A critic claiming AI is a scam

‌
To be honest, I had some doubts myself.
They say there is no fundamental way to fix hallucinations...
I thought the programming industry's productivity must have increased overwhelmingly due to AI,
but I heard there are claims that productivity isn't actually that high because people have to spend extra time reviewing code line by line just in case due to those hallucinations... It's a very subtle issue..
But a truly massive scam prevents any criticism while the public is in the midst of enthusiasm and hype.
They act as if to say, "What do you know?"
When someone is at the peak of their popularity, their fandoms use that person's influence to prevent anyone from criticizing them.
But even this summary below was generated by AI..
It's obvious that AI itself shouldn't be completely dismissed and it is indeed amazing,
but I wonder if current AI stocks have been driven up by the public buying without looking at the prices.
It's a sensitive topic, but that's why I wonder if there might be a bubble..
Even Warren Buffett said that no matter how good a stock is, buying it at an expensive price is a bad investment.
If this really turns out to be a scam,
later on, people might go... "Hmm...? Come to think of it, it wasn't actually that great?" and their perspective might shift abruptly.
---
This is a summary of the main contents of the YouTube video you provided.
This video is from 'The Diary Of A CEO' podcast, featuring tech critic and PR expert Ed Zitron, who strongly criticizes the current Generative AI craze and bubble in an interview.
Summary of Key Points
1. "Generative AI is close to a con" [00:00]
Ed Zitron argues that Big Tech companies are deceiving the public and investors by greatly exaggerating the capabilities and economic value of AI.
He points out that while AI is packaged as 'magic' that can replace all jobs or cure cancer, it is actually just simple cloud software that incurs massive costs and frequent errors.
2. Unsustainable AI revenue structures and losses [00:44]
Massive deficits: Major AI companies such as OpenAI and Anthropic are recording astronomical losses (e.g., OpenAI recorded losses of several billion dollars or more last year).
Internal circular investment: He highlights the structural risk where Big Tech (Microsoft, Amazon, Google, etc.) provides funds to loss-making AI startups, and those startups in turn use the Big Tech companies' cloud/GPU servers [04:48].
Actual unit cost issues: He explains that compared to subscription fees in the $20–$200 range, the computing costs (token usage costs) are much higher, creating an imbalanced structure where the more users use it, the greater the company's loss [13:02].
3. Representative delusions surrounding the AI bubble
The delusion of job replacement: There is insufficient economic data to support the claim that AI is immediately and completely replacing human labor [00:55].
Exaggerated productivity and Hallucination risks: In fields where accuracy is essential, such as development, medicine, or finance, AI's hallucination problem is fatal, and there is a benchmark illusion where figures only appear correct on the surface [25:50].
The pretext of the US-China AI war: He criticizes the phenomenon of fueling national competition under the pretext of technology dissemination, leading to massive amounts of capital being poured into GPU purchases [01:06].
4. Conclusion and Message
He emphasizes that the excessive bubble in the tech industry risks reaching its limit around 2027, and that rather than being swept away by a short-term boom, a critical perspective and a realistic understanding of the technology are required [02:27:25].