OpenAIは"Stawbridy"のAIモデルを公開し,思考能力は向上したが,透明性の懸念は高まった. OpenAI unveils "Strawberry" AI models with improved reasoning skills but raises transparency concerns.

OpenAIは,その新しいAIモデル"Strawreview"(O1-previewとo1-mini)を導入し,思考能力が著しく向上している. OpenAI has introduced its new AI models, "Strawberry" (o1-preview and o1-mini), which show major improvements in reasoning skills. AIは,手続で問題を解決することで,人間のように複雑な作業を分解する,連鎖的な思考方式を用います. The AI uses a chain-of-thought method, breaking down complex tasks like humans do by solving problems step-by-step. この自己形成プロセスは,その推理能力を向上させるが,透明性が欠如しており,悪用や誤認の潜在的懸念を生じさせる. This self-verification process enhances its reasoning abilities but lacks transparency, raising concerns about potential misuse and misinformation.