Consider long-term investments in non-tech companies like banks (Morgan Stanley, JP Morgan) and industrials (Deere, Caterpillar) that are effectively using AI to boost productivity, as this is a key theme for 2026 and beyond. Apple (AAPL) is also positioned as a potential long-term AI winner due to its on-device strategy, which avoids the massive capital spending of its rivals. Be cautious with the memory stock sector, as the recent mania in names like Micron (MU) shows signs of a speculative late-cycle bubble that could correct sharply. The traditional SaaS sector also faces significant headwinds from AI disruption, with stocks like Salesforce (CRM) already down nearly 40% from their peak. While central to the AI buildout, be aware that NVIDIA (NVDA)'s stock has stalled recently and faces risks from circular financing practices.

AI Market vs. Dot-com Bubble

• The current AI boom is frequently compared to the dot-com bubble of the late 1990s. The podcast draws several parallels and distinctions.
• Similarities:
  • A major, potentially disruptive new technology wave is driving excitement (AI now vs. the Internet then).
  • The NASDAQ 100 performance since the launch of ChatGPT (Nov 2022) is similar to its performance after the Netscape IPO (Aug 1995). The speaker suggests there may still be "a ways to go" for the current rally.
  • Early winners are the "picks and shovels" companies building the foundation (NVIDIA today vs. Cisco and Sun back then). The application layer is expected to come later.
  • "Circular financing" is a concern. Today, companies like NVIDIA and Google are investing in customers who then buy their products. This is similar to vendor financing in the dot-com era by companies like Lucent and Cisco, which "didn't turn out very well."
  • Companies are adding "AI" to their names to boost stock prices, similar to how companies added ".com" in the bubble.
• Differences:
  • One analyst argues the AI bubble is a "late-cycle bubble" at the end of the 50-60 year information technology wave, whereas the dot-com bubble was a "mid-cycle bubble."
  • This implies AI is being built on existing infrastructure (the internet), leading to faster adoption but potentially different outcomes.
  • The capital expenditure (CapEx) burden is now on the hyperscalers (Microsoft, Google, Meta) and chip foundries, whereas in the dot-com era, tech companies got a "free ride" off telecom companies that overbuilt fiber networks.

Takeaways

• Investors should be aware of the historical parallels to the dot-com bubble, particularly regarding "picks and shovels" rallies and risky financing schemes.
• The current rally could still have room to run, but the risk of an "overbuild" in data centers and a subsequent downturn is a key theme to watch.
• The argument that this is a "late-cycle" bubble suggests that established incumbents (Microsoft, Google) may be the primary long-term winners, with fewer disruptive new companies emerging than in the dot-com era.

NVIDIA (NVDA)

• NVIDIA is identified as the primary "picks and shovels" vendor of the AI boom, providing the foundational hardware (GPUs).
• The company is involved in circular financing, investing in companies like CoreWeave (a "neocloud" company), which then turn around and buy NVIDIA's products. This is presented as a risk factor to watch.
• NVIDIA's investments are also seen as strategic, such as buying into Grok to bring a potential competitor "under the umbrella."
• The stock's performance has stalled recently, noted as "trading at the same spot it was trading six months ago." This could indicate that the initial, easiest gains for the stock are over.

Takeaways

• While NVIDIA is central to the AI build-out, investors should be cautious about the risks of circular financing, where NVIDIA funds its own customers. This can create artificial demand.
• The stock's recent sideways movement after a massive run-up suggests investors are becoming more selective and may be rotating capital into other areas of the AI ecosystem.
• Watch for NVIDIA's ability to maintain its market dominance as competitors (like Google's TPUs and custom silicon from Broadcom) and open-source models from China gain traction.

SaaS (Software-as-a-Service) Stocks

• The entire SaaS sector is facing significant pressure and uncertainty due to the rise of AI.
• Salesforce (CRM) is highlighted as a company whose stock is down nearly 40% from its all-time high. A recent $5.6 billion contract with the Army was seen as a "rounding error" for a company of its size, and the stock sold off after an initial pop.
• There is a fear that AI could eventually replace legacy software, allowing companies to develop their own solutions in-house instead of buying from vendors like Workday (WDAY) or Salesforce.
• However, large enterprises like JP Morgan still value having a "throat to choke," meaning they prefer relying on third-party vendors for support and accountability, which could slow the disruption.
• Pricing pressure is a major headwind. The "per-seat" pricing model is expected to shift to a "consumption-based" or "value-added" model, making revenue growth harder to achieve.
• Valuations are also a concern. Many SaaS stocks are still coming down from "outlandish" price-to-sales ratios. High stock-based compensation is another issue, as it masks the true profitability of these companies on a GAAP basis.

Takeaways

• The SaaS sector is a high-risk area. Investors should be wary of traditional software companies that do not have a clear and compelling AI strategy.
• Expect continued pressure on software stock valuations and a shift in business models away from simple per-seat pricing.
• Look for software companies that provide a "system of record" with deeply embedded data, as they may be more resilient to disruption than others.

Memory Stocks (Micron, Seagate, etc.)

• Memory stocks like Micron (MU), SanDisk, Western Digital (WDC), and Seagate (STX) have seen "outright mania," with stocks up hundreds of percent in the last six months.
• This rally is attributed to a supply-side crunch and tightness in the market for memory needed in AI systems.
• The podcast presents a theory that these massive price hikes are characteristic of a "late-cycle bubble," comparing it to the oil crisis of 1973.
• The implication is that while the price surge can continue for a while due to supply issues, it may be followed by a collapse once the supply/demand imbalance resolves.

Takeaways

• The rally in memory stocks is seen as a potentially speculative, late-stage phenomenon in the current AI cycle.
• While the theme is powerful right now due to supply constraints, investors entering at these high levels should be aware of the cyclical nature of the memory industry and the risk of a sharp correction if and when supply catches up to demand.

Non-Tech Incumbents (Industrials & Banks)

• A key future theme is that the biggest winners from AI may come from "unexpected places" outside of the tech sector.
• The focus is shifting to how traditional, labor-intensive industries can use AI to drive productivity and efficiency gains.
• Sectors mentioned include retailers, banks, and industrials.
• Specific examples of companies reportedly ahead in their use of AI include Morgan Stanley (MS) and JP Morgan (JPM). The speaker suggests their stock performance may already be reflecting this differentiation.
• Industrial companies like Deere (DE) and Caterpillar (CAT) are also cited as potential beneficiaries.
• The argument is that even small efficiency gains in low-margin businesses can lead to significant earnings growth.

Takeaways

• Investors should start looking beyond pure-play AI and tech stocks to identify established companies in traditional sectors that are effectively implementing AI.
• While it's still early, companies in banking, insurance, and industrials that can demonstrate quantifiable benefits from AI could be the next wave of winners.
• This is a long-term theme for 2026 and beyond. Be cautious of valuations running ahead of actual, demonstrated productivity gains.

Other Mentioned Companies & Themes

• OpenAI: The company faces "big questions" and significant risk. It needs to raise massive amounts of capital (potentially $100 billion), and an inability to do so could prevent it from going public or lead to a "public down round." It is described as a potential "accident waiting to happen."
• Oracle (ORCL): Used as a "poster child" for the risks of the AI build-out. The stock surged on a contract from OpenAI but then gave back all its gains and more. This was attributed to concerns about Oracle having to raise debt to fund the build-out, unlike cash-rich hyperscalers.
• Corning (GLW): The stock hit a new all-time high, reaching levels from the year 2000, after announcing a $6 billion fiber optic cable contract from Meta. The speaker calls this a potential "ringing the bell" moment, suggesting it might be a sign of market froth, similar to the fiber overbuild in the dot-com bubble.
• Apple (AAPL): Positioned as a potential long-term AI winner. By focusing on "inference on the edge" (on the device itself), Apple has avoided the massive CapEx spending of the hyperscalers. If this strategy proves successful, Apple will be "doubly blessed."
• Palantir (PLTR): Highlighted for its "truly astounding" valuation, trading at 64 times forward sales. The stock is down 20% from recent highs, suggesting investors are becoming more critical of such extreme valuations.
• C3.ai (AI): Mentioned as a cautionary tale. Despite having "AI" in its name long before the current boom, the stock is down 93% from its all-time high, showing that a name alone does not guarantee success.