What actually changed across these 3 phases?
Generative AI → produces tokens
Reasoning AI → improves quality of tokens
Agentic AI → turns tokens into actions
That’s why this isn’t just a model upgrade cycle.
It’s a shift in how compute is consumed.
🧵 1/At a recent talk with Morgan Stanley, Jensen Huang (CEO of NVIDIA) said:
“In the last 2 years, we went through three inflection points in AI”
He’s referring to a specific shift:
1. Generative AI
2. Reasoning AI
3. Agentic AI
Each one changed how AI is used in the real world.
4. Agentic AI
This is the current inflection and the most important one.
The shift is simple but profound:
AI is no longer just answering questions.
It’s starting to do things.
- use tools
- access files
- browse
- plan
- execute tasks over time
Jensen Huang (NVIDIA CEO):
“If you don’t own everything, you have a 0% chance.”
Next era of computing is clear:
Partial ownership isn’t enough, control the stack or you’re out.
Agentic AI isn’t theory anymore.
The future isn’t devices, it’s autonomous systems working 24/7.
Meanwhile $TAO is ripping (+15–28%), back above $300.
Subnets like @tplr_ai are exploding.
Open + proprietary AI aren’t rivals—they’re complements.
DeAI summer loading? 🔥
Jensen Huang (CEO of NVIDIA) just gave a major nod to decentralized AI 👀
On All-In, Covenant-72B was highlighted:
→ 72B params
→ Trained permissionlessly
→ 70+ contributors
→ Everyday hardware
Built on Bittensor.
No clusters. No gatekeepers. Just distributed compute.
Jensen called it a “pretty crazy technical accomplishment”
and compared Bittensor to a “modern Folding@Home.”
That’s NVIDIA validating DeAI as legit + scalable.
NVIDIA just revealed something bigger than AI tools.
At GTC 2026, Jensen Huang broke down a concept that most people are still underestimating:
OpenClaw is not just a framework. It’s an operating system for AI agents.
Here are the 5 most important insights:
We are witnessing the birth of a new computing model:
- Not apps
- Not copilots
But autonomous agent systems
The real question now isn’t:
“How do we use AI?”
It’s:
“How do we design, control, and scale AI agents safely?”