Telecom's Artificial General Intelligence (AGI) Vision: Beyond the GenAI Frontier

被引:1
|
作者
Chaccour, Christina [1 ]
Karapantelakis, Athanasios [2 ]
Murphy, Timothy [3 ]
Dohler, Mischa [4 ]
机构
[1] Ericsson Inc, Plano, TX 75024 USA
[2] Ericsson Res, S-16480 Kista, Stockholm, Sweden
[3] Ericsson Inc, BA Cloud Software & Serv, Montreal, PQ H4S 0B6, Canada
[4] Ericsson Inc, Adv Technol Grp, Santa Clara, CA 95054 USA
来源
IEEE NETWORK | 2024年 / 38卷 / 05期
关键词
Artificial intelligence; Transformers; Telecommunications; Decoding; Artificial general intelligence; Task analysis; Probability distribution; artificial intelligence (AI); machine learning (ML); causality; semantic communications; explainability; generative AI (GenAI); artificial general intelligence (AGI);
D O I
10.1109/MNET.2024.3425594
中图分类号
TP3 [计算技术、计算机技术];
学科分类号
0812 ;
摘要
This paper unveils the groundbreaking impact of Generative AI (GenAI) as the dawn of a transformative era in 5G/6G networks and beyond. Exploring its disruptive potential across the value chain-from network design to agile and robust automation-we showcase GenAI as a catalyst for innovation and unparalleled efficiency. While tracing its historical journey from conception to practical implementation, the paper positions GenAI not as the sole solution but as the inception of a new era shaping network design, deployment strategies, and synchronous optimization dynamics. We also scrutinize the role of causal AI and transparent frameworks, such as explainable AI and neuro-symbolic AI in fostering trust and seamlessly integrating domain knowledge. Looking ahead beyond GenAI, we envision a future AI landscape composed of semantic communications, collaborative GenAI and discriminative agents. We also examine the challenges related to scalability and complexity that must be overcome to achieve sustainable AI deployments. Finally, we highlight the significance of emerging computing technologies and frameworks, such as quantum and neuromorphic computing, that play a pivotal role in the broader trajectory towards artificial general intelligence.
引用
收藏
页码:21 / 28
页数:8
相关论文
共 50 条