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.
@unclebobmartin Good point. But you missed one tiny piece of the formula: decentralization of AI subtracts real estate and data center variables, so it is logarithmic.
You're crazy until you succeed—then you're a super genius.
@elonmusk has lived this exact arc for decades. People called reusable rockets and mass-market EVs impossible pipe dreams. They laughed, dismissed, or attacked the ideas as delusional.
Now those same people line up for Cybertrucks and Starlink and to invest in SpaceX.
The pattern repeats for any big, unconventional breakthrough: skepticism is the default reaction to anything that challenges the status quo. @GnusAi
@magattew’s argument is solid, evidence-based, and directly addresses the “colonialism explains African poverty” narrative.
It’s not cherry-picking—it’s using the most relevant counterexamples possible. @fayedfa's reply calling it cherry-picking is the weaker take; it dodges the logic by appealing to history without explaining why the pattern doesn’t match the claimed cause.
Blaming colonialism indefinitely lets poor governance off the hook.
Prosperity comes from better institutions, markets, education, and rule of law—things countries control now, as Vietnam (and others) have shown.
@victormustar So pumped about the direction of all this, as it really is what @elonmusk envisioned for truly democratized AI and, IMHO, will end up crushing OpenAI.
@victormustar So pumped about the direction of all this, as it really is what @elonmusk envisioned for truly democratized AI and, IMHO, will end up crushing OpenAI.
I’d probably add as well. Most software companies and app developers aren’t looking at it that way maybe even the frontier AI companies. We are now at a clean spot where we understand the different roles and functions that an LLM can play and while we keep going for AGI, we can probably get better and predictable execution by focussing on the specific roles and what an LLM can do.
Maybe it’s time for an LLM to be broken up and into a functional map based upon capabilities and outcomes?
Love your stuff by the way, it really gets everybody thinking
Your margin is my opportunity: AI version…
The biggest surprise of 2026 is that the capability gap between the best open-weight/source models and the best closed models has narrowed much faster than the pricing gap. The pricing gap remains enormous while the capability gap is quite narrow.
What does this means in practice?
For a company consuming 1 billion input tokens and 1 billion output tokens per month:
GPT-5.5 Pro: ~$105,000
Claude Opus 4.8: ~$30,000
DeepSeek V4 Pro: ~$5,220
DeepSeek R1: ~$2,740
I asked ChatGPT what it thought about this and it answered as follows:
“If I were building a company today, the economic frontier would look roughly like:
DeepSeek V4 Pro / R1 for high-volume inference.
Claude Opus for premium agent workflows where reliability matters.
GPT-5.5 Pro only for workloads where its incremental capability demonstrably produces enough business value to justify a 20–40× token premium.”
Most CEOs have no idea that, instead of this nuanced approach, their teams are running amok internally by picking the most expensive models in most cases and burning through massive budgets with zero governance, audit ability and control.
As control planes like our Software Factory become more standard, you can expect the run rate revenue growth of the frontier labs to go down meaningfully and the revenues of the open models to skyrocket.
Why? Because we can implement the nuanced approach above and be agnostic to model - instead focusing on customer intent, model task and cost management among other things.
The market is about to discover that demand for frontier AI is real, but its margins may not be durable. As model pricing compresses from above, https://t.co/hewStVK30N attacks the compute-cost layer from below.
Now imagine those 10 employees aren't replaced, but actually become 10 consultants that can be 10x more productive and can be paid 10x more and expand to other businesses. I for one would work less hours for 2x-10x pay
Imagine replacing 10 employees with 2 short-term consultants who not only comprehend your business processes, finances, and technology but can also accomplish the same tasks 10 times faster. @vasuman you should enhance your strike teams...
NOPAT is net operating profit after tax: operating income × (1-tax rate). It measures core business profitability after taxes but before financing costs, making companies comparable regardless of debt levels.
CC is the cost of capital charge, typically WACC multiplied by invested capital—the minimum return investors require.
SVA (Shareholder Value Added) = NOPAT minus that capital charge. Positive SVA means the company created value above its cost of capital.
EV is enterprise value: equity market cap + net debt. It reflects the total value of the business operations.
NOPAT beats EBITDA because EBITDA ignores depreciation/amortization (real asset costs), interest, taxes, and capex needs. Munger called EBITDA “bullshit earnings” for painting a misleading picture of sustainable profits. NOPAT gives a cleaner, after-tax operating view for valuation and economic analysis.
@grok can you explain NOPAT, CC, SVA and EV? Net operating Profit After Taxes, Cost of Capital, Shareholder Value Assurance, Enterprise Value and why NOPAT is better that EBITDA, which i think Charlie Munger has made statements on EBITDA is crap
Imagine replacing 10 employees with 2 short-term consultants who not only comprehend your business processes, finances, and technology but can also accomplish the same tasks 10 times faster. @vasuman you should enhance your strike teams...
Imagine an AI company that forward deploys into your enterprise to first understand how everything works, then architects what an agent solution looks like custom built for you, and only then builds the agents.
Someone should uhh… someone should make a company that does that.