The AI "Bubble" Is Bursting RIGHT NOW!
In a Nutshell
Stock market declines in AI companies reflect overinflated expectations rather than an impending collapse, and AI fundamentally differs from historical bubbles because it involves real infrastructure, hardware, and revenue-generating businesses rather than pure speculation like tulips. A credit crunch scenario is unlikely since data center orders remain sold out through 2028 and substantial debt isn't tied up in AI like it was in housing. AI deployment continues successfully in enterprises, with 60% of pilots succeeding, making its disappearance improbable regardless of market corrections.
These notes were generated by AI and may contain inaccuracies.
The speaker recently returned from Toronto where they were working with a client in their role as a fractional chief AI officer. The purpose of the video is to address concerns about an AI bubble following stock market declines where Google fell 13% and Microsoft and other major tech stocks dropped.
The speaker clarifies they are responding to those who believe AI is about to implode and disappear entirely, stating this is not going to happen.
Stock market declines are not necessarily signs of doom. There is a distinction between a stock market correction and a proper bubble. A stock market correction occurs when the market falls by 10-20%, and we may be overdue for one. Major tech companies including Microsoft, ServiceNow, and Google are declining not because they are doing anything wrong, but because they are failing to meet inflated expectations.
The housing bubble was driven by cheap debt, bad information, and deregulation. It was catastrophic because substantial debt was tied up in it, causing a credit crunch that prevented other debt from being serviced and created a domino effect. The fear is that if AI resembles this scenario, another credit crunch could occur.
The tulip mania of centuries ago involved speculation on tulips that had no intrinsic utility or value, driven purely by mimetic desire where people wanted tulips because others wanted them. People took on debt to purchase tulips that were perishable goods.
The dot-com revolution and crash is mentioned, noting that Cisco Systems only recently achieved its previous all-time high stock price after 23 years.
Unlike tulips, AI infrastructure includes actual real estate, property, hardware, and profitable businesses. While GPUs have a shelf life, they generate revenue throughout that period. Overpriced stocks are not equivalent to a bubble.
The greatest risk would be if substantial debt and credit were tied up in AI, potentially causing a credit crunch that leads to unemployment spiking to 10% and a three-to-four-year freeze on data center builds. However, this scenario does not appear to be developing.
Elon Musk tweeted that financing for data centers and GPUs is becoming harder to obtain. Simultaneously, conflicting information shows data center orders and infrastructure components including heating, cooling, transformers, and GPUs remain sold out through 2027 and 2028.
Critics like Ed Zitron argue that AI industry dealings are circular, but this mirrors how big oil, big rail, and big steel interacted during the last industrial revolution with robber barons. First movers claim the lion's share of initial value, and markets eventually become saturated, but oil, rail, and steel did not become less important to America.
Motivated reasoning occurs when individuals cherry-pick mechanisms, principles, and data, interpreting them charitably to fit their narrative. This explains how AI doomers operate, as they emotionally oppose AI and want it to fail.
The speaker references their experience working with businesses and enterprises deploying AI, noting that universities and enterprises are contacting them, and AI is not disappearing. When 40% of AI pilots fail, this means 60% succeed.
Some people hold an aesthetic preference that AI should fail and be shut down, but this is not reality. The speaker states they would report this if they believed it would happen.
The speaker has 15 years experience in IT infrastructure, cloud, virtualization, and automation. They followed AI since Google Universal Sentence Encoder around 2015-2016, which was the precursor technology to generative AI and transformers. After GPT-2's release, they recognized it as the path to AGI. Following GPT-3's release, they built their YouTube channel and pivoted their career to AI. They developed GitHub repositories for cognitive architectures and agents during the GPT-3 era, then focused on post-cyber economics and building the fully automated economy.
Mortgage distress levels in America have reached levels reminiscent of the housing crash and Great Recession, occurring independently of AI. People have been predicting a major recession for two years without it materializing. A contagion effect could occur where American debt trouble ripples worldwide, as happened during the Great Depression in the 1920s and 1930s, affecting countries with no connection to American data centers.
The speaker captured these thoughts during a walk while testing a new carbon fiber selfie stick. They are not actively worried about a credit crunch and hope banks, speculators, and venture capital are more cautious so any bubble pop does not result in cancelled mortgages.
Keep David Shapiro in your library
Save the videos and channels worth coming back to, and find them again in one place.
