- AI Fire
- Posts
- ⚠ 'Vibey' GPT‐5 = Whole Crisis?
Read time: 5 minutes
AI is testing the limits of money. Big Tech is pouring billions into data centers, but half the tab is being pushed onto Wall Street. Does the AI gold rush just slow down?
What are on FIRE 🔥
IN PARTNERSHIP WITH SECTION
On September 25, get the expert formula for turning AI spend into ROI at your oganization at Section's virtual AI ROI conference.
No gatekeeping, no pitching, no BS. Hear from top AI leaders like Scott Galloway, Mo Gawdat, and May Habib and leave with a clear path to AI ROI for your company.
AI INSIGHTS
What’s the biggest capital project of our lifetime looks like? (Hint: not railroads, not highways, not even cloud.) It’s AI compute, and the bill is staggering.
Morgan Stanley says we’re on track for $2.9 trillion in AI data center spend by 2029. Big Tech will cover about half… leaving a $1.5 trillion funding hole. And the projects aren’t small.
Meta → “Prometheus” supercluster ($100B+)
xAI → “Colossus” ($100B+)
OpenAI → “Stargate” ($100B+)
By 2026, Google, Amazon, Microsoft, and Meta alone will spend $400B on data centers in a single year. So who’s actually paying? The short answer: borrow like crazy:
2025 → $60B in loans already funneled into ~$440B of projects
Meta’s Ohio & Louisiana centers → $29B raised, $26B of that pure debt
It’s Wall Street’s cash, leveraged to the hilt. There’re some red flags nobody wants to talk about: Overcapacity risk, obsolescence risk, exit risk.
→ It’s déjà vu of past infrastructure bubbles: railroads, telecom, dot-com. Same story, new GPUs.
Why It Matters: Instead of slowing down, they’re outsourcing the risk, letting private capital eat the downside if demand doesn’t match the build-out. It’s a financial system stress test. AI’s biggest risk might not be technical.
PRESENTED BY BELAY
Economic pressure is rising, and doing more with less has become the new reality. But surviving a downturn isn’t about stretching yourself thinner; it’s about protecting what matters most.
BELAY matches leaders with fractional, cost-effective support — exceptional Executive Assistants, Accounting Professionals, and Marketing Assistants — tailored to your unique needs. When you're buried in low-level tasks, you lose the focus, energy, and strategy it takes to lead through challenging times.
BELAY helps you stay ready for whatever comes next.
TODAY IN AI
AI HIGHLIGHTS
🖼 Google’s magen 4 image generator is now GA in the Google AI studio, with up to 2K resolution and a new fast model for quicker outputs. You should give it a try here.
🕹️ In a viral test on X, GPT-5 earned 8 badges in the Pokemon Red video game in just 6,000 steps, almost 70% fewer than o3, which just came out 6 mons ago. See the test.
🚀 Cognition just gathered SF’s most ambitious AI builders to push GPT-5, Claude Opus, Gemini Pro to their limits. And an ex-OpenAI mem picked the very best. Here’re top demos featured.
😡 ‘Cheapfake’ AI celeb videos are rage-baiting people on YouTube. 100+ channels used AI to create lazy clips. Watch this viral video (25.3M likes), see if it’s fake (or not?)
😶 Anthropic gave Claude Opus 4 & 4.1 the rare ability to end a conversation (not to ghost you), but only in extreme cases. It can say no & walk away. Here’s an example.
📓 GPT, Gemini, now we got Claude 'Learning Mode'. It will help users learn by guiding them to the answers instead of handing them over. Watch it in action here.
💰 AI Daily Fundraising: Meta is raising up to $29 billion for AI projects, including a $14.3 billion investment in Scale AI, securing a 49% stake and involving Scale's CEO in Meta's AI team.
AI SOURCES FROM AI FIRE
NEW EMPOWERED AI TOOLS
AI QUICK HITS
🎶 ElevenLabs Jingle Maker was built to turn any website into a Jingle.
🤑 OpenAI might add ads to ChatGPT, eyeing big earnings by 2029.
💼 Meta’s Superintelligence Lab just added 3 more OpenAI researchers.
📈 ChatGPT's app revenue skyrocketed by 673% in 2025, now $1.35B.
🗣️ NVIDIA launched Granary over 1M hours of audio in 25 languages.
AI CHART
OpenAI finally dropped GPT-5, its most-hyped model ever, instead of jaw-dropping demos or AGI-level breakthroughs, people got a mid-tier upgrade that left the community wondering:
Is AI slowing down?
So, What’s New in GPT‑5? Not much. A bit faster & cheaper. The “vibes” are better. Even AI insiders were surprised: “People expected something totally new. We didn’t get that.”
GPT‑5 is fine. Just don’t expect magic. The whole vibe of the AI race just shifted. Before:
“We’re building AGI”
“Every model is exponentially smarter”
“We’ll automate everything soon”
Now:
“Let’s build useful products”
“Maybe AGI is further off than we thought”
“LLMs are infrastructure, not magic”
Under Biden, AI regulation was about existential risk and AGI safety. Under Trump? The mood is way more chill. We’re entering the “post-hype” era of AI, where the smartest people in the room are shifting their goals.
We read your emails, comments, and poll replies daily
How would you rate today’s newsletter?Your feedback helps us create the best newsletter possible |
Hit reply and say Hello – we'd love to hear from you!
Like what you're reading? Forward it to friends, and they can sign up here.
Cheers,
The AI Fire Team
Reply