THE AI BUBBLE POP HAS STARTED...

AI Summary

Mutahar discusses the ongoing 'AI bubble pop,' asserting that the technology is overvalued and its economic correction will resemble the 2008 financial crisis, albeit ultimately beneficial for tech enthusiasts. He highlights how major AI companies like Anthropic are increasing pricing, cutting access, and experiencing a slowdown in growth. Specifically, Anthropic has changed its billing model to charge per-use for features like third-party agents, making it expensive for power users who previously benefited from subsidized $20 plans. Mutahar explains that running large language models (LLMs) is computationally intensive and costly, requiring expensive GPUs and significant electricity, leading to data center delays and cancellations in the US. He criticizes the current state of AI, noting a decline in model quality, increased hallucination, and a general 'AI fatigue' among users who find AI-generated content sloppy and pervasive. Mutahar argues that the only truly useful application for AI is coding, but even that market may not be large enough to sustain current organizations. He points out that many users still fact-check ChatGPT's sources, questioning the utility of AI search over traditional search engines. Despite the challenges, he acknowledges advancements like Google's Turbo Quant, which allows larger models on single GPUs. Mutahar concludes that the bubble pop is a necessary correction, as current AI companies have priced out other technologies and are causing more damage than value, advocating for a future where technology is more accessible and owned by individuals rather than centralized corporations.

Claims Extracted (12)