實例，chatGPT解數學競賽題相當厲害，但分數計算出錯

娃參加Beestar的小學初中網上數學競賽，第二輪有幾道題挺難，我看了也不確定，索性下來用正時髦的chatGPT試試，結果讓我大開眼界，chatGPT幾秒鍾解五年級競賽難題， 並且給出詳細解題步驟。但chatGPT也在簡單的分數計算出錯，莫名其妙。。 分享一下：

 題目 （5年級競賽難題)：

Box A contains 5 red balls and 7 green balls. Box B contains 3 red balls and 2 green balls. A box is chosen at random, and one ball removed at random from this box. Given that the ball removed was green, what is the probability that it came from box A?

這題目其實家長碰上都得仔細想想。輸入chatGPT， 幾秒鍾給出超詳細的解題過程：

Let's use Bayes' theorem to solve this problem:

Let A be the event that the ball was drawn from box A, and G be the event that a green ball was drawn. We want to find P(A|G), the probability that the ball came from box A given that it was green.

By Bayes' theorem:

P(A|G) = P(G|A)P(A) / P(G)

where:

P(G|A) is the probability of drawing a green ball given that the box is A. This is 7/12 since box A has 7 green balls out of 12 total balls.

P(A) is the prior probability of choosing box A, which is 1/2 since either box could have been chosen at random.

P(G) is the total probability of drawing a green ball, which can be calculated using the law of total probability:

P(G) = P(G|A)P(A) + P(G|B)P(B)

where P(G|B) is the probability of drawing a green ball given that the box is B, and P(B) is the prior probability of choosing box B, which is also 1/2.

We can calculate P(G|B) as 2/5 since box B has 2 green balls out of 5 total balls.

Plugging these values into Bayes' theorem, we get:

P(A|G) = (7/12) * (1/2) / [(7/12)(1/2) + (2/5)(1/2)]

≈ 0.636

Therefore, the probability that the ball came from box A given that it was green is approximately 0.636 or 63.6%.

太厲害了！chatGPT 真嚇我一跳。但我拿著結果在Beestar卻沒有對應的選擇答案。後來發現chatGPT竟把自己給出的簡單計算算錯了，(7/12) * (1/2) / [(7/12)(1/2) + (2/5)(1/2)] = 35/59 ≈ 0.593， 不是chatGPT給出的0.636。 難解題思路步驟都對了，最後簡單計算卻錯了，讓我無語，挺滑稽！搞AI的大牛們解釋一下，是不是AI 分數計算能力有點弱？

 

所有跟帖： 

• ChatGpt 正確率不到50% -Lucky2022- ♂ (114 bytes) () 03/19/2023 postreply 13:49:42

• ChatGPT是生成答案，而不是計算出答案。 -光猜不想- ♂ (180 bytes) () 03/20/2023 postreply 13:08:58