Guides
All analyses › Nate B Jones
You’re on YouTube for AI entrepreneurs. Read the analysis and decide whether the video is worth your time.All analyses →
AI Just Got Cheaper. Here's Why That's a Problem for NVIDIA.
Author’s title: AI Just Got Cheaper. Here's Why That's a Problem for NVIDIA.

Actually: Businesses are switching to cheaper, specialized models, causing flat revenue despite skyrocketing usage.

N
Nate B JonesAI strategy for executives, without the hype · 107K views · 2 days ago
6/10
BUSINESS SCORE
📚 General overview
Practicality5
Specifics7
Novelty6
✓ The analysis is enough. Summary covers the economic shift and cost-saving tactics. The summary is sufficient for most business owners. Watch the first 5 minutes if you are specifically looking to optimize your AI API costs. · video 11:00 · ✓ analysis double-checked against the video

🔎 What’s inside

Businesses are shifting from expensive 'frontier' AI models to cheaper, specialized ones, causing a disconnect where AI usage is skyrocketing but revenue for providers remains relatively flat.
Promise vs. reality
Honest. The video explains the economic shift where cheaper intelligence reduces the profit margins for chipmakers and AI labs despite high demand.

⚡ The gist in 30 seconds

  • Businesses are 'trading down' to cheaper AI models that handle 80-90% of tasks at a fraction of the cost, leading to flat revenue growth for labs despite high usage.
  • NVIDIA faces a potential oversupply 'avalanche' because they committed to production based on growth forecasts that may not match actual business spending.
  • New tools like Google's 'Orchestrator' automatically route tasks to the cheapest capable model, further reducing the need for expensive frontier models like GPT-4.
  • AI adoption is shifting from engineering to customer support, where fast, economical answers are prioritized over deep reasoning.
  • Over-reliance on AI-generated code and the loss of senior engineering expertise are cited as risks to system reliability, as seen in recent major tech outages.
👤 Nate B Jones, a tech analyst and commentator with professional experience at Amazon.

⏱ Key moments click — the video jumps there

Ramp AI Index shows token spending remained flat across August and September.
Hypothetical 'Usage vs. Bill' scenario where usage goes up but the bill stays flat due to cheaper models.
Introduction of Google Gemini's specialized models: Carbon (reasoning) and Hydrogen (speed).
Customer Success AI spend grew from 0.8% in June to 13.3% in October.
Discussion of NVIDIA's financial exposure due to committed chip supply.
Research finding Blackwell chip fingerprints in Chinese open-source models despite export controls.
Analysis of the Microsoft 365 outage and its relation to engineering culture.

💰 Numbers

Token spending growth (Aug-Sept)Flat (Ramp data)
Total AI spend increase (Oct)+1.65% (Ramp data)
Frontier models share of spend69.5% (Ramp data)
Customer Success AI spend13.3% of total (author's claim)
Google Orchestrator efficiency24% less quota used (Google claim)
Numbers are quoted from the video authors.

🛠 Tools from the video

Ramp — Spend management platform providing AI market dataGoogle Gemini Carbon — AI model optimized for deep reasoningGoogle Gemini Hydrogen — AI model optimized for fast executionGoogle Orchestrator — Tool that auto-switches between AI models to save costsNVIDIA Blackwell — High-end AI chips subject to export controls

🚀 How to apply it in your business

  1. Audit your AI API usage to identify tasks currently using expensive models that could be handled by 'mini' or 'flash' versions.
  2. Implement a routing layer in your AI applications to send simple queries to low-cost models automatically.
  3. Focus AI implementation on customer support and routine operations where speed and cost-efficiency provide the highest ROI.
  4. Maintain human oversight and senior engineering review for all AI-generated code to prevent long-term technical debt.
  5. Monitor industry spend indexes (like Ramp) to ensure your AI budget is aligned with market benchmarks.

Who will this analysis save an hour?We’ll send the link with the “actually” takeaway — they won’t have to watch the whole video.

⚠️ Watch out: The data regarding spending trends is heavily reliant on reports from Ramp, which represents a specific subset of business users and may not reflect the entire market.

❓ Questions and answers

Why is AI getting cheaper for businesses?

Businesses are becoming 'smarter' by using smaller, specialized models and automated routing tools to avoid paying for expensive frontier intelligence for every task.

Is NVIDIA in financial trouble?

The video suggests a risk of oversupply if the massive capacity NVIDIA built doesn't match the slower-than-expected growth in actual customer spending.

What are Gemini Carbon and Hydrogen?

These are Google's new specialized models: Carbon is for complex reasoning, while Hydrogen is optimized for speed and lower cost.

How can I reduce my company's AI costs?

By using an orchestrator or manual routing to ensure only the most difficult problems use the most expensive models like GPT-4 or Gemini Ultra.

Why are tech outages becoming more common?

The author argues that a declining engineering culture, loss of senior talent, and over-reliance on AI-generated fixes may be hindering system reliability.

Found a mistake in the analysis? Tell us what’s wrong and we’ll fix it.
Want analyses like this for your niche? We’ll send the best videos for your business.