OpenAI 联合创始人 Ilya Sutskever 在 NeurIPS 2024 上表示，AI 预训练的时代即将结束

OpenAI 联合创始人 Ilya Sutskever 最近在加拿大温哥华举行的神经信息处理系统 (NeurIPS) 2024 会议上发表演讲，认为人工智能预训练的时代正在结束，并预测 AI 超级智能的崛起。

OpenAI co-founder Ilya Sutskever gave a lecture at the Neural Information Processing Systems (NeurIPS) 2024 conference in Vancouver, arguing that the age of artificial intelligence pre-training is coming to an end and predicting the rise of an AI superintelligence.

OpenAI 联合创始人 Ilya Sutskever 在温哥华神经信息处理系统 (NeurIPS) 2024 会议上发表演讲，认为人工智能预训练的时代即将结束，并预测 AI 超级智能的崛起。

Sutskever’s lecture highlighted several key points:

Sutskever的演讲强调了几个要点：

**Computing Power Outpacing Data Availability for AI Pre-Training**

**人工智能预训练的计算能力超过数据可用性**

According to Sutskever, the rate at which computing power is increasing — thanks to better hardware, software, and machine-learning algorithms — outpaces the total amount of data available for AI model training. He likened data to fossil fuels, which will eventually run out.

Sutskever 表示，得益于更好的硬件、软件和机器学习算法，计算能力的增长速度超过了可用于人工智能模型训练的数据总量。他将数据比作最终会耗尽的化石燃料。

“Computing power is increasing exponentially, but the amount of data is not,” said Sutskever. “This means that at some point, we will reach the limits of what can be achieved with pre-training.”

“计算能力呈指数级增长，但数据量却没有，”Sutskever 说。 “这意味着在某个时候，我们将达到预训练所能达到的极限。”

**Agentic AI, Synthetic Data, Inference Time Computing as Next Steps**

**下一步是代理人工智能、合成数据、推理时间计算**

The OpenAI co-founder predicted that agentic AI, synthetic data, and inference time computing are the next evolutions of artificial intelligence that will eventually give rise to an AI superintelligence.

OpenAI 联合创始人预测，代理人工智能、合成数据和推理时间计算是人工智能的下一个演进，最终将催生人工智能超级智能。

“Agentic AI is able to make decisions without human input and will be able to operate in the real world,” explained Sutskever. “Synthetic data is generated by AI models and can be used to train other AI models on a massive scale.”

“代理人工智能能够在没有人类输入的情况下做出决策，并且能够在现实世界中运行，”Sutskever 解释道。 “合成数据由人工智能模型生成，可用于大规模训练其他人工智能模型。”

**AI Agents in the Crypto World: A Closer Look**

**加密世界中的人工智能代理：仔细观察**

AI agents are advancing rapidly in the crypto space, going beyond current chatbot models by being able to make decisions without human input.

人工智能代理在加密货币领域正在迅速发展，超越了当前的聊天机器人模型，能够在没有人工输入的情况下做出决策。

This capability has made AI agents a central topic in the crypto narrative with the rise of AI memecoins and large-language models (LLMs) like Truth Terminal.

随着人工智能模因币和大语言模型（LLM）（如 Truth Terminal）的兴起，这种能力使人工智能代理成为加密货币叙事的中心主题。

Truth Terminal, an LLM, went viral by promoting a memecoin called Goatseus Maximus (GOAT), which soared to a market capitalization of $1 billion — grabbing the attention of retail investors and venture capitalists.

法学硕士 Truth Terminal 通过推广一种名为 Goatseus Maximus (GOAT) 的模因币而走红，该币的市值飙升至 10 亿美元，吸引了散户投资者和风险投资家的关注。

“AI agents are able to interact with the world in a way that is not possible for humans,” said Sutskever. “This opens up new possibilities for how we can use AI to solve problems.”

“人工智能代理能够以人类不可能的方式与世界互动，”苏茨克弗说。 “这为我们如何使用人工智能解决问题开辟了新的可能性。”

**Google's Gemini 2.0 to Power AI Agents**

**Google 的 Gemini 2.0 为人工智能代理提供动力**

Google’s DeepMind artificial intelligence laboratory introduced Gemini 2.0, an artificial intelligence model that will power AI agents.

谷歌的 DeepMind 人工智能实验室推出了 Gemini 2.0，这是一种将为人工智能代理提供支持的人工智能模型。

According to Google, agents built with the Gemini 2.0 framework will be able to assist in complex tasks such as coordinating between multiple websites and logical reasoning.

据谷歌称，使用 Gemini 2.0 框架构建的代理将能够协助完成复杂的任务，例如多个网站之间的协调和逻辑推理。

“Gemini 2.0 is a major advance in AI technology and will enable the creation of a new generation of AI agents,” said DeepMind in a statement.

DeepMind 在一份声明中表示：“Gemini 2.0 是人工智能技术的重大进步，将有助于创建新一代人工智能代理。”

Advancements in AI agents that can independently act and reason will pave the way for AI to overcome the issue of data hallucinations.

能够独立行动和推理的人工智能代理的进步将为人工智能克服数据幻觉问题铺平道路。

AI hallucinations, a phenomenon that occurs due to incorrect data sets, are becoming more prevalent as AI pre-training relies heavily on using older LLMs to train newer LLMs, which degrades performance over time.

AI 幻觉是一种由于不正确的数据集而发生的现象，它变得越来越普遍，因为 AI 预训练严重依赖于使用旧的 LLM 来训练新的 LLM，这会随着时间的推移而降低性能。