Bununla ilgili yazdığım İngilizce yazıyı kopyala yapıştır yapıcam, gelişmiş yapay zekaları daha gelişmiş beyinler daha iyi kullanır.

There are two testing methods that require a high test environment and generally require more manpower as you develop an AI. And in these tests, especially the 1st test, unfortunately, AIs such as deepseek perform poorly because they focus entirely on performance.

1. SimpleQA (Correct):
- SimpleQA is a dataset for evaluating simple question-answer systems.
- This dataset is often used to test the ability of natural language processing models to find the correct answer to a given question.
- Questions are usually short and clear, and answers are usually a single word or short phrase.
- Example: "Which country is Paris in?" to the question "France".

2. FRAMES (Acc.):
- FRAMES is a dataset used to evaluate dialog systems and chatbots.
- This dataset simulates how users interact with a dialog system to complete a specific task.
- FRAMES measures the ability of dialog systems to understand user intentions, follow context, and provide appropriate responses.
- Example: A dialog between a user and a chatbot to make a restaurant reservation.

Gemini 1.5 Flash generally performs best in these tests, while Chatgpt outperforms other advanced AIs with its advanced AIs.
This makes it pointless to use advanced AI in chatbots and other similar things where there are stupid users who are ignorant about AI.

#ai #deepseek
@cleverestech