Google Just Raised Prices.
By Jose Ruiz · · 5 min read
Good morning, Crowd,
Jose here.
Thursday morning.
Four stories this week that prove the AI race just moved from "who has the best model" to "who can actually afford to run it."
Buckle up.
Read time: 5 mins
SPONSORED OF THE DAY
Check this Special Offer we brought you today!
The New Rules of Online Visibility
Your customers are searching in places your strategy doesn’t reach.
So before your business is buried and left behind, you need to understand the new rules of SEO.
BELAY's report explains how search is changing, what AI means for your visibility, and the practical steps small businesses like yours can take to stay visible.
BELAY’s U.S.-based Marketing Assistants turn strategy into execution, helping your business stay visible, credible, and competitive in every search.
LATESTDEVELOPMENTS
1️⃣ GOOGLE PIXEL PRICES JUST WENT UP $100 |
What: Google announced its new Pixel 11 phones will cost $100 more than the previous generation.
Why: Memory is expensive now. Like, really expensive.
One gigabyte of RAM cost $2.80 last year. Now it's $12. That's a 6x jump in one year.
Why the sudden spike? AI companies are buying up all the memory for their data centers, leaving slim pickings for phone makers.
The real talk: Google tried to absorb the costs. They failed. Now you pay more.
And it's not just Google. Every phone maker, laptop maker, and console maker is about to raise prices.
Jose's take: This is the clearest signal possible that AI infrastructure is running hot. When hardware makers have to raise consumer prices just to break even, the scarcity is real.
Capital is going to flow hard toward anyone who can solve the memory problem
AI |
2️⃣ MOONSHOT AI JUST BEAT OPENAI ON PRICE |
What: A Chinese AI startup called Moonshot released a model called Kimi K3 last month.
Now they're crashing the party.
The numbers: Moonshot's models are now in the top spots on OpenRouter (a platform that routes AI requests across different providers).
American developers are using it. American companies are using it. Why? It's way cheaper than OpenAI or Anthropic.
This week, Moonshot had to temporarily stop accepting new subscriptions because they got crushed by demand.
What's wild: They're not just cheaper. The model is actually good. Fast at coding. Fast at reasoning.
This isn't "cheap and bad." This is "cheap and good."
Jose's take: OpenAI and Anthropic just learned they're not protected by a moat. They're protected by network effects and relationships.
But if Moonshot proves you can get 80% of the capability for 20% of the cost, some customers will absolutely switch.
This is the first real competitive pressure on frontier AI labs. And it matters.
SAMSUNG |
3️⃣ SAMSUNG AND SK HYNIX JUST COMMITTED $518B TO BUILDING NEW CHIP FACTORIES |
What: South Korea announced an $880 billion investment plan for semiconductors, AI, and robotics.
Samsung and SK Hynix alone are committing $518 billion to building new chip manufacturing plants.
The timeline: These new factories take 5-10 years to build.
The signal: If you're betting half a trillion dollars on something, you believe it's going to matter for the next 20 years.
This isn't a "maybe" bet. This is an "all in" bet.
What it means: Memory and compute will stay expensive for the next 5-10 years.
Then, when these new factories come online around 2031-2035, prices crash and margins compress.
If you're an AI infrastructure company, you have a 5-year window of premium pricing. Use it wisely.
Jose's take: Samsung betting half a trillion dollars is the most bullish signal possible on AI.
It also means founders have a small window to build infrastructure moats before competition gets cheap.
Build fast. Build defensively. Get customers locked in.
After 2031, only the most efficient infrastructure companies survive.
PHARMA |
4️⃣ BRISTOL MYERS IS NOW BUYING NVIDIA CHIPS DIRECTLY |
BM
What: A major pharmaceutical company announced it's buying Nvidia's SuperPOD systems directly.
They're not renting from Amazon or Google. They're buying the chips themselves.
Why: Bristol Myers is using AI to speed up drug discovery. They reduce development timelines by 20-30%.
More capacity means screening more drug candidates faster.
That's worth buying your own infrastructure for.
The bigger story: For three years, the assumption was that only cloud companies would buy chips.
Everyone else would rent from AWS or Google.
Bristol Myers just proved that assumption wrong.
Large enterprises are going to start building their own AI infrastructure. Because cloud rental isn't fast enough. Or cheap enough. Or controllable enough.
Jose's take: The AI market is about to split into two.
On one side: Cloud companies renting AI capacity to smaller businesses.
On the other side: Large enterprises (pharma, oil, defense, finance) building their own AI infrastructure.
Both paths have capital flowing. Both paths have winners.
If you're a founder, pick one. Don't try to do both.
PREMIUM
📰Join Founderscrowd Premium
Join Founderscrowd Premium to get:
Unique investment opportunties
Our detailed valuation analysis (why this entry point makes sense)
The specific risks to watch before IPO
How to structure your position ($25K minimum)
Real-time alerts when secondary tranches open
Step-by-step process to invest
Premium members also get first dibs on allocation — the difference between getting in and getting locked out.
$100/month. Cancel anytime. No commitment.
This is where founders and serious investors actually allocate capital.
HERE'S THE PATTERN
All four stories point to the same thing:
The AI race isn't about models anymore. It's about hardware.
Models are table stakes. Everyone has a pretty good model now.
The constraint is compute. Memory. Power. Data center capacity.
Capital flows toward constraints.
So capital is flowing toward anyone who can:
Make compute cheaper (Etched, inference specialists)
Make memory cheaper (new fabs, alternative tech)
Make compute more efficient (optimization software)
Help you use expensive compute better (enterprise AI platforms)
If you're building an AI chatbot, you just got a harder road. Unit economics got worse because compute costs went up.
If you're building AI infrastructure, you just got a gift. Premium margins for the next 5 years.
WHAT FOUNDERS SHOULD DO
Building AI applications: Your margins are compressing. Either get more efficient or raise prices. You've got maybe 18 months before this gets painful.
Building AI infrastructure: You're in the best position possible. Shortage is real. Capital is flowing. Grab market share like your life depends on it. You've got 5 years of premium pricing.
Building enterprise AI: Pick a lane. Either build software so valuable that expensive compute doesn't matter. Or build infrastructure so efficient that compute costs don't matter.
WHAT THIS MEANS FOR YOUR WALLET
Phone prices are going up.
Memory prices are staying high.
New laptops will cost more.
Consoles will cost more.
This is the year that AI infrastructure costs finally show up in consumer prices.
Welcome to the real AI economy.
HUNT ACCORDINGLY ☕
Jose.
Keep reading — it's free.
Subscribe to unlock the full issue and get the next one in your inbox every morning. Already subscribed? Enter your email again.
Disclaimer: Founderscrowd is a media company. We are not a broker-dealer, funding portal, investment adviser or financial planner, and we do not collect or hold investor funds. All content is for informational and educational purposes only and is not investment, legal or tax advice or an offer to buy or sell any security. Views on any company or deal are our writers' opinions. Some content is sponsored. Private market investments are speculative, illiquid and may result in the loss of your entire investment; many offerings are limited to accredited investors. Read full disclaimer
