The race to AGI is a battle for compute, and how DePIN can counter it
Quote725 words
Argues AGI development is bottlenecked by finite compute, with Nvidia at about 80% of GPU supply and big clouds absorbing most of it. Says geo-distributed clustering by DePINs like io.net can even the playing field.
Key claims
Nvidia holds about 80% of GPU production
Most GPUs go to major cloud providers and AI giants
The race to Artificial General Intelligence is a battle for compute. 🤖 ✨
Billions of $$$ are spent to amass a finite resource.
Compute.
Why?
How can decentralized networks even the playing field?
And WTF is Artificial General Intelligence?
AGI refers to artificial intelligence capable of understanding, learning, and applying knowledge across a wide range of tasks with human-like proficiency.
Unlike narrow (use case specific) AI, which is designed to perform certain tasks (like language translation or image recognition), AGI would be capable of performing any intellectual task that a human can do (just better and quicker).
Why does AGI matter?
It matters for two main reasons.
☞ The Military
☞ The Economy
In military terms, AGI can improve strategic planning, automate battlefield decisions, optimize logistics, and manage advanced defense systems with unmatched speed & precision.
Its offensive potential includes AGI-based weapons that can autonomously identify, target & engage threats, execute cyberattacks, and deploy autonomous drones or robotic units, making it a significant force multiplier that can decide over global military supremacy.
In economic terms, AGI can drive innovation, optimize financial markets, automate business processes & analyze global trends, leading to increased efficiency, reduced costs, and likely a significant competitive advantage in global markets for the ones in control of this technology.
Not surprising, that the idea of AGI is often associated with significant ethical, societal & philosophical questions and concerns as it could potentially surpass human intelligence quickly, significantly impacting many aspects of society, including employment or global security.
What is the current state of AGI?
AGI is a theoretical concept at this stage, as no existing AI systems possess the broad, adaptable intelligence that AGI would require.
However, many companies (including @OpenAI & co) as well as state actors are working towards this all-mighty form of artificial intelligence.
The development of AGI involves solving complex challenges in areas like understanding and reasoning, learning from small amounts of data, and generalizing knowledge across different domains.
That's where we tie it all back together with the initial question:
Why is the race to AGI a battle for compute?
The creation of AGI involves training large-scale models on massive datasets.
The more data and the more complex the model, the more computational power is required.
Deploying AGI will mean running these models across various environments and tasks, which requires ongoing and substantial computational power to handle the diversity & complexity of tasks an AGI might encounter.
Current state-of-the-art models already require extensive compute, and AGI will likely push these requirements to new heights.
But even after training, AGI is expected to process & analyze data in real time, understanding and responding to human-style tasks with little to no delay.
Achieving this requires not just huge amounts of, but also highly efficient compute resources to run inference at scale.
Compute is a finite resource and that is evidenced by the global GPU shortage that results from the insatiable demand of tech companies around the world to get access to high-performance GPUs.
With the supply side of the market dominated by essentially a single company (Nvidia) in terms of production (80% of market), access to compute in general is very exclusive.
To make matters worse, most of the GPUs that are being produced and sold end up in the hands of major cloud compute providers such as @awscloud, @googlecloud or @Azure or directly in the hands of AI giants that have raised billions of dollars in funding (e.g. @OpenAI) to be first to reach AGI.
By controlling the major part of supply, these actors effectively gain an unfair advantage over other players in the market that can not compete with the unlimited resources of these companies.
However, that doesn't mean that the end of the story is already written.
Decentralized networks that can cluster compute from geo-distributed locations can fix this.
How?
By clustering under-utilized compute resources (GPUs) across the globe, DePINs like @ionet open up a new source of compute supply.
They can amass not only large amounts of compute, but also serve it to endusers (e.g. AI/ML companies) at a competitive price, evening the playing field and countering the monopolizing pressure arising from the centralization of traditional compute markets.
People don't understand yet how important these geo-distributed clustering layers are. That's why they are sleeping on $IO.
Soon everyone will wake up.
NFA but DYOR anon.
QUOTING @zerokn0wledge_
Is the world at risk of facing an AI dystopia? The big 5 tech companies are in control of the major distribution channels (products and use cases) for AI to reach mass adoption. They (and the companies they fund, see post below) also control the majority of compute, which is a
Is the world at risk of facing an AI dystopia? The big 5 tech companies are in control of the major distribution channels (products and use cases) for AI to reach mass adoption. They (and the companies they fund, see post below) also control the majority of compute, which is a
How the covered assets moved · since 24 Aug 2024
Prices as of 23 Oct 2026This piece lays out a thesis on the sector without making a call on the token, so price moves since publication are shown as market context.
Covered in this piece
Since then
Zero has written 30 more pieces on io.net and $IO since this one.
- io.net IDE tokenomics: #1 revenue DePIN at 5.3x MC/ARRBullish call on $IO
- DeAI as crypto's endgame: compute, agents and autonomous economiesOn io.net
- Why decentralized compute solves the startup GPU cost squeezeOn io.net
- Big compute is cracking: summary of the DeAI case against centralized cloudOn io.net
Daily closes in USD from DefiLlama and Yahoo Finance. Only a bullish or bearish thesis is read as a call; explainers, maps and mentions are context. Not financial advice.