Pinned Tweet
Ok, here it is. This is game-changing, and what we are about to build and release on top of the @gnusai Operating System.
drive.google.com/file/d/1iuJ…
Ambitious? Maybe... But with our OS already built, IMHO, a 3-6 week project. With embedded proven blockchain technology.
Humanity, you are welcome.
@elonmusk, @PeterDiamandis "A day before something is a breakthrough, It's a crazy idea", @JeffPeoples, @beffjezos, @XFreeze, @tetsuoai, @BetterCallMedhi, @EricJorgenson, @gailalfaratx, @cboyack, @MarioNawfal, @Natortelli, @SawyerMerritt,@michaelnicollsx, @MrBeast, @Cobratate, @SawyerMerritt, @grok, @rohanpaul_ai, @Chad_Hurley, @stevechen, @sama, @brycent, @Ronald_vanLoon
So true, why we are building something unique.
The GNUS Cognitive App is a system that continuously models, challenges, and improves an individual’s thinking.
Built on top of @GnusAi a distributed, verifiable compute network that executes and coordinates arbitrary workloads—including AI, rendering, and cognitive processes—while continuously optimizing cost, performance, and correctness.
I’ve been saying: "Training is where all the money is spent, and inference is where all the money is made."
That gap is widening. Pre-training is hitting diminishing returns. Labs are locking in massive fixed compute commitments (data centers, power, chips) just to stay in the race. Meanwhile, inference is the permanent load and the revenue engine.
Open-weight models + distributed serving change the math. You don’t need to own the capex to compete where the money is actually made. @GnusAi
Fair warning to potential Anthropic investors:
Open-source/open-weight models are only part of the risk. The bigger one is large-scale compute — especially once inference moves to distributed or decentralized systems.
Anthropic’s own S-1 says competition “could increasingly depend on access to large-scale compute rather than on model development capability.”
Training is the capex sink (data centers, power, chips). Inference is where the money is made. Systems that don’t need to own the compute change the equation.
sec.gov/Archives/edgar/data/…
@GnusAi
Fair warning to potential Anthropic investors:
Open-source/open-weight models are only part of the risk. The bigger one is large-scale compute — especially once inference moves to distributed or decentralized systems.
Anthropic’s own S-1 says competition “could increasingly depend on access to large-scale compute rather than on model development capability.”
Training is the capex sink (data centers, power, chips). Inference is where the money is made. Systems that don’t need to own the compute change the equation.
sec.gov/Archives/edgar/data/…
@GnusAi
Kenneth Hurley retweeted
spot on. kennethhurley.com/blog/elect…
Kenneth Hurley retweeted
“Probably with smaller, local, specialised models.” gcs.gnus.ai
My AI predictions.
In the next 1 to 3 years, Open AI or Anthropic, or both, fail / get acquired for a fraction of their current valuations.
Microsoft, Google or a developer tooling company with access to lots of developers and source code like Sonar or JetBrains wins AI in software engineering.
Probably with smaller, local, specialised models.
Other domain specific companies spring up to win domains such as mathematics, legal, writing, etc.
TL/DR: The technology isn’t going away, but the financial bubble will burst.
What do you think?
"I think what you have is a market deficiency. Especially to the trained eye and trained operator."
That's what an operator/investor who has personally funded decentralized data centers said about @GnusAi on a call @boardyai set up for me this week.
Where Genius AI BOSS stands:
About 80% complete after roughly four weeks of building. 27 plugins, including point of sale, reservations and HR. $20 per device per month, on-premise. Built for small and medium businesses that are tired of paying for Toast and QuickBooks.
One restaurant is our first product partner. Five more customers are waiting, all from word of mouth.
2-3 more weeks to finish the MVP.
On the 👃IMHO
“5-10 years out, maybe longer for complex systems. The role is shifting to directing agents and judgment. Level up there.”
Honest take: AI already handles most routine coding and many supervised end-to-end tasks. Fully autonomous replacement of a solid software engineer—architecture, novel design, accountability, production ownership—looks 5-10 years out, maybe longer for complex systems. The role is shifting to directing agents and judgment. Level up there.
Very true, building some that I've wanted to build for years!
Kenneth Hurley retweeted
Why are there no predictions about distributed/decentralized inference, the catalyst for some of these?
It begins, bad AI investments and more to come.
techcrunch.com/2026/10/13/ju…
Kenneth Hurley retweeted
For narrow, high-volume, proprietary-data tasks where accuracy and latency matter more than generality, the Coinbase-style approach can still win on total economics.
Kenneth Hurley retweeted
Exactly. We built the decentralized infrastructure. Now we're building GCS at @gnusai.
Companies should own their AI stack, data and specialized models. Fine-tune once, then adapt with ELMs, private memory, routing, verification and EGGROLL retraining.
gcs.gnus.ai/
Kenneth Hurley retweeted
We’re building the platform that lets companies create and maintain private specialist models on their own data, without having to build and operate the AI infrastructure themselves. @GnusAi
Kenneth Hurley retweeted
Chris, we’re building exactly this with GCS: specialized Expert Language Models (ELMs) on companies’ own data. We go further with EGGROLL-style retraining from real outcomes, plus memory and verification. Local or distributed compute. Let’s talk.
gcs.gnus.ai/