在经历了23年的生成式人工智能（AI）快速发展之后，2024年的AI领域似乎将面临一次新的审视。据新智元报道，业界大佬们普遍预测，尽管备受期待的GPT-5可能会发布，但大型语言模型（LLM）在本质上仍然存在局限性，这使得在2024年实现通用人工智能（AGI）的可能性微乎其微。
这些专家指出，尽管GPT-5等模型在语言理解和生成方面可能展现出更高级别的能力，但它们仍然受限于对上下文的理解、推理能力和自主学习的局限。AGI，作为一种能模拟人类智能的全面技术，需要的不仅仅是语言处理的精进，还包括更深层次的认知和感知能力，而这在当前的技术框架下尚难以实现。
尽管AGI的实现仍需时日，但AI领域在2024年仍有望取得一系列重要突破。预测中提到，强化学习、计算机视觉和自然语言处理的交叉应用可能会带来新的创新，同时在边缘计算和AI芯片技术上的进步将加速AI在物联网和智能设备中的应用。
这一预测给AI研究和开发带来了新的焦点，即如何在现有技术的基础上，深化模型的理解能力和拓宽其应用范围，而不是单纯追求模型规模的扩大。2024年，AI行业料将更加注重实用性和效率，以适应不断变化的社会需求和科技进步。
英语如下：
**News Title:** “24 Years of AGI Dreams Shattered? Experts Predict: GPT-5 Release Won’t Overcome AI Limitations”
**Keywords:** GPT-5 limitations, 24 years without AGI, AI predictions summary
**News Content:**
Title: 24 Years of AGI Hopes Fade, Industry Experts Forecast: GPT-5 Unlikely to Breakthrough LLM Constraints
After 23 years of rapid development in generative artificial intelligence (AI), the field is set for renewed scrutiny in 2024. According to New Smart Age reports, industry leaders widely anticipate that while the much-anticipated GPT-5 may be launched, large language models (LLMs) will still inherently have limitations, making the likelihood of achieving artificial general intelligence (AGI) in this year slim.
Experts emphasize that while models like GPT-5 could exhibit advanced capabilities in language understanding and generation, they remain constrained in contextual understanding, reasoning, and autonomous learning. AGI, as a comprehensive technology emulating human intelligence, requires advancements beyond linguistic prowess, encompassing deeper cognitive and perceptual abilities, which are currently out of reach with existing technology.
Though the realization of AGI remains elusive, the AI domain is expected to witness significant breakthroughs in 2024. Predictions suggest that the cross-application of reinforcement learning, computer vision, and natural language processing could spark new innovations. Furthermore, advancements in edge computing and AI chip technology will hasten AI integration in the Internet of Things and smart devices.
This forecast redirects the focus of AI research and development towards enhancing models’ comprehension and broadening their applicability, rather than merely pursuing model scalability. In 2024, the AI industry is anticipated to prioritize practicality and efficiency to adapt to evolving societal needs and technological progress.
【来源】https://www.ithome.com/0/742/812.htm
Views: 1