‘Big Short’ Investor Explains How the AI Bubble Will Burst 💥

Welcome to another edition of Horizon AI,

In today’s issue, we take a look at an interview with legendary ‘Big Short’ investor Steve Eisman, in which he discusses his concerns about the current AI market bubble and the systemic risks within the industry.

Let’s jump right in!

Read Time: 4.5’ min

Here's what's new today in the Horizon AI

• Chart of the week: These U.S. States Have the Most Jobs Exposed to AI

• Google Launches a 'Universal' Gemini Agent for Work

• AI Findings/Resources

• AI tools to check out

• Video of the week

TOGETHER WITH DELL

Built for the unstoppable.

The Dell Pro powered by Intel® Core™ Ultra with Intel vPro® adjusts power usage based on how you work, so your battery life never slows you down.

Explore More

Chart of the week

These U.S. States Have the Most Jobs Exposed to AI

• Washington has the highest share of workers in AI-exposed occupations at 5.7%, followed by Virginia (4.6%) and Washington, D.C. (4.5%).

• California has the largest number of workers in highly AI-exposed roles, with 724,000 positions, followed by Texas at 500,000.

• Jobs involving programming, mathematics, writing, and data processing face significant potential disruption, though AI exposure does not necessarily mean those jobs will disappear.

Source

AI News

GOOGLE

Google Launches a “Universal” Gemini Agent for Work

At Gemini at Work 2026, Google Cloud announced its new Gemini agent, describing it as a “universal agent for work” that answers your questions, handles your knowledge work, creates your images and media, and writes and runs code.

Details:

• The agent connects with Google Workspace apps, including Gmail, Drive, Docs, Sheets, and Calendar, while also supporting access through mobile devices, desktop computers, the web, Slack, and Microsoft 365.

• Because it operates in the cloud, Gemini can maintain context across connected devices, allowing users to start tasks in one place and continue them elsewhere.

• Google says the agent can coordinate with specialized sub-agents for different tasks and operate as a dedicated AI coworker, complete with its own identity and its own @agents.company.com email.

• The system can also select the AI model best suited to each task, with the goal of handling different workflows without requiring users to manage the underlying technology.

Gemini agent is currently only available to enterprise customers in private preview, but it could hint at where Google wants its AI assistant to head, one that answers questions and gets work done from a single window.

Source

AI Findings/Resources

⭐ Married with Children, the classic 1980s sitcom, gets an AI crossover with movie legends

🤨 Can AI be moral without being sentient?

👉 Ben Affleck is an AI nerd, and the internet is impressed

👀 Every app is being CHAT-ified

🤔 Boris Cherny, the creator and head of Claude Code at Anthropic, shares how he prompts Claude, saying you should talk to it the way you would to a coworker

AI Tools to check out

📹 Ankon AI: Turn any idea into a whiteboard video.

📝 Vunote: Turn YouTube videos into organized knowledge with timestamped notes, AI-powered insights, summaries, and more.

✨ Figr: Design with an AI that already understands your product.

🦾 Tadata: An AI employee in Slack that helps you get more done.

Video of the week

‘Big Short’ Investor Eisman Lays Out the Biggest Risks Facing the AI Market

In this interview, ‘Big Short’ investor Steve Eisman, best known for betting against the housing market before the global financial crisis, shares his critical view of the current AI market and the systemic risks he sees in the industry.

Concentration Risk

Eisman highlights that the AI supply chain, from Nvidia to major hyperscalers like Google, Amazon, and Microsoft, relies heavily on just two companies: OpenAI and Anthropic. They are burning through massive amounts of cash, and he warns that if they fail, it could trigger a significant recession.

Circular Financing

Much of the capital powering these AI giants comes from the very companies they pay for cloud services, creating a circular funding dynamic that may be unsustainable .

Lack of Moats

Eisman argues that the Large Language Model (LLM) industry currently lacks true "moats" or competitive advantages because developers switch between models easily, and companies are becoming more cost-conscious, ending the era of "token maxing".

Manufactured Hysteria

He suggests that the ongoing narrative about AI ending the world is a strategic distraction intended to prompt government regulation, which could effectively create a "legal moat" or duopoly for established players.

Power Constraints

The growth of data centers is creating a massive demand for electricity, which the US grid is struggling to meet. Eisman notes that companies involved in physical power infrastructure, such as GE Vernova, are seeing significant growth as a result.

Investing Perspective

Eisman advises against investing in LLM-specific companies due to the lack of competitive moats. Instead, he suggests focusing on "picks and shovels", infrastructure companies that benefit from the capital expenditure of the hyperscalers, such as Nvidia, Micron, Arista Networks, Cisco, and Eaton.

That’s a wrap!

Thanks for sticking with us to the end! Let’s stay connected on LinkedIn and Twitter.

Not subscribed yet? Sign up here and send it to a colleague or friend!

See you in our next edition!

Gina 👩🏻‍💻