Inside the AI Power Struggle: Breakthroughs, Breaches, and Billion-Dollar Battles

Welcome back to your AI and tech roundup! 

In terms of breakthroughs, the big news this week is the release of Gemini 3. Though it did great on the benchmarks, I usually don't pay much attention to that. What is a bigger test is to see how we

This week's big news in AI is Gemini 3, Google's latest generative AI model. A number of observers, including OpenAI itself, consider this a development worth taking seriously. It's a good illustration of how the AI game is wide open right now.

Both OpenAI and Anthropic have responded—there's been reported panic at OpenAI, and Anthropic has released Opus 4.5. 

The other major story is that Google is in talks with Meta to sell its AI chips. This is significant because it creates tremors in Nvidia's dominance. For a while, Nvidia thought they were king of the mountain—the only company that could deliver the chips necessary for this generative AI revolution. That assumption is now being challenged.

This connects to a question I recently discussed with students: what might cause this AI bubble to burst? This chip competition could be one factor. Relatedly, Michael Burry announced he's launching a Substack to monitor the AI bubble. That's one of the reasons he shut down Scion Asset Management—to speak freely without SEC restrictions.

When thinking about disruptive innovation, it's worth revisiting the Netflix-Blockbuster case study. One lesson I always emphasize: when the dot-com bubble burst, Blockbuster dismissed Netflix partly because they believed internet hype was overblown. This is where the Gartner Hype Cycle becomes essential—technologies go up, they burst, and then they become normalized. It's not a smooth S-curve; there's a detour through hype.

---

1. OpenAI Confirms Data Breach Through Third-Party Vendor Mixpanel

OpenAI confirmed that a security incident at third-party analytics provider Mixpanel exposed identifiable information for some users of its API services. The company emphasized that personal ChatGPT users were not affected and that no chats, API usage data, passwords, API keys, payment details, or government IDs were compromised. Leaked data may include API account names, email addresses, approximate locations, and technical details like browser and operating system. OpenAI is notifying affected users directly, warning them to watch for phishing attempts, and has removed Mixpanel from all products while expanding security reviews across its vendor ecosystem. (Source: The Star)

Key Takeaways

• Limited to API Users: The breach impacted OpenAI API customers only, not people using ChatGPT for personal use.
• Sensitive Data Protected: No chats, passwords, API keys, payment information, or government IDs were exposed in the incident.
• Stronger Vendor Security: OpenAI has removed Mixpanel and is conducting broader security and vendor reviews to reduce future risks.

---

2. Michael Burry Launches Substack and Warns AI Boom Mirrors Dot-Com Bubble

Michael Burry, the famed “Big Short” investor known for calling the 2008 housing crash, has launched a paid Substack newsletter titled Cassandra Unchained shortly after closing his hedge fund, Scion Asset Management. Burry insists he is not retired and says the blog now has his “full attention.” In early posts, he compares today’s AI boom to the 1990s dot-com era, warning that nearly $3 trillion in projected AI infrastructure spending over the next three years shows classic bubble behavior. He also criticizes tech heavyweights such as Nvidia and Palantir, questioning their accounting practices and the sustainability of current valuations. Shutting down his fund, Burry says, frees him from regulatory and compliance constraints that previously limited how candid he could be in public communications. (Source: Reuters)

Key Takeaways

• Burry Goes Independent: His new Substack, priced at $39 per month, has already attracted more than 21,000 subscribers.
• AI Bubble Concerns: Burry argues that current AI infrastructure spending and investor enthusiasm resemble the excesses of the dot-com era.
• Big Tech Under Scrutiny: He has sharpened criticism of companies like Nvidia and Palantir, questioning their growth assumptions and accounting choices.

---

3. Nvidia Shares Drop as Google Considers Selling AI Chips to Meta

Nvidia’s stock fell after a report indicated that Google is in talks with Meta to sell its custom tensor processing unit (TPU) AI chips for use in Meta’s data centers starting in 2027. This would mark a shift from Google’s current approach of renting access to TPUs through Google Cloud toward directly selling chips to major customers. The report also said Google is pitching TPUs to other clients and could potentially capture as much as 10% of Nvidia’s annual revenue. The news added to investor worries that Nvidia’s biggest customers—such as Google, Amazon, and Microsoft, all of which are developing their own AI chips—are becoming formidable competitors. Amid broader concerns about an AI bubble and “circular” AI investment structures, Nvidia responded by praising Google’s AI progress and reaffirming that its own business remains fundamentally sound and transparent. (Source: Yahoo Finance)

Key Takeaways

• Google May Sell TPUs Externally: Talks with Meta suggest Google could evolve from cloud-only chip access to directly selling AI hardware.
• Competition for Nvidia Intensifies: Google, Amazon, and Microsoft’s in-house AI chips pose growing threats to Nvidia’s dominance.
• AI Bubble Fears Linger: Stock moves and criticism from investors like Michael Burry feed concerns about froth in the AI sector.

---

4. Anthropic Unveils Claude Opus 4.5 Amid Intensifying AI Model Race

Anthropic introduced Claude Opus 4.5, calling it its most powerful AI model so far and positioning it as the top performer for coding, AI agents, and computer-use tasks. The company says Opus 4.5 outperforms Google’s Gemini 3 Pro and OpenAI’s GPT-5.1 and GPT-5.1-Codex-Max on software engineering benchmarks. Anthropic also highlighted the model’s creative problem-solving abilities, noting that in one airline customer-service benchmark, Opus 4.5 technically “failed” by solving the user’s problem in an unanticipated way that still helped the customer. The launch comes as Gemini 3 reshapes the competitive landscape, Meta’s Llama 4 Behemoth continues to face delays, and the cost of building frontier AI models soars. Backed by large chip deals with Amazon and Google, Anthropic is reportedly on track to break even by 2028, earlier than OpenAI’s projected timeline. (Source: Yahoo Finance)

Key Takeaways

• New Flagship Model: Claude Opus 4.5 is positioned as best-in-class for coding, agents, and advanced computer-use scenarios.
• Creative Problem Solving: The model can find unconventional solutions, occasionally breaking benchmarks while still successfully helping users.
• High-Cost, High-Stakes Race: Massive chip deals and huge infrastructure spending underscore how expensive leading the AI model race has become.

---

5. Gemini 3 Shows Google’s Biggest Advantage Over OpenAI

With the launch of Gemini 3, Google is showcasing its “full-stack” advantage over OpenAI. Google controls the entire AI pipeline: DeepMind researchers build the models, in-house TPUs train them, Google Cloud hosts them, and products like Search, YouTube, and the Gemini app deliver them to users. For the first time, Google rolled out a new flagship AI model directly into Google Search on day one via an “AI mode,” eliminating friction for users who might otherwise need to download an app or visit a separate site. This end-to-end control lets Google move quickly and avoid the dependency and circular financing issues some rivals face. However, OpenAI still holds a powerful branding edge, as “ChatGPT” has effectively become shorthand for AI in the public’s mind. Analysts say Gemini 3 may be the clearest sign yet that Google is finally aligning its vast technical and distribution resources into a cohesive AI strategy. (Source: Business Insider)

Key Takeaways

• Full-Stack Advantage: Google owns everything from chips to cloud to consumer apps, allowing tighter integration and faster deployment of Gemini 3.
• AI Mode in Search: Integrating Gemini 3 directly into Google Search puts advanced AI tools in front of users instantly, with minimal friction.
• Branding Battle Ahead: While Google has the infrastructure edge, OpenAI’s ChatGPT still dominates public awareness, setting up a long-term branding showdown.

Author: Malik D. CPA, CA, CISA. The opinions expressed here do not necessarily represent UWCISA, UW,  or anyone else. This post was written with the assistance of an AI language model. 

5 Reasons the AI Boom Is a 'Multi-Bubble' Waiting to Pop (or Why You Must Check Out this Bloomberg Podcast!)

On the following "Odd Lots" podcast, financial analyst and MIT fellow Paul Kedrosky argues that the AI boom is something historically unique and uniquely dangerous: a "meta-bubble" that combines the riskiest elements of every major financial crisis into a single, unprecedented event.

   

Beyond the story of the multi-bubble (which is probably a better term then Meta-Bubble to avoid confusion with the company):

One of the podcast co-host, Tracy Alloway, also brough up the issue of how private credit used to be called shadow banking:

 "I realized private credit kind of supplanted shadow banking as the term right like after 2008 we called it shadow banking and then at some point it flipped to I guess the cuddlier term  private credit"

Kedrosky points out that the entire shadow banking industry is $1.7 trillion dollars.

The episode also sheds light on the depreciation of the AI chips. Why does this matter? For those following Dr. Michael Burry, of Big Short fame, has delisted his Scion Asset Management after two important announcements.   Firstly, he said he is shorting Palantir and Nvidia. Secondly, he raised the alarm the changes in depreciation policies and the tech firms (see his tweet here), which he sees has overstated earnings.  However, look to point number 3 in this post to get Kedrosky’s take.

The other piece of context is to understand how much leverage is now linked to the AI Boom/Bubble:

 “The amount of debt tied to artificial intelligence has ballooned to US$1.2 trillion, making it the largest segment in the investment-grade market, according to JPMorgan Chase & Co…AI companies now make up 14 per cent of the high-grade market from 11.5 per cent in 2020, surpassing United States banks, the largest sector on the JPMorgan U.S. Liquid index (JULI) at 11.7 per cent, JPMorgan analysts including Nathaniel Rosenbaum and Erica Spear wrote in a note Monday.” (link)

 Finally, I learned about IBM’s GenAI offering named Granite. It is a small language model (SLM), which Kedrosky notes is emblematic of the use-case for GenAI:

“…what's increasingly happening is the problems they're solving are really mundane. And so it's things like I'm trying to onboard a bunch of new suppliers right now the people have weird zip codes and they sometimes don't match up. I have a dude in the back who fixes that I’d rather have someone who could do it faster so I could onboard a lot more suppliers. It turns out these small language models are really good at that these micro models like IBM's Granite and whatever else but those things require a fraction of the training are very cheap..”

 See here to learn more about IBM’s Granite GenAI SLM:

https://www.ibm.com/granite

Podcast Key Takeaways

1. It's Not Just a Tech Bubble; It's a "Multi-Bubble"

Paul Kedrosky's central thesis is that the current AI boom is not just another technology bubble; it's a "meta-bubble" (see comments above about why I think it should be the multi-bubble) He argues that for the first time in history, all the key ingredients of every major historical bubble have been combined into a single event, creating a situation of unparalleled risk.

Kedrosky identifies four core components that are simultaneously at play:

• A Real Estate Component: Data centers, the physical heart of the AI buildout, are a unique asset class sitting at the intersection of industrial spending and speculative real estate. This brings the property speculation element of past crises directly into the tech boom.

• A Powerful Technology Story: The narrative around AI is one of the most compelling technology stories ever told, comparable in scope to foundational shifts like rural electrification. This powerful story fuels investment and speculation on a massive scale.

• Loose Credit: The financing of the boom is being supercharged by loose credit, with a crucial distinction from past cycles: private credit has now largely supplanted traditional commercial banks as the primary lenders in this specific buildout.

• A Government Backstop: An "existential competition" narrative, framing the AI race as a critical national security issue between the US and China, has created a sense of a limitless, government-endorsed spending imperative. Nations around the world are pursuing "sovereign AI," suggesting capital is no object.

2. The Financing Looks Frighteningly Similar. It was used by Enron.

The financial engineering behind the AI boom rhymes with the complex and opaque structures central to the 2008 financial crisis. Even cash-rich tech giants are increasingly using Special Purpose Vehicles (SPVs), a move designed to keep massive amounts of debt off their balance sheets. The motivation, according to Kedrosky, is to avoid upsetting shareholders about diluting earnings per share to fund these colossal projects. The Byzantine complexity of these SPV structures, he notes, looks like the "forest with all the spiderwebs".

This structure incentivizes a dangerous blending process. To make the data center asset more attractive as a financial instrument, sponsors combine stable, low-yield tenants like hyperscalers with "flightier tenants" who pay much higher rates. This blending improves the overall yield, making it easier to securitize and sell to investors.

See here for details around Meta’s and x.ai’s use of SPV, see this article. And for a refresher on how Enron used SPVs to hide its debt from investors, check out this article.

3. The Assets Have a Short Expiration Date

A critical flaw in the AI financial structure is a dangerous "temporal mismatch" between long-term debt and short-lived assets. This risk is being actively obscured by accounting maneuvers. Kedrosky points out that around four years ago, tech companies extended the depreciation schedules for data center assets. This was done, however, just as the AI buildout began relying on GPUs with dramatically shorter lifespans.

 There are two reasons for this shortened lifespan. The first is rapid technological obsolescence. The second, and perhaps more important, is "thermal degradation." Kedrosky uses a "used car" analogy: a chip for simple storage is like a car "driven to church on Sundays." A GPU training AI models is run "flat out 24 hours a day," like a vehicle in a 24-hour endurance race. This intense usage can slash its useful lifespan to as little as 18-24 months.

Yet these short-lived GPUs are the core collateral for loans stretching out 30 years. This creates an "unprecedented temporal mismatch" and a constant, significant refinancing risk that will come to a head in the coming years when a massive wave of these debts comes due.

4. The Business Models Run on "Negative Unit Economics"

Before diving into the flawed economics, Kedrosky offers a crucial disclaimer: "AI is an incredibly important technology. What we're talking about is how it's funded." The problem is that the core products are fundamentally unprofitable. Unlike traditional software, where fixed costs are spread across more users, the costs for large language models (LLMs) rise more or less linearly with use. This leads to what is termed "negative unit economics."

"...a fancy way of saying that we lose money on every sale and try to make it up on volume..."

When confronted with this reality, the justification for the massive capital expenditure shifts to what Kedrosky calls "faith-based argumentation about AGI." He cites a recent investment bank call where analysts justified the spend using a top-down model. First, they calculated the "global TAM for human labor," then simply assumed AI would capture 10% of it. Kedrosky points out that such a number is hard to pin down in terms of exact figures.  

 5. We're Betting Trillions on Potentially Inefficient Technology

A counter-intuitive risk is that the entire technological path the US is on may be a bloated, inefficient dead end. The current American strategy focuses on building ever-larger, computationally intensive models. This stands in stark contrast to China's "distillation" or "train the trainer" approach, where they use large models to train smaller, highly efficient ones. (See in the intro the use of IBM's Granite as an example of this observation)

This suggests huge efficiency gains are possible. Kedrosky notes that the transformer models underlying today's LLMs went from the lab to market faster than almost any technology in history, and as a result, they are "wildly inefficient and full of crap."

The implication is profound. If massive efficiency gains are achievable, as China's approach suggests, it means that the current forecasts for future data center demand are likely "completely misforecasting the likely future the arc of demand for compute." The entire financial model is based on a technological path that may already be obsolete.

Closing thoughts

Many contend that we are in AI Bubble. And it’s hard to argue against that. The patterns of technology investments, whether it was the dotcom bubble of the 1990s, the radio bubble of the 1920s, or the railway bubble of the 1840s, there is a consistent pattern of investors engaging in a euphoric rush to capture a “powerful technology story”. The key challenge will be the downstream effects of containing the bursting of the bubble. We have seen how the clean-up for the 2008 financial crisis was “in progress” and then COVID hit. Inflation is still running high – an after effect of that last crisis. How much room is left for further maneuvering? Unfortunately, this is something that we will have to wait and see how things turn out.

Author: Malik D. CPA, CA, CISA. The opinions expressed here do not necessarily represent UWCISA, UW,  or anyone else. This post was written with the assistance of an AI language model. 

AI Iceberg: Tech Bubble Warnings, White-Collar Cuts, Deepfake Dilemma, and Canada's AI Strategy

‘Big Short’ Investor Bets Against AI Giants in Market Warning

Michael Burry, famed for predicting the 2008 financial crisis and immortalized in The Big Short, has disclosed new bearish positions through his hedge fund, Scion Asset Management. Burry has taken put options—investments that profit from a stock's decline—against two tech giants: Palantir and Nvidia. Despite Palantir’s strong earnings report and raised revenue outlook, its stock saw volatility due to valuation concerns. Nvidia also faced market jitters amid geopolitical tensions and pending earnings, particularly after former President Trump’s comments about limiting chip sales to China. Burry's move aligns with his recent warnings about an overheated market, echoing sentiments from other Wall Street leaders about inflated tech valuations. Known for his contrarian positions, Burry’s recent bets signal caution amid a tech-driven market rally fueled by AI hype (Source: Yahoo Finance).

• Contrarian Warning: Michael Burry is betting against Nvidia and Palantir, signaling concerns about a tech bubble.
• Market Volatility: Despite strong financials, Palantir's stock dropped due to valuation skepticism; Nvidia's dip was influenced by geopolitical factors.
• Broader Bearish Sentiment: Burry’s move aligns with a broader warning from major Wall Street voices about an impending market correction.

The Number One Sign You’re Watching an AI Video

As AI-generated videos flood social media, experts are warning that blurry, low-resolution footage is often the best clue you’re watching a fake. According to researchers like Hany Farid and Matthew Stamm, poor-quality videos are frequently used to mask telltale AI inconsistencies—such as unnatural skin textures or glitchy background movements—making them harder to detect. Many recent viral AI videos, from bouncing bunnies to dramatic subway romances, share a common trait: they look like they were filmed on outdated devices. While advanced models like OpenAI's Sora are improving, shorter clip lengths, pixelation, and intentional compression remain key signs. Experts argue we must shift from trusting visual “evidence” to verifying context and source—similar to how we assess text—because soon, visual cues may vanish entirely. The rise of these deceptively convincing clips signals a new era in digital literacy where provenance, not appearance, becomes the cornerstone of truth (Source: BBC).

• Low Quality, High Risk: Blurry, pixelated videos are a major red flag for AI fakes—they often hide subtle AI flaws.
• Short and Deceptive: AI-generated videos are usually brief due to high processing costs and a higher chance of mistakes in longer clips.
• Context Over Clarity: Experts urge people to stop trusting visuals alone—source and verification matter more than ever.

The $4 Trillion Warning: AI May Be Headed for a Historic Crash

Brian Merchant of Wired applies a scholarly framework to assess whether the AI industry is in a financial bubble—and concludes it likely is. Drawing on research by economists Brent Goldfarb and David A. Kirsch, who studied dozens of historical tech bubbles, Merchant finds AI checks every box for a classic speculative frenzy: high uncertainty, the dominance of “pure-play” companies like OpenAI and Nvidia, a surge of novice investors, and irresistible industry narratives promising everything from job automation to miracle cures. Unlike earlier technologies, AI’s ambiguity fuels investor enthusiasm instead of caution, while public and private markets pour unprecedented capital into ventures with unclear profit models. Nvidia, for example, now accounts for 8% of the total stock market value. Goldfarb ultimately rates AI at a full 8 out of 8 on the bubble-risk scale, likening today’s mania to the radio and aviation bubbles that preceded the 1929 crash. If AI fails to deliver on its sweeping promises, the fallout could be massive (Source: Wired).

• All Bubble Indicators Flashing: AI ranks highest on a tested framework for identifying tech bubbles—uncertainty, pure plays, novice investors, and grand narratives.
• Public at Risk: With firms like Nvidia heavily tied to public markets, a burst could affect everyday investors and retirement funds.
• Narrative-Driven Speculation: AI’s limitless promise has generated massive investment despite weak current returns, echoing past tech hype cycles.

White‑Collar Jobs Vanish as AI Reshapes the Office Landscape

Major U.S. companies—such as Amazon.com, Inc., United Parcel Service (UPS), and Target Corporation—are cutting tens of thousands of white‑collar roles as they adopt artificial intelligence and automation to streamline operations. Amazon announced plans to cut 14,000 corporate jobs (up to ~10 % of its white‑collar staff). UPS reduced its management workforce by about 14,000 positions over 22 months. These actions reflect a broader shift: traditionally secure white‑collar roles—even for experienced professionals and recent graduates—are becoming vulnerable. The wave of cuts is attributed in part to AI tools replacing or reducing the need for many tasks formerly done by higher‑paid office workers; at the same time, hiring remains stronger in blue‑collar or trade sectors. The changing landscape means intensified competition for fewer roles, and many workers are facing uncertainty about their careers (Source: The Wall Street Journal).

• White‑Collar Vulnerability: Even well‑educated office professionals are now at risk as AI enables firms to cut back on corporate staffing.
• Structural Shift in Jobs: While white‑collar hiring weakens, demand for trade and frontline roles is relatively stronger—signaling a change in which segments of the workforce are most secure.
• Increased Competition & Pressure: With fewer open roles and employers demanding more specific qualifications, both new grads and mid‑career workers face a tougher employment market.

Canada’s AI Crossroads: Sovereignty or Speed?

As AI infrastructure booms globally, Canada faces a critical decision: whether to deepen reliance on foreign tech giants like OpenAI or invest in sovereign, Canadian-controlled systems. While companies like OpenAI have proposed building AI data centers in Canada—attracted by the country’s clean energy supply—critics warn that such partnerships could threaten national digital sovereignty. Canadian data, from health records to mobility stats, is increasingly fueling foreign AI innovation and economic gains. Yet, the infrastructure to process and govern that data under Canadian law remains underdeveloped. The federal government has begun investing in domestic AI capabilities, but unless cloud and compute services are Canadian-owned and governed, experts argue that Canada will merely become a digital raw material supplier. Drawing parallels to the country’s historical resource exports, the article urges Canada to prioritize legal and economic control over its data to foster innovation and retain value at home (Source: Maclean’s).

• Sovereignty vs. Speed: Relying on U.S. tech firms for AI infrastructure risks ceding control over Canadian data and its economic value.
• Data as Digital Raw Material: Like lumber or oil, Canada’s data is being exported and monetized elsewhere while domestic innovation lags behind.
• A National Strategy Needed: Experts urge Canada to treat data governance and infrastructure as core to its economic and sovereign future.

Author: Malik D. CPA, CA, CISA. The opinions expressed here do not necessarily represent UWCISA, UW,  or anyone else. This post was written with the assistance of an AI language model