The AI opportunity CFOs must know
OpenAI sets dates for AI researchers and GPT-6. Humanoids hit the market. Compute turns into capex factories. Your job? Allocate judgment.
The AGI countdown has begun
OpenAI just drew a line in time.
Yes, they gave AGI a date.
By September 2026 (< 2 years): An intern-level AI research assistant. This is the automation of entry-level knowledge work, analysis, data gathering, and preliminary research.
By March 2028 (< 3.5 years): A legitimate AI researcher. This is the automation of core R&D, strategy, and problem-solving functions.
And in 6 months we might get GPT-6.
That changes everything.
OpenAI’s restructure is complete. A nonprofit foundation now governs a public-benefit corporation, with Microsoft locked in as its compute backbone.
They already have $1.4 trillion in motion toward a $7 trillion vision.
Imagine the moment models can run autonomously for days … intelligence stops being scarce. The only limit is how much compute you can buy.
This reframes massive AI spending from a questionable cost center into an existential necessity. Your AI budget is no longer about ROI. It’s about survival.
If you control compute, you control acceleration.
The first mass-market Humanoid Robot is here
1X Neo is the first humanoid robot for consumers and offices Shipping early 2026.
Price: $20,000 or $499/month.
Strength: 150 lbs.
Weight: 66 lbs.
Labor is no longer hired. It’s subscribed to.
When you can lease muscle that never sleeps, every operation model flips. Imagine CFOs who start testing humanoids in logistics, retail, or facility work now will own the benchmarks everyone else copies later.
You can hand NEO a daily list.
Laundry, shelves and tidying You can schedule tasks, or just say the word and watch it move. NEO can fold, organize, and reset. And when it hits a task it doesn’t know, a 1X Expert takes over remotely, guides it through the motion, and trains it to do it next time, all while the job still gets done.
This is exciting stuff.
Today it folds laundry.
Tomorrow it builds infrastructure.
The chip that thinks like a brain
A startup called Extropic just announced a computing revolution.
The Thermodynamic Sampling Unit (TSU).
Instead of executing commands, it samples probabilities. Like a human brain predicting its next thought.
It’s reportedly 10,000× more efficient than GPUs.
If that holds, your AI cost curve collapses. Training that costs $10 million today could cost thousands. Every company becomes an AI company overnight.
Extropic is making computing more efficient. Scaling up energy production requires the support of a nation-state, but a more efficient computer can be built by a dozen people in a garage outside Boston.
Nvidia’s $1B Power Move
Nvidia just invested $1 billion in Nokia. The deal is Nokia builds 6G networks that run on Nvidia chips. Nvidia funds the network and sells it the hardware.
It’s vertical control, not really a synergy.
Nearly half of ChatGPT’s 800 million weekly users connect via mobile. 6G will need to handle billions of AI-native devices. Drones, AR glasses, autonomous robots.
All demanding inference in milliseconds.
This is where NVIDIA’s move becomes strategic gold.
The want to colonize the edge.
Expect every trillion-dollar firm to follow, funding their own supply chains to corner compute, data, and distribution. If your partners are building closed ecosystems, your procurement strategy becomes your survival plan.
Elon Musk wants to use every idle Tesla as a mini data center
10 million cars. 1 kilowatt each. That’s 100 gigawatts of distributed inference.
Your car drives by day, trains AI by night, and earns you money in between.
This is what the future of assets looks like: Nothing sits still. Everything works, learns, and earns.
For CFOs and senior finance leaders, I think it’s more of a mindset shift.
Ask where your idle assets could create compute value.
Of course, Microsoft Edge launched Copilot Mode
An agentic browser that works across Windows and Mac.
Every finance team will soon live inside an AI workspace that drafts reports, validates numbers, and summarizes data in real time.
Is it really time to question your browser?
I’ll leave that up to you.
The Bottom Line
Amazon just made massive layoffs. 14,000 people. And here’s the thing: these are corporate roles.
So they say they’re preparing for wide-scale AI adoption. But is that true?
Amazon overhired for years. They’re trimming layers, flattening the org, and getting lean. But here’s what’s different this time:
AI isn’t just helping them cut. It’s helping them rebuild.
Amazon’s CEO Andy Jassy said it himself:
As AI rolls out across Amazon, they’ll need fewer people doing the old jobs and more people doing entirely new ones.
This is the shift. This is the signal.
It’s not just layoffs. It’s a recomposition of the workforce.
14,000 thousand gone but a new kind of company emerging.
One where agents run workflows, and humans move higher up the stack.
Thinking, designing, supervising, creating.
And if you’re reading this, you’re already ahead of the curve.
Because what’s happening at Amazon right now is what’s coming for every enterprise.
The old structure will be gone.
The middle layers will be automated.
The new edge will be human + machine.
This isn’t the end of jobs.
It’s the beginning of the next era of work.
And you must adapt now.
Or you’ll be out of the game before it even starts.
And that’s all for today.
See you on Thursday!
Whenever you’re ready, there are 2 ways I can help you:
Advertise with the AI CFO Office. Sponsorships for 2026 are now open.
If you’re building an AI-powered CFO tech startup, I’d love to hear more and explore if it’s a fit for our investment portfolio.
I’m Wouter Born. A CFOTech investor, advisor, and founder of finstory.ai
Find me on LinkedIn