AI CEO Sam Altman downplays wild online speculation from fans about his company's capabilities

在有传言称该公司接近超级智能之后，OpenAI 首席执行官萨姆·奥尔特曼（Sam Altman）淡化了粉丝们在网上对其公司能力的疯狂猜测。

OpenAI CEO Sam Altman has played down wild speculation online by fans about his company's capabilities, following rumors that the company is close to superintelligence.

此前有传言称 ChatGPT 制造商已经秘密实现了通用人工智能（AGI），粉丝们对此感到兴奋不已。据《财富》，包括 Altman 在内的 OpenAI 工作人员上个月一直在努力推广一系列人工智能推理模型和大量更新。随后，一位名叫格温·布兰文 (Gwern Branwen) 的人工智能作家加剧了谣言，声称该公司的新推理模型即将取得巨大突破。

Fans were excited after rumors that the makers of ChatGPT had secretly implemented artificial general intelligence (AGI). According to Fortune, OpenAI staff, including Altman, have been working hard over the last month to promote a series of artificial intelligence inference models and numerous updates. Then an AI writer named Gwern Branwen fueled the rumors, claiming that the company's new inference model was about to achieve a huge breakthrough.

Altman 呼吁粉丝降低对 OpenAI 的期望

Altman urges fans to lower expectations for OpenAI

然而，OpenAI 老板不得不在X 平台上发帖奥特曼告诉粉丝们要降低他们的期望，尽管该公司正在开发一些“很酷的东西”。

However, OpenAI bosses were forced to tell fans to lower their expectations in a post on Platform X, despite the company having some "cool stuff" in the works.

“推特炒作再次失控。我们为您准备了一些非常酷的东西，但请冷静下来，将您的期望降低 100 倍。”

"The Twitter hype is out of control again. We have some really cool stuff for you, but calm down and lower your expectations 100x."

奥特曼。

Ultraman.

社交媒体平台上的猜测一直很高，尤其是 OpenAI 下个月将部署 AGI 的 X 平台。然而，首席执行官表示，他们不会部署任何通用人工智能，也尚未建造它。

Speculation has been high on social media platforms, especially with OpenAI set to deploy AGI's X platform next month. However, the CEO said they won't be deploying any general artificial intelligence and haven't built it yet.

Altman 撰文反驳了有关 OpenAI 下一步行动的猜测。一些狂热者同意奥特曼的观点，尽管他们承认他们对公司的发展感到兴奋，而另一些人则表示“我们的期望保持在基础上真是太好了”。

Altman wrote an article to refute speculation about OpenAI's next move. Some enthusiasts agreed with Altman, though they admitted they were excited for the company's growth, while others said "it's great that our expectations are staying grounded."

“我感觉到你了！控制期望是关键。迫不及待地想看看你们为我们准备了什么，即使下个月不是 AGI。为我们即将推出的酷玩意儿感到兴奋！”安基什写道

"I feel you! Managing expectations is key. Can't wait to see what you guys have in store for us, even if it's not AGI next month. Excited for the cool stuff we have coming soon!" Ankish wrote

其他工作人员也介入，试图平息网上的炒作。该科技公司的推理专家诺姆·布朗表示，社交媒体上有“很多模糊的人工智能炒作”。虽然承认对人工智能的进展有“充分的理由感到乐观”，但布朗强调，许多研究问题仍然需要解决方案，并补充说 OpenAI“尚未实现超级智能”。

Other staff members also stepped in to try to quell the online hype. Norm Brown, an inference expert at the tech company, said there was "a lot of vague AI hype" on social media. While acknowledging that there are "good reasons to be optimistic" about the progress of artificial intelligence, Brown emphasized that many research problems still require solutions, adding that OpenAI "has not yet achieved superintelligence."

他的言论直接挑战了 OpenAI 内部最近的声明。一月份，OpenAI 研究员 Stephen McAleer写道：“当我们不知道如何创造超级智能时，我有点怀念做人工智能研究的日子。” McAleer 的观点表明 OpenAI 至少找到了一条通向所谓的人工智能 (ASI) 的清晰道路。

His comments directly challenge recent statements within OpenAI. In January, OpenAI researcher Stephen McAleer wrote: “I kind of miss the days of doing AI research when we didn’t know how to create superintelligence.” McAleer’s point suggests that OpenAI has at least found a path toward what’s known as artificial intelligence (ASI). clear path.

在加入 OpenAI 之前，布朗曾在 Facebook 人工智能研究中心 (FAIR) 工作，在那里他开发了在扑克和外交等复杂游戏中击败人类玩家的人工智能系统。他在扑克 AI Libratus 等系统上的工作探索了“测试时计算”的概念，表明给人工智能更多的计算时间可以带来更好的游戏动作。

Before joining OpenAI, Brown worked at Facebook Artificial Intelligence Research (FAIR), where he developed artificial intelligence systems that beat human players in complex games such as poker and diplomacy. His work on systems such as the poker AI Libratus explored the concept of "computation at test", showing that giving an AI more computing time can lead to better game moves.

推特炒作再次失控。

Twitter hype is out of control again.

我们不会在下个月部署 AGI，也不会构建它。

We won't deploy AGI next month, nor will we build it.

我们为您准备了一些非常酷的东西，但请冷静下来，将您的期望降低 100 倍！

We have some really cool stuff for you, but calm down and lower your expectations 100x!

— Sam Altman (@sama) 2025 年 1 月 20 日

— Sam Altman (@sama) January 20, 2025

奥特曼是不是自相矛盾？

Is Ultraman contradicting himself?

然而，奥特曼的言论与他之前的言论相矛盾。他说，在人工智能时代，知识渊博、记忆事实或知道去哪里找到它们，远不如学习如何提出好问题那么有价值。

However, Ultraman's comments contradicted his previous statements. In the age of artificial intelligence, being knowledgeable, memorizing facts or knowing where to find them is not nearly as valuable as learning how to ask good questions, he said.

“将会有一种我们仍然真正重视的能力，但它不会是同等程度的原始智力马力，”奥特曼在沃顿商学院组织心理学家亚当·格兰特的“重新思考”播客中告诉沃顿商学院组织心理学家亚当格兰特，在十二月发布的一集中。 “弄清楚要问什么问题比找出答案更重要，”奥特曼补充道。

“There will be an ability that we still really value, but it won’t be the same level of raw intellectual horsepower,” Altman told Wharton on Wharton organizational psychologist Adam Grant’s “Rethink” podcast Business school organizational psychologist Adam Grant, in an episode released in December. “It’s more important to figure out what questions to ask than to figure out the answers,” Altman added.

奥特曼的观察可以用多种方式来解释，向其他人提出很好的问题或为人工智能聊天机器人设计问题，这样他们就会给你需要的答案，这个过程被称为即时工程。

Altman's observation can be interpreted in a variety of ways, asking other people good questions or crafting questions for an AI chatbot so they give you the answers you need, a process known as just-in-time engineering.

CNBC采访时表示，向他人提出清晰、简洁的问题可以表现出同理心或建立可信度dent 莉迪亚·洛根 (Lydia Logan) 在 2024 年 6 月表示，对即时工程师的需求非常旺盛，有些工作的年薪超过 10 万美元。

Asking clear, concise questions of others can show empathy or build credibility, said a CNBC interview. Lydia Logan said in June 2024 that demand for just-in-time engineers is very strong, Some jobs pay more than $100,000 a year.

奥特曼预测，人工智能最终可能会接管大多数工作场所的管理任务，但它不会完全取代人类智力。相反，人们需要帮助技术学习批判性思维技能，以加强论据并提出新的想法，他说。

Altman predicts that artificial intelligence may eventually take over most workplace management tasks, but it won't completely replace human intelligence. Instead, people need to help technology learn critical thinking skills to strengthen arguments and come up with new ideas, he said.

“我确实从必须创造性地推理问题并找出以前没有人想到的答案中获得了最大的职业乐趣，”奥特曼说。

“I really get the most professional joy out of having to reason creatively about problems and come up with answers that no one has thought of before,” Altman said.

“我期望现实中会发生的是，我们将会有一种新的方式来解决难题，”他补充道。

"What I expect will happen in reality is that we will have a new way to solve difficult problems," he added.