谷歌DeepMind的首席执行官Demis Hassabis近期在接受《连线》(WIRED)杂志采访时表达了对人工智能(AI)未来发展持乐观但谨慎的态度。他认为,尽管OpenAI的ChatGPT近期在科技界掀起热潮,但AI领域最重大的突破仍未到来,而这并不仅仅关乎硬件的提升,如芯片技术的进步。
Hassabis指出,实现通用人工智能(AGI)的道路上,需要的不仅仅是大规模的计算能力。他强调,创新将是推动AI发展至关重要的因素。回顾DeepMind的历史,早在AlphaGo时代,公司就已经在代理学习、强化学习和规划等领域开始了深入研究。这些技术的进步,Hassabis认为,对于构建更加智能、适应性强的AI系统至关重要。
Hassabis的言论揭示了AI研究的多维度性,暗示未来的突破将可能来自对现有技术的深度挖掘和综合应用,而不仅仅是硬件层面的竞赛。这一观点为AI领域的未来发展提供了新的思考方向,也预示着在追求AGI的道路上,科研人员需要在算法和理论层面做出更多创新性的贡献。
英语如下:
News Title: “DeepMind CEO: Major AI Breakthroughs Lie Ahead Beyond Chip Advances”
Keywords: AI breakthroughs, DeepMind, General Artificial Intelligence
News Content: Demis Hassabis, the CEO of Google’s DeepMind, recently expressed a cautious yet optimistic outlook on the future of artificial intelligence (AI) in an interview with Wired magazine. He believes that while OpenAI’s ChatGPT has sparked excitement in the tech industry, the most significant breakthroughs in AI are yet to come – and these will not be limited to improvements in hardware, such as advancements in chip technology.
Hassabis underscores that achieving General Artificial Intelligence (AGI) requires more than just massive computational power. He emphasizes the importance of innovation in driving AI’s progress. Looking back at DeepMind’s history, the company started delving into deep research in areas like agent learning, reinforcement learning, and planning as early as the AlphaGo era. Hassabis contends that the advancements in these technologies are crucial for developing more intelligent and adaptable AI systems.
Hassabis’s remarks highlight the multidimensional nature of AI research, suggesting that future breakthroughs might stem from a deep exploration and integration of existing technologies, rather than just a race in hardware. This perspective offers a new angle for contemplating the future of AI and indicates that in the pursuit of AGI, researchers will need to contribute more innovative work at the algorithmic and theoretical levels.
【来源】https://www.wired.com/story/deepmind-ceo-demis-hassabis-interview-artificial-intelligence-scale/
Views: 1