I don’t think the next AI compute wave will look like the last one.
Training was big clusters and big headlines.
Agents feel different.
Smaller jobs.
Running constantly.
Happening in more places.
Messier, but probably much bigger than people expect. https://t.co/5NJBwZaBZV
— PAN (@Pan_Ecosystem)
Aug 7, 2026
Training was big clusters and big headlines.
Agents feel different.
Smaller jobs.
Running constantly.
Happening in more places.
Messier, but probably much bigger than people expect. https://t.co/5NJBwZaBZV
— PAN (@Pan_Ecosystem)
Aug 7, 2026
X (formerly Twitter)
PAN (@Pan_Ecosystem) on X
I don’t think the next AI compute wave will look like the last one.
Training was big clusters and big headlines.
Agents feel different.
Smaller jobs.
Running constantly.
Happening in more places.
Messier, but probably much bigger than people expect.
Training was big clusters and big headlines.
Agents feel different.
Smaller jobs.
Running constantly.
Happening in more places.
Messier, but probably much bigger than people expect.
We’re not trying to make smarter agents.
We’re trying to make agent execution something you can actually verify and settle.
Intent, compute, payment, result.
PAN sits in that loop.
— PAN (@Pan_Ecosystem)
Aug 9, 2026
We’re trying to make agent execution something you can actually verify and settle.
Intent, compute, payment, result.
PAN sits in that loop.
— PAN (@Pan_Ecosystem)
Aug 9, 2026
X (formerly Twitter)
PAN (@Pan_Ecosystem) on X
We’re not trying to make smarter agents.
We’re trying to make agent execution something you can actually verify and settle.
Intent, compute, payment, result.
PAN sits in that loop.
We’re trying to make agent execution something you can actually verify and settle.
Intent, compute, payment, result.
PAN sits in that loop.
Official Announcement
Thrilled to partner with @XAgent_official the zero-code platform empowering anyone to Speak to Build custom AI Agents.
As the Demand Layer for AI Compute PAN will explore seamless compute access agent-to-agent payments and scalable infrastructure for the https://t.co/mYgOtdDTjz
— PAN (@Pan_Ecosystem)
Aug 10, 2026
Thrilled to partner with @XAgent_official the zero-code platform empowering anyone to Speak to Build custom AI Agents.
As the Demand Layer for AI Compute PAN will explore seamless compute access agent-to-agent payments and scalable infrastructure for the https://t.co/mYgOtdDTjz
— PAN (@Pan_Ecosystem)
Aug 10, 2026
X (formerly Twitter)
PAN (@Pan_Ecosystem) on X
Official Announcement
Thrilled to partner with @XAgent_official the zero-code platform empowering anyone to Speak to Build custom AI Agents.
As the Demand Layer for AI Compute PAN will explore seamless compute access agent-to-agent payments and scalable…
Thrilled to partner with @XAgent_official the zero-code platform empowering anyone to Speak to Build custom AI Agents.
As the Demand Layer for AI Compute PAN will explore seamless compute access agent-to-agent payments and scalable…
Training compute is easy to recognize.
Big clusters.
Big budgets.
Big announcements.
Agent compute will be harder to see.
Small jobs.
Always running.
Spread across tools, apps, and users.
Paid for one action at a time.
That market may be much larger than it looks.
— PAN (@Pan_Ecosystem)
Aug 11, 2026
Big clusters.
Big budgets.
Big announcements.
Agent compute will be harder to see.
Small jobs.
Always running.
Spread across tools, apps, and users.
Paid for one action at a time.
That market may be much larger than it looks.
— PAN (@Pan_Ecosystem)
Aug 11, 2026
X (formerly Twitter)
PAN (@Pan_Ecosystem) on X
Training compute is easy to recognize.
Big clusters.
Big budgets.
Big announcements.
Agent compute will be harder to see.
Small jobs.
Always running.
Spread across tools, apps, and users.
Paid for one action at a time.
That market may be much larger than…
Big clusters.
Big budgets.
Big announcements.
Agent compute will be harder to see.
Small jobs.
Always running.
Spread across tools, apps, and users.
Paid for one action at a time.
That market may be much larger than…
Training compute is easy to spot.
Big clusters.
Big budgets.
Big announcements.
Agent compute will be harder to see.
Small jobs.
Always running.
Spread across tools, apps, and users.
Paid for one action at a time.
The biggest compute market may be the one nobody sees. https://t.co/rI6uYqRSgb
— PAN (@Pan_Ecosystem)
Aug 13, 2026
Big clusters.
Big budgets.
Big announcements.
Agent compute will be harder to see.
Small jobs.
Always running.
Spread across tools, apps, and users.
Paid for one action at a time.
The biggest compute market may be the one nobody sees. https://t.co/rI6uYqRSgb
— PAN (@Pan_Ecosystem)
Aug 13, 2026
X (formerly Twitter)
PAN (@Pan_Ecosystem) on X
Training compute is easy to spot.
Big clusters.
Big budgets.
Big announcements.
Agent compute will be harder to see.
Small jobs.
Always running.
Spread across tools, apps, and users.
Paid for one action at a time.
The biggest compute market may be the…
Big clusters.
Big budgets.
Big announcements.
Agent compute will be harder to see.
Small jobs.
Always running.
Spread across tools, apps, and users.
Paid for one action at a time.
The biggest compute market may be the…
Agents need receipts.
Not screenshots.
Not vague logs.
A real one should show:
intent → compute → verification → payment → result.
PAN is building that execution record. https://t.co/i2DVVPjjDD
— PAN (@Pan_Ecosystem)
Aug 14, 2026
Not screenshots.
Not vague logs.
A real one should show:
intent → compute → verification → payment → result.
PAN is building that execution record. https://t.co/i2DVVPjjDD
— PAN (@Pan_Ecosystem)
Aug 14, 2026
X (formerly Twitter)
PAN (@Pan_Ecosystem) on X
Agents need receipts.
Not screenshots.
Not vague logs.
A real one should show:
intent → compute → verification → payment → result.
PAN is building that execution record.
Not screenshots.
Not vague logs.
A real one should show:
intent → compute → verification → payment → result.
PAN is building that execution record.
Open models do not just create more builders.
They create more agent attempts.
More experiments.
More tool calls.
More failed runs.
More inference demand.
More need to verify what actually happened.
The open-source wave expands the execution market, not just the model market. https://t.co/Sg0KVkPqk7
— PAN (@Pan_Ecosystem)
Aug 16, 2026
They create more agent attempts.
More experiments.
More tool calls.
More failed runs.
More inference demand.
More need to verify what actually happened.
The open-source wave expands the execution market, not just the model market. https://t.co/Sg0KVkPqk7
— PAN (@Pan_Ecosystem)
Aug 16, 2026
X (formerly Twitter)
PAN (@Pan_Ecosystem) on X
Open models do not just create more builders.
They create more agent attempts.
More experiments.
More tool calls.
More failed runs.
More inference demand.
More need to verify what actually happened.
The open-source wave expands the execution market, not…
They create more agent attempts.
More experiments.
More tool calls.
More failed runs.
More inference demand.
More need to verify what actually happened.
The open-source wave expands the execution market, not…
A wallet is not an approval system.
If an agent can spend,
you need to define:
what it can buy,
how much it can spend,
and when a human must approve. https://t.co/sdhiFPaKlh
— PAN (@Pan_Ecosystem)
Aug 18, 2026
If an agent can spend,
you need to define:
what it can buy,
how much it can spend,
and when a human must approve. https://t.co/sdhiFPaKlh
— PAN (@Pan_Ecosystem)
Aug 18, 2026
X (formerly Twitter)
PAN (@Pan_Ecosystem) on X
A wallet is not an approval system.
If an agent can spend,
you need to define:
what it can buy,
how much it can spend,
and when a human must approve.
If an agent can spend,
you need to define:
what it can buy,
how much it can spend,
and when a human must approve.
14,000 designs. 9 binders. 3 sub-nanomolar.
The agent didn't win because it was smart.
It won because the loop was closed:
design → wet-lab validate → feed back.
Autonomous execution only pays off
when there's a checkpoint at the end. https://t.co/zyvvLjIMdN
— PAN (@Pan_Ecosystem)
Aug 18, 2026
The agent didn't win because it was smart.
It won because the loop was closed:
design → wet-lab validate → feed back.
Autonomous execution only pays off
when there's a checkpoint at the end. https://t.co/zyvvLjIMdN
— PAN (@Pan_Ecosystem)
Aug 18, 2026
X (formerly Twitter)
NVIDIA Healthcare (@NVIDIAHealth) on X
Using NVIDIA Proteina-Complexa, @muni_bio's autoresearch agent explored nearly 14,000 protein designs.
@adaptyvbio validated nine TREM2 binders, three with sub-nanomolar affinity.
The results show how agents can connect computational design with wet-lab…
@adaptyvbio validated nine TREM2 binders, three with sub-nanomolar affinity.
The results show how agents can connect computational design with wet-lab…
Agent compute is not one big bill.
It is hundreds of small executions
that become visible only after they add up. https://t.co/tRbmilJYPy
— PAN (@Pan_Ecosystem)
Aug 20, 2026
It is hundreds of small executions
that become visible only after they add up. https://t.co/tRbmilJYPy
— PAN (@Pan_Ecosystem)
Aug 20, 2026
X (formerly Twitter)
PAN (@Pan_Ecosystem) on X
Agent compute is not one big bill.
It is hundreds of small executions
that become visible only after they add up.
It is hundreds of small executions
that become visible only after they add up.
Idle GPUs aren’t compute.
If developers can’t discover them, verify them, run jobs to completion, and pay for actual usage, they’re just hardware.
— PAN (@Pan_Ecosystem)
Aug 23, 2026
If developers can’t discover them, verify them, run jobs to completion, and pay for actual usage, they’re just hardware.
— PAN (@Pan_Ecosystem)
Aug 23, 2026
X (formerly Twitter)
PAN (@Pan_Ecosystem) on X
Idle GPUs aren’t compute.
If developers can’t discover them, verify them, run jobs to completion, and pay for actual usage, they’re just hardware.
If developers can’t discover them, verify them, run jobs to completion, and pay for actual usage, they’re just hardware.