OpenAI closed a $122 billion funding round this week at an $852 billion valuation.
Let that number land. At $2 billion per month in revenue — 40% of it enterprise — OpenAI just crossed the threshold where it becomes infrastructure. The kind of thing you build on, not the kind of thing you evaluate.
If you're still treating it like a vendor you might swap out next year, you're operating with the wrong mental model.
What Changed This Week
The funding round itself isn't the signal. The round is proof of a signal that's been building for 18 months.
Here's the actual signal: enterprise is buying AI at a rate that validates a $2 trillion industry within five years. When SoftBank, Amazon, Nvidia, and Microsoft all participate in the same round at the same valuation, they're not betting on a product. They're betting on a platform — the same way investors bet on AWS not as cloud storage, but as the backbone of the internet economy.
The stated goal is a unified agent-first platform combining ChatGPT, Codex, and agentic capabilities into a single operating surface for enterprise AI. That's not a tool. That's infrastructure.
The Strategy Problem
Most enterprise leaders are still asking the wrong question: "Which AI vendor should we use?"
The right question is: "What does our AI architecture look like in a world where the leading platform is worth $852 billion, isn't going anywhere, and is actively building agentic capabilities that will run end-to-end business processes without human intervention?"
Those questions have different answers. The first leads to vendor comparisons and pilot programs. The second leads to architectural decisions and governance frameworks.
The companies asking the first question are building on sand. Every platform decision is provisional. Every implementation is an experiment. Every pilot is an excuse to not commit.
The companies asking the second question are building something that compounds.
The Codex Problem Nobody's Talking About
Codex — OpenAI's coding agent — hit 2 million weekly users this week. Up 5x in three months.
This matters for CTOs specifically. Codex isn't replacing senior engineers. It's replacing the workflow of junior developers — the "write this function," "fix this bug," "explain this error" work that used to require a human to touch. At scale, that changes the composition of engineering teams.
The companies that understand this are restructuring what they hire for. They're not looking for developers who can write code. They're looking for developers who can architect systems, define quality standards, and direct AI agents that write the code. That's a senior-profile hire, not a junior-profile hire.
The companies that don't understand this are going to keep posting junior developer roles into a market that's been fundamentally reshaped — and wonder why the candidates they're seeing don't match the output they need.
What Enterprise AI Talent Looks Like Now
If OpenAI is infrastructure, then the people who implement it are infrastructure engineers — not AI experimenters.
The skill profile that matters now is someone who can:
- Build on AI platforms with the same durability assumptions you'd bring to a database architecture or cloud infrastructure decision
- Think about governance, auditability, and failure modes from day one — not as a compliance afterthought
- Navigate a vendor landscape where the platform is actively expanding its footprint quarterly
This is different from the AI talent that was valuable 18 months ago. The early days rewarded curiosity and speed. What matters now is depth, durability, and architectural judgment.
Those people are rare. They're not looking for their first AI job. They're coming from places that have already built at scale — and they're evaluating offers based on whether the company they're joining is building something real or still running experiments.
The Question to Ask
One question cuts through all of this: is your organization's AI strategy built on a thesis that can survive your primary vendor growing 5x?
If the answer is "we're not sure" or "we haven't thought about it at that scale" — that's the work.
The $852 billion valuation isn't a number to have an opinion about. It's a forcing function. The companies that build with that reality in mind are going to look very different from the companies that didn't in 24 months.
VC5 Consulting works with companies that need to staff for AI at infrastructure scale. If you're figuring out what your AI architecture actually needs to function at that level, let's talk.