.@akashnetAI processed over 7 billion tokens yesterday, a new ATH and has been consistently processing over 5 billion tokens everyday on an average.
Another proof point that Inference on decentralized compute is works great at scale.
Akash Mainnet 18 upgrade will occur at block #27230465 on Thursday, June 11th, at ~14:00 UTC.
This upgrade enables:
→ Oracle v2, replacing block-height references with wall-clock timestamps for staleness detection, TWAP, and price queries.
→ Resource Reclamation (AEP-82), giving providers a negotiated grace period before lease termination instead of immediate closure.
→ Market Order Close Event fix, ensuring EventOrderClosed fires when a deployment closes while its order is still open.
Current deployments will NOT be affected.
.@akashnetAI processed over 7 billion tokens yesterday, a new ATH and has been consistently processing over 5 billion tokens everyday on an average.
Another proof point that Inference on decentralized compute is works great at scale.
Razer is pitching their Blade 18 as your local AI stack.
In our latest podcast, we broke down how meaningful it would be if Razer systems could take part in @akashnet HomeNode, with hardware like this earning $AKT when you're not using it.
Your laptop. Your GPU. Your revenue.
$AKT remains a top project for me.
I have to say, @gregosuri is one of the most dedicated and ethical founders in the space. Never witnessed a single red flag over the years. Happy to have been part of the community since day 1.
With nonstop innovation Akash will survive and thrive!
Funny how the crowd always forms after @nvidia says it.
@gregosuri has been saying the same thing for years:
There are millions of idle GPUs already plugged into the grid.
Don't build the datacenter.
Unlock it. 🔥
#DePIN#Ai#GPU $AKT #web3#DeAI@akashnet#homenode
I just realised that Google might actually be buying this compute in part to stop anyone else from buying that capacity.
Blocking other companies from buying compute could alone be worth billions of dollars (making sure competitors like Anthropic are capacity constrained).
We risk losing control of AI.
That was @AnthropicAI's warning yesterday, when they called for a global pause in AI development.
The truth is something very different.
"We" never had control. They did. From the beginning.
@OpenAI, @AnthropicAI, and @Google account for 89% of all AI revenue.
Centralized clouds and AI companies have hoarded GPUs, forced out competition, and set the agenda for what gets built, and who gets to build it.
But AI isn't the problem. Concentrated power is.
When compute is accessible and affordable, and models are open-source, everyone, everywhere can build.
That changes everything.
It's not technology engineered for corporate profits. It's technology built for people.
That's what @ionet is here for.
https://t.co/yJIA1KRh0H
Robots using $AKT for compute. Virtual worlds on demand. Claude Code on @akashnetAI models. A brand new Console Air. And idle 4090s earning ~$10/day on homenode.
A lot happened on @akashnet last month.
Sit down, you don't want to miss this one:
https://t.co/wFdbnjJco9
Greg Osuri on the supply chain crisis hiding behind the AI boom
"Most new data centers are using LNG. There's no other way to power these systems"
"Now we have a 4-year lead time to get a transformer and a 14-year lead time to get a turbine"
"Every part of the supply chain is constrained right now"