make judgments only on specific problems, which of the following is the most appropriate description of “weak AI”?
Explanation
It shows the result of a judgment without understanding the question or thinking like a human — weak AI processes one specific task and has no human-like mind behind it.
Weak (narrow) AI is built for a particular problem: image recognition, translation, a game, a recommendation. It learns patterns from data and outputs an answer, but it does not grasp the meaning of what it is asked, and it cannot carry its skill over to unrelated problems. Every AI in practical use today is weak AI. Strong (general) AI would have human-equivalent intelligence that understands and reasons across any problem; it does not yet exist.
Answer (b)
Why not others:
- •(a) — intelligence equivalent to a human's, reproduced on a computer and judging as a human does, is the definition of strong AI
- •(c) — an AI smarter than humans that improves itself without help goes beyond even strong AI: superintelligence, the idea behind the technological singularity
- •(d) — applying its own knowledge and abilities autonomously and flexibly to any situation is again strong (general) AI
Key rule: Weak AI = task-specific, produces results without understanding; strong AI = human-like general intelligence that understands and reasons across problems.
Basic Theory · AI-written, may contain errors
This question comes from the past questions IPA Japan publishes, translated into English by ITPEC Practice. ITPEC Practice is an independent study tool and is not affiliated with IPA or ITPEC. See the original paper (PDF) or Report an issue.