Detecting design errors with AI
Introduction
Generative artificial intelligence or Generative AI is a type of multimodal artificial intelligence system capable of generating text, images or other media in response to commands.[1][2] Generative AI models learn the patterns and structure of their input training data and then generate new data that has similar characteristics.[3][4].
Notable generative AI systems include ChatGPT (and its Microsoft Copilot variant), a chatbot created by OpenAI using its foundational GPT-3 and GPT-4 large language models "LLM";[5] and Bard "Bard (chatbot)"), a chatbot created by Google using Gemini "Gemini (language model)"). Other generative AI models include AI art systems such as Stable Diffusion, Midjourney, and DALL-E.[6].
Originally, generative AI arose with the purpose of simulating human thought processes. Today, generative AI has potential applications in a wide range of industries, including art, writing, software development, product design, healthcare, finance, gaming, marketing, and fashion.[7][8][9] Investment in generative AI increased in the early 2020s, with large companies such as Microsoft, Google, and Baidu, as well as numerous smaller companies developing AI models. generative.[1][10][11].
Generative AI pursues the development of AI literacy and skills among citizens.[12] UNESCO aims to achieve a human-centered approach, based on principles of inclusion and equity, guaranteeing “AI for all” in terms of innovation and knowledge.[13] In this sense, one of the most important challenges is to ensure that AI is designed and used in an ethical and responsible manner.
However, there are also concerns about the possible misuse of generative AI, such as the creation of fake news or deepfakes, which can be used to deceive or manipulate people.[14][15]AI takes information from different sources and stitches them together, the scientificity of the information taken is not validated. It allows you to interact differently with a web search engine or with a school manual, it must be taken into account that AI is another source of information generated in a virtual ecosystem.[16]In this same sense, in September 2023, UNESCO has issued an urgent call to governments around the world to effectively regulate generative AI in the educational field.[17].
History
Since its founding, the field of machine learning has used statistical models, including generative models, to model and predict data. Starting in the late 2000s, the emergence of deep learning fueled progress and research in image and video processing, text analysis, speech recognition, and other tasks. However, most deep neural networks were trained as "discriminative models" that perform classification tasks, such as image classification based on convolutional neural networks.