論文研究關鍵字

首頁Home › 人工智慧與新興科技 AI & Emerging Tech

NO.802

生成式AI

Generative AI

1這是什麼What is it?

生成式 AI(generative AI)指能依據訓練資料的分布,產生新文字、圖像、音訊、影片或程式碼的模型,核心架構包括 GAN(2014)、變分自編碼器、擴散模型與 Transformer(Vaswani 等,2017)。與以辨識與分類為主的判別式模型不同,它學習資料的生成機制;其輸出具機率性,因此帶有幻覺(hallucination)與偏見的風險。

Generative AI refers to models that produce new text, images, audio, video or code according to the distribution of training data, with core architectures including GANs (2014), variational autoencoders, diffusion models and the Transformer (Vaswani et al., 2017). Unlike discriminative models centered on recognition and classification, it learns the generative mechanism of data, and its outputs are probabilistic and thus carry risks of hallucination and bias.

2常被用在哪種研究問題Typical research questions

影響/關聯型Effect / association比較型Comparative

常見問法:生成式 AI 如何影響設計創意與品質?

示範寫法Example

生成式 AI 提供新的創作工具。本研究比較使用與不使用生成式 AI 的設計歷程,並以專家評分與訪談評估差異。

Generative AI offers new creative tools. This study compares design processes with and without generative AI and assesses differences through expert ratings and interviews.

撰寫內容,非真實研究結果,請與指導教授討論後改寫。

Written as a model example, not a real finding. Please revise with your advisor.

3怎麼測它How to measure it

比較有/無 AI 輔助的作品+專家評分

Measurement notes are written in Chinese. Switch to 中/EN to read them.

4代表文獻Key references

英文論文(2000 年後)English paper (post-2000)

Oppenlaender, J. (2022). The creativity of text-to-image generation. MindTrek '22, ACM.

中文論文(2000 年後)

中文論文為預填檢索連結,請自行挑選一篇 2000 年後的期刊論文並核對作者、年份、出處後再引用。

英文引用來自模型整理,引用前請務必核對並閱讀原文。English citations are compiled by a model. Verify and read the original before citing.

‹ 人工智慧Artificial IntelligenceAI繪圖 ›AI Image Generation