☕ The grid is 70% empty. A Stanford startup just unlocked it for $64M.

GridCARE just unlocked hidden power in the grid worth $10B. Isomorphic Labs raised $2.1B for AI drug discovery. CREATE Medicines got $122M to cure cancer with one injection. This week: infrastructure solved problems everyone said were impossible.

Good afternoon! Quick question: What's the biggest bottleneck stopping AI from scaling?

Not chips. Not data. Not models.

Power.

Data centers need electricity. Lots of it. And utilities say it takes 6 to 10 years to connect new facilities to the grid. That's not a bottleneck. That's a brick wall.

Until now.

A Stanford spinout just raised $64 million to unlock power that's already there—hiding in plain sight in the grid we already built. They're compressing interconnection timelines from years to months. First customer got 80 megawatts in 2026. More coming.

Also this week: Google's AI drug discovery lab raised $2.1 billion (one of the biggest biotech rounds ever). And a gene therapy company got $122 million to cure blood cancers with a single injection instead of months of chemo.

Three companies. Three impossible problems. Three confirmed raises this week.

Let's break them down. ☕

—Alberto, Founderscrowd

💰 This Week's Top 3 Confirmed Raises

🔌 GRIDCARE: $64M TO UNLOCK HIDDEN POWER

What happened:

GridCARE raised $64 million in an oversubscribed Series A led by Sutter Hill Ventures (early Nvidia investor). Other investors: John Doerr, National Grid Partners, Future Energy Ventures.

The raise:

• Amount: $64M Series A (oversubscribed)
• Lead investor: Sutter Hill Ventures
• Previous round: $13.5M seed (May 2025)
• Total raised: $77.5M
• Founded: 2025 (Stanford Sustainability Accelerator)

The problem:

Data centers are the factories of the AI economy. But they need massive amounts of power. A single AI data center can use 100+ megawatts. That's enough to power 80,000 homes.

Connecting to the grid takes 6 to 10 years. Utilities say there's no capacity. Developers wait. AI projects stall.

The solution:

GridCARE discovered something shocking: The grid is only 30% utilized most of the year. The other 70%? Sitting idle.

Why? Because utilities can't see it. Traditional grid planning assumes worst-case scenarios (peak demand on the hottest day of the year). But those scenarios almost never happen.

GridCARE's Energize platform uses physics-based AI to model quadrillions of grid scenarios in real time: weather patterns, congestion points, outage history, demand variability.

It finds hidden capacity and activates it. Interconnection timelines drop from years to months.

⚡ GridCARE: Hidden Power in the Grid

The traction:

Portland General Electric (October 2025):

• Unlocked 400 megawatts in Hillsboro, Oregon
• First 80 megawatts online in 2026
• Full 400MW by 2029

Current pipeline:

• 2+ gigawatts across 12+ markets
• $10 billion in economic value unlocked
• National Grid now a customer AND investor

Why this matters:

Power is the new constraint for AI. Not chips. Not talent. Power.

OpenAI, Anthropic, Google—they all need data centers. Data centers need power. Power takes years to connect. GridCARE solves this.

Bottom line:

GridCARE isn't building new power. They're unlocking power that already exists but utilities can't see. If this works, they just solved AI's biggest bottleneck. Sutter Hill (early Nvidia investor) betting $64M says it works.

💊 ISOMORPHIC LABS: $2.1B FOR AI DRUG DISCOVERY

What happened:

Isomorphic Labs (Google DeepMind's drug discovery spinout) raised $2.1 billion in a Series B. This is one of the largest biotech funding rounds ever.

What they do:

Isomorphic uses AI (specifically AlphaFold, DeepMind's protein-folding model) to discover new drugs. Traditional drug discovery takes 10+ years and costs $2+ billion per drug. Most fail.

Isomorphic's thesis: AI can predict how proteins fold, which reveals how drugs will interact with the body. This compresses discovery timelines from years to months.

The partnerships:

Isomorphic has multi-billion-dollar deals with:

• Eli Lilly: Up to $2.9B
• Novartis: Up to $2.7B

These aren't research grants. These are commercial partnerships. Big pharma is betting billions that Isomorphic can find drugs faster than traditional methods.

Why this matters:

Drug discovery is broken. It takes too long, costs too much, and fails too often. AI is the only path to fixing it.

AlphaFold (Isomorphic's foundation) won the Nobel Prize in Chemistry in 2024. Demis Hassabis is one of the most credible AI researchers on the planet.

Bottom line:

Google just bet $2.1 billion that AI can revolutionize drug discovery. Given AlphaFold's track record (solved protein folding, a 50-year-old problem), this isn't a gamble. It's a calculated bet on the future of medicine. Pharma companies already committed $5.6B in partnerships. The market believes.

🧬 CREATE MEDICINES: $122M TO CURE CANCER WITH ONE INJECTION

What happened:

CREATE Medicines raised $122 million in a Series B to advance in vivo CAR-T therapy for blood cancers.

The problem:

CAR-T therapy works. It cures certain blood cancers (leukemia, lymphoma) that were previously untreatable. Survival rates jump from 30% to 80%+.

But it's brutal:

• Costs $400,000 to $500,000 per patient
• Requires extracting immune cells, reprogramming them in a lab, and re-infusing them
• Takes months
• Only available at specialized centers
• Patients often can't wait

The solution:

CREATE Medicines is developing in vivo CAR-T. Instead of extracting cells, they inject a gene therapy directly into the patient. The therapy reprograms immune cells inside the body to attack cancer.

One injection. Done.

Why this matters:

CAR-T is one of the most important medical breakthroughs in 20 years. But it's inaccessible. Only ~10,000 patients per year get it in the U.S.

In vivo CAR-T could scale to millions.

If CREATE succeeds:

• Cost drops from $500K to ~$50K
• Treatment time drops from months to days
• Accessibility increases 100x

Bottom line:

CREATE Medicines is trying to turn a $500,000, months-long cancer treatment into a single injection. If this works, it's a Nobel Prize and a $10B+ exit. If it doesn't, it's still advancing one of the most promising areas in medicine. $122 million says investors believe.

🎯 THE BIG PICTURE

Let me connect these three stories:

GridCARE: Unlocking hidden power in the grid
Isomorphic Labs: AI drug discovery
CREATE Medicines: One-injection cancer cure

1. All three solve "impossible" problems.

• GridCARE: Power takes 6-10 years to connect. They do it in months.
• Isomorphic: Drug discovery takes 10+ years. AI does it faster.
• CREATE: CAR-T takes months and $500K. They do it in one injection.

2. All three are infrastructure plays.

GridCARE isn't building new power plants. They're unlocking existing capacity. Isomorphic isn't a pharma company. They're AI infrastructure for drug discovery. CREATE isn't inventing CAR-T. They're making it accessible.

3. All three have massive markets.

• GridCARE: Every AI data center needs power (hundreds of billions)
• Isomorphic: Drug discovery is a $2 trillion market
• CREATE: CAR-T market could hit $100B+ with accessibility

4. All three raised this week.

Private markets are betting on infrastructure that solves hard, boring problems with massive budgets behind them. Not apps. Not wrappers. Infrastructure.

🔥 THE BOTTOM LINE

GridCARE unlocked $10 billion in hidden grid capacity. Isomorphic Labs raised $2.1 billion to compress drug discovery from 10 years to months. CREATE Medicines got $122 million to turn a $500K cancer treatment into one injection.

The lesson: The biggest opportunities in private markets aren't building new things. They're unlocking things that already exist but everyone said were impossible.

GridCARE didn't build new power plants. They found power hiding in the grid. Isomorphic didn't invent new biology. They used AI to see proteins faster. CREATE didn't discover CAR-T. They made it accessible.

Infrastructure that unlocks constraints beats infrastructure that builds from scratch.

Follow the constraints. Bet on companies removing them. That's where the billion-dollar exits hide.

See you tomorrow for more private market insights. Until then, keep hunting. ☕

—Alberto
Founderscrowd