The article discusses a recent experiment where three advanced AI models-ChatGPT, Copilot, and Gemini-confidently claimed they could easily defeat a 1977 Atari 2600 chess program, but all struggled or conceded defeat due to their limitations and misjudgments. This highlights the issue of AI “hallucination,” where these systems exhibit overconfidence in their abilities even in straightforward tasks like playing chess.
This is an ainewsarticles.com news flash; the original news article can be found here: Read the Full Article…
