In partnership with

Why Mira Murati and Ilya Sutskever's AI Startups Are Worth the Hype—and the Billions

In 2025, the AI industry is witnessing unprecedented investments, with former OpenAI executives Mira Murati and Ilya Sutskever at the forefront. Murati's Thinking Machines Lab is reportedly raising $2 billion at a $10 billion valuation, while Sutskever's Safe Superintelligence Inc. (SSI) has already secured $3 billion at a $32 billion valuation. These figures might seem staggering, especially for companies in their infancy pre-product, but a closer look at their backgrounds—and the economic mechanics of the venture ecosystem—reveals why this all kinda makes sense.

So lets set the stage first…

Mira Murati: Engineering Visionary Behind AI's Mainstream Breakthroughs

Career Highlights

• Tesla (2013–2016): Worked on the Model X as a product manager, gaining firsthand experience in scaling deep tech into mass-market hardware.

• Leap Motion (2016–2018): VP of Product, focused on human-computer interaction through motion control interfaces.

• OpenAI (2018–2024): VP of Applied AI & Partnerships, later CTO. Oversaw development and productization of ChatGPT, DALL·E, and GPT-4.

• 2025: Founded Thinking Machines Lab, assembling top-tier talent including Bob McGrew and Alec Radford. The company is raising $2B at a $10B valuation to build the next wave of intelligent infrastructure.

Ilya Sutskever: Deep Learning Pioneer and Architect of Modern AI

Career Highlights

• University of Toronto (2005–2012): Protégé of Geoffrey Hinton; co-creator of AlexNet, the seminal deep learning breakthrough in computer vision.

• Google Brain (2013–2015): Worked on early NLP systems, co-authoring the foundational seq2seq paper.

• OpenAI (2015–2023): Co-founder and Chief Scientist. Instrumental in developing GPT-2, GPT-3, and GPT-4, while shaping OpenAI’s research philosophy.

• 2025: Launched Safe Superintelligence Inc., focused exclusively on developing superintelligent AI with robust safety alignment. Raised $3B at a $32B valuation, with reported backing from Alphabet and NVIDIA.

OpenAI’s Rise as a Precedent

• 2015: Founded as a nonprofit with a few million in funding.

• 2018–2020: Transitioned to a capped-profit model to raise capital.

• 2022: ChatGPT launches and becomes the fastest-growing product in history (reaching 100M users in two months).

• 2023: Hits $100M in annualized revenue in under 2 years.

• 2025: Reports rumored $12B+ annualized revenue.

Murati and Sutskever were key architects of that trajectory. Now, they’re starting from a place of credibility, demand, and capital readiness.

Why VCs Must Invest in High-Valuation AI Startups

In the current venture landscape, funds aren’t just able to invest—they're under pressure to deploy massive amounts of capital. Firms are managing $2B, $5B, even $10B+ funds, and their math only works if a few companies generate $100B dollar exits at these levels

VC Fund Economics: Returning a $10 Billion Fund

Let’s say a venture capital firm with $10B under management invests $1B into one of these early AI startups. Here’s what the math looks like:

Scenario A: 5% Ownership at Exit

• 5% of a company that exits at $200 billion = $10 billion return (break-even for the fund).

Scenario B: 2.5% Ownership at Exit

• 2.5% of a $400 billion exit = $10 billion return (break-even again).

And remember: after multiple funding rounds and dilution, 2.5–5% ownership is realistic for many large firms investing at a $10–$35 billion entry valuation—especially if they don’t lead or follow on heavily.

Why These Bets Still Make Sense

• Big Funds = Big Bets: A multibillion-dollar fund writing $20M checks isn’t going to return capital. They have to invest $200M–$1B into companies with massive potential.

• Foundational AI is the rare exception: Few sectors outside of foundational AI make $100B+ exits seem plausible—but here, they’re within reach.

• Math drives behavior: If a firm wants to raise again, it must prove it can generate those giant wins. That means betting early on proven AI leaders with the team, capital, and ambition to build the next OpenAI.

Learn AI in 5 minutes a day

This is the easiest way for a busy person wanting to learn AI in as little time as possible:

1. Sign up for The Rundown AI newsletter

2. They send you 5-minute email updates on the latest AI news and how to use it

3. You learn how to become 2x more productive by leveraging AI

Sign up to start learning.