We’ve heard it from Gary Smith and Jeffrey Funk. But now, once again, from AI analyst Gary Marcus: The AI bubble created, in part, by Large Language Models (LLMs) or chatbots is nearing its peak:
我们从加里·史密斯和杰弗里·芬克那里听到了这一点。但现在，人工智能分析师加里·马库斯（Gary Marcus）再次指出：部分由大型语言模型（ LLMs ）或聊天机器人创造的人工智能泡沫已接近顶峰：

The economics are likely to be grim. Sky high valuation of companies like OpenAI and Microsoft are largely based on the notion that LLMs will, with continued scaling, become artificial general intelligence. As I have always warned, that’s just a fantasy. There is no principled solution to hallucinations in systems that traffic only in the statistics of language without explicit representation of facts and explicit tools to reason over those facts.经济形势可能会很严峻。 OpenAI 和微软等公司的天价估值很大程度上是基于这样的理念：随着规模的不断扩大， LLMs将成为通用人工智能。正如我一直警告的那样，这只是一个幻想。对于仅传输语言统计数据而没有事实的明确表示和对这些事实进行推理的明确工具的系统中的幻觉，没有原则上的解决方案。

LLMs will not disappear, even if improvements diminish, but the economics will likely never make sense: additional training is expensive, the more scaling, the more costly. And, as I have been warning, everyone is landing in more or less the same place, which leaves nobody with a moat. LLMs such as they are, will become a commodity; price wars will keep revenue low. Given the cost of chips, profits will be elusive. When everyone realizes this, the financial bubble may burst quickly; even NVidia might take a hit, when people realize the extent to which its valuation was based on a false premise.即使改进减弱， LLMs也不会消失，但其经济学可能永远没有意义：额外的培训是昂贵的，规模越大，成本就越高。而且，正如我一直警告的那样，每个人都或多或少地落在同一个地方，这使得没有人拥有护城河。像他们这样的LLMs将成为一种商品；价格战将使收入保持在低水平。考虑到芯片的成本，利润将难以捉摸。当每个人都认识到这一点时，金融泡沫可能会很快破灭；当人们意识到其估值在多大程度上是基于错误前提时，即使是英伟达也可能会受到打击。

“CONFIRMED: LLMs have indeed reached a point of diminishing returns,” November 9, 2024“已确认： LLMs确实已达到收益递减点”， 2024 年 11 月 9 日

Maybe the reason there is no principled solution to AI hallucinations is that we cannot create artificial human minds. Maybe this will be a good opportunity to test that.
也许人工智能幻觉没有原则性解决方案的原因是我们无法创造人造人类思维。也许这将是一个很好的测试机会。

You may also wish to read: Model collapse: AI chatbots are eating their own tails. The problem is fundamental to how they operate. Without new human input, their output starts to decay. Meanwhile, organizations that laid off writers and editors to save money are finding that they can’t just program creativity or common sense into machines.
您可能还想阅读：模型崩溃：人工智能聊天机器人正在吃自己的尾巴。这个问题对于他们的运作方式至关重要。如果没有新的人类输入，它们的输出就会开始衰减。与此同时，为了省钱而解雇作家和编辑的组织发现，他们不能仅仅将创造力或常识编入机器中。