Talk:AI bubble

I have added the POV template to this article because it doesn't have any information about the speculation that there is no current AI bubble. ―Panamitsu (talk) 11:08, 26 August 2025 (UTC)

I think it was removed Arkoftwo4 (talk) 12:38, 1 October 2025 (UTC)

Yes, I removed it because I wasn't sure if I was correct in adding it. ―Panamitsu (talk) 20:45, 1 October 2025 (UTC)

I fixed some of the grammar on this article because there were many typos that would be considered hard to read by a person reading this Wikipedia article TheNewOne62 (talk) 15:27, 14 October 2025 (UTC)

These aren't "typos", but... Anyways, unsure whether to accept the change from New Zealand English to American English, but ah hell, Wikipedia is a US-based establishment, isn't it? (Learn more please about English-language spelling reform and American and British English spelling differences if you can.) George Ho (talk) 18:35, 14 October 2025 (UTC)

No we don't do that for no reason. See MOS:RETAIN. ―Panamitsu (talk) 02:29, 15 October 2025 (UTC)

I understand you're from New Zealand, but using a small country's variety of the language on an article completely unrelated to it is a bit silly. American or British English is the most convenient for an encyclopedia. Staryu★ 00:55, 4 November 2025 (UTC)

I didn't learn until now about the term "market correction" when I watched a CNBC interview with Cathie Wood, an Ark Invest CEO. Wood said there won't be an AI bubble but rather market (or stock) correction on AI investments. Others have said the same. (Source: CNBC article) George Ho (talk) 04:32, 31 October 2025 (UTC)

For what it's worth it could be both. We'll just have to see. It might be worth including if you want to add it in the Opposing Views section Catboy69 (talk) 13:29, 31 October 2025 (UTC)

Are "Marxist" interpretations necessary on a largely speculative article? Marxism doesn't even make an appearance on the Great Recession or Dot-com bubble pages, which are the pages/subjects most relevant to this one. It seems to be WP:Fringe, at least for now. Catboy69 (talk) 20:56, 11 February 2026 (UTC)

I've just now reverted Jonathonrobust's additions of Marxist stuff and then moved the references down to "Further reading" section. George Ho (talk) 21:22, 11 February 2026 (UTC)

Much of the article is framed as speculation, which would be valid for the time it was wrote, but now evidence towards the bubble has emerged and I think it should be updated.

While most ai-investing companies haven’t yet reported downturns, OpenAI (https://techstrong.ai/articles/openais-financial-crisis-a-1-4-trillion-gamble/) and Microsoft (https://futurism.com/artificial-intelligence/microsoft-ai-efforts-faceplanting) could act as good examples of such. ~2026-13038-54 (talk) 02:06, 28 February 2026 (UTC)

Let's not considered this actually happening yet. The one about OpenAI says this: The analyst rejects characterizations of an AI bubble, pointing to five key factors that suggest the technology’s growth trajectory remains in its early stages. The Futurism article doesn't mention "AI bubble". George Ho (talk) 04:23, 28 February 2026 (UTC)

People have been talking about whether or not Sora shutting down means the AI bubble is bursting. Whatever you think, there are now some articles on this, so I think it's worth mentioning here. IdfbAn (talk) 19:58, 27 March 2026 (UTC)

Added Mary McNamara's assertion that Sora's demise won't mean AI bubble or a domino effect... yet. George Ho (talk) 21:14, 27 March 2026 (UTC)

Regarding this revert: The argument seems to be that because it was in 2024, it belongs in the article.

Regardless of when it was published, it's just one of countless opinions on the topic, and citing a quote from a prolific columnist's own work is arbitrary. It's not even close to the first example of someone talking about an AI bubble. For example, with a perfunctory search I found a blog post from February 2023. (The blog post is interesting, but it's not a reliable source, it's to demonstrate the issue). Is there some reason this one Motley Fool columnist's quote is significant? Grayfell (talk) 20:27, 8 April 2026 (UTC)

I have just now shoved the Motley Fool piece down to the "Further reading" section. Seems that the editor wants it included totally. George Ho (talk) 21:20, 8 April 2026 (UTC)

Thank you for bringing this up for discussion. The purpose of the "History" section in this article should be to provide all the relevant context about the genesis of the idea that an AI Bubble might be occurring. Otherwise the article incorrectly suggests that no one was foreseeing an AI Bubble and that the DeepSeek moment was the first time people questioned the market values of the AI companies. I'd be happy to include the blog post you found, since its author, Joshua Brown, does seem like a notable and respected commentator in this space. The fact that The Motley Fool wrote an entire article specifically about the AI Bubble in late 2024, mentioning it by name, and correctly timing the market correction in early 2025, makes it extremely relevant to how that term began to be used, and therefore it belongs chronologically before the 2025 paragraph, not as a suppressed dangling footnote. Please reconsider your objection to including it. ~2026-20098-04 (talk) 06:44, 9 April 2026 (UTC)

Unfortunately, your use of the Motley Fool piece was just too basic, IMO. As I see, the author wrote down reasons to support the theory. Why not summarize the reasons the author made? George Ho (talk) 06:52, 9 April 2026 (UTC)

That's a great point, actually. Another source (USA Today) for the same article clarifies that the analyst is giving 4 distinct reasons for his assessment. I can create a summary of his reasons, so the paragraph would look like this instead:

At the end of 2024, Sean Williams writing for The Motley Fool published a prediction that "The Artificial Intelligence (AI) Bubble Will Burst in 2025." The four reasons he gave were: parabolic rallies have always historically come to an end, increasing the supply of GPUs would reduce Nvidia's profit margin, US government tariffs and export controls could reduce GPU sales, and the current price to sales ratio for AI stocks is larger than for other companies.

How does that look? ~2026-20098-04 (talk) 09:23, 10 April 2026 (UTC)

You're just copying section headers, aren't you? If you're unable to summarize his piece well, then... let's not use the piece but rather let another editor (who is not you) read it and figure out how to incorporate it. George Ho (talk) 11:52, 10 April 2026 (UTC)

No, I tried to reduce the content of each section down into a short summary, which is the same process that Williams did when he wrote the section headers for his article. I did try to make sure that my summaries aligned with his wording though, as I didn't want to be accused of contributing original research. If you think you can come up with a better summary then you're welcome to suggest it here, and I'll give you my own constructive criticism, but right now it seems like you are creating your own policies and applying personal taste to what is a simple factual addition to this article. ~2026-20098-04 (talk) 16:46, 11 April 2026 (UTC)

Excuse me? This article was previously nominated for deletion; the result was "no consensus". More likely, moving portions into other articles might be slightly likely. With all the AI talk heating up, seems that trying to incorporate the Motley Fool into the content would, IMO, add more fuel to the fire, wouldn't it? Indeed, not just the Motley Fool, there should be other past predictions, right? George Ho (talk) 17:15, 11 April 2026 (UTC)

If a reliable, independent source mentions Williams's 2024 column or Brown's 2023 blog post, let's use that to provide context. Without that context, this is a subtle form of original research since we're picking sources based on convenience or our own opinions, instead of due weight or lasting significance. As I said, the blog post was after a perfunctory search, and as a blog post, it's not a reliable source. Explaining that people have identified this issue from the beginning would be a good thing, so hopefully we can do this without cherry picking and without original research. Grayfell (talk) 18:40, 9 April 2026 (UTC)

I take your point about Brown's blog post not being an independent source, but his Wikipedia article makes clear that he is a qualified expert in this field, and it mentions his blog by name. Perhaps his post was just a fringe idea when it was first published, though, so I shouldn't consider it a key part of the history of this term. After all, he didn't publish it as the official position of Ritholtz Wealth Management, but just as personal opinion. In any case, I still think it is worth including it in the "Further reading" section, for people who want to understand the full evolution of the usage of this term, and so that the article doesn't implicitly give the wrong impression that concerns about an AI Bubble only began at the end of 2024.

As for the William's article, I don't think you can claim that USA Today, Yahoo Finance, and The Motley Fool have conflicts of interests. The Wikipedia:IS policy applies this standard: "The newspaper is not owned or controlled by the company. Therefore, it is an independent source". Removing valid sources to focus on one preferred narrative is just as much cherry picking as including irrelevant details, so I hope you want to avoid both of those extremes. ~2026-20098-04 (talk) 10:01, 10 April 2026 (UTC)

This is not about COI, this is about due weight, context, and neutrality. Williams's column was from a The Motley Fool, which is a financial advice site, and the discussion of the bubble was primarily a framing to give advice on Nvidia stock at that specific time. Without that context, this is cherry picking, and with that context, this is excessive.

The internet is an all-you-can-eat buffet of opinions. Or to put it another way, Wikipedia:Opinions are like arseholes. The way to demonstrate that this opinion is important enough to include as an example is via outside sources which describe it as an example.

You will not find much support among other editors for adding blog posts to 'further reading' sections, for a lot of different reasons. Grayfell (talk) 19:34, 10 April 2026 (UTC)

I agree that the article was primarily (though not exclusively) about Nvidia's stock price, but that's also a major focus of the rest of the History section. It shouldn't be surprising that the biggest company in the world gets most of the attention for an article like this, and it's also not surprising that Williams focused on that stock when talking about the bubble that he was noticing.

You may also like to consider WP:RS, which says "Wikipedia articles should be based on reliable, published sources, making sure that all majority and significant minority views that have appeared in those sources are covered". We have a professional analyst talking about an AI bubble, published on reputable news sites, at an important moment early in the history of the phenomenon, which was not even controversial, as you've noted the idea had already been around for a year by that time.

It feels like you're trying to censor history by preventing any mention of this term which pre-dates your personally preferred narrative that the bubble was only noticed in 2025. ~2026-20098-04 (talk) 17:14, 11 April 2026 (UTC)

Not to put to fine a point on it, but I have already considered WP:RS many times. I would like to explain that the bubble has been obvious almost since its beginning (I did mention the 2023 blog) but we need to cite and neutrally summarize reliable sources to do this. We can't just pick and choose unreliable sources like WP:FORBESCON, blogs, or Motley Fool columns. Any quote from an unreliable source is going to be arbitrary. This is my main objection. These quotes do not add enough info to justify thee space they occupy based on their weight and reliability.

The use of contemporaneous sources is a recurring issue with how Wikipedia deals with history. These examples are all primary sources for the author's opinions/predictions. As a tertiary source, Wikipedia focuses on summarizing WP:SECONDARY sources, so ideally, we need sources about this history.

Grayfell (talk) 19:08, 11 April 2026 (UTC)