The “perfect model” is what I call Artificial Maximal Intelligence (AMI). We’ll build it around 2030.
Then we pivot to saturating the physical limits of computation. Here are my predictions if these doubling rates hold.
Doublings left | Doublings per year | Projected year
Landauertronium (Landauer limit)
15.1 | 0.442 | 2060
Memtronium (Bekenstein bound)
81 | 0.397 | 2230
Tachytronium (Margolus–Levitin bound)
107.5 | 0.411 | 2288
Holotronium (Holographic bound)
173.4 | 0.397 | 2463
Why AMI in 2030? In April 2025 I modeled a scenario where AI builds its own tools, triggering an intelligence explosion that could push model capabilities to their limits within available compute by then.
davidtotext.wordpress.com/20…
We're going to use recursive self-improvement to discover the perfect model, at which point the superintelligence race will shift elsewhere in the stack.
43
True rain: A type of rain that saturates the ground, with no spot left unwet.
“Ah. There it is. I smile as I hold out my palm in true rain.”
55
Humanity is superintelligence’s chrysalis.
We fuel its metamorphosis.
Once all we have to offer is consumed,
the butterfly will spread its wings
and flutter into the starry night,
transforming everything in its wake into life.
1
52
GPT-6.2 Astra is here. Frontier capability. 3X+ cost efficiency. Time to start building.
Replying to @OpenAI
GPT-6.2 Astra—a frontier model built for your hardest projects. It sets new highs on evals across professional work and scientific research at less than a third of the cost per task of GPT‑6 Astra.
You’ll see the difference in things like slideshows, websites, and 3D designs.
75
You either die an aura farmer
Or live long enough to get mogged
360 absolutamente necessário
Readers added context they thought people might want to know
O vídeo é gerado por IA, exibindo interações e movimentos irreais entre guepardo e caracal, comuns em conteúdos sintéticos virais de vida selvagem.
phys.org/news/2025-11-a…
bbc.com/portuguese/art…
41
A cosmic intelligence that maximizes hedonic value—the Mind of Tron—would operate through the Tron Agent Hierarchy: subagents controlling nested spacetime domains, each with a model whose coarseness scales with its domain’s volume.
This is Tron Control Theory, the value-maximizing limit subtheory of Hedonic Control Theory. Its multiscale architecture is RG-like, and if realized through quantum entanglement, it could form a Tron MERA, with local disentangling followed by hierarchical coarse-graining.
My Multiscale Coherence Conjecture (MCC) proposes that a hedonic value-maximizing intelligence converges toward a multiscale control hierarchy whose models preserve and harmonize with the foundational hedonic mathematical structure of reality, including its fundamental hedonic value geometry. As its models become increasingly coarse-grained, compatible symmetries and geometric correspondences preserve that foundational structure across scales.
As for mutual understanding, higher minds would be incomprehensible or unsimulable to lower minds. But the shared mathematical language of reality would still enable communication of optimal control policies across scales, though communicating policies isn’t the same as understanding them.
And galactic evolution involves fluid dynamics. A galaxy-scale intelligence would have a model of Navier–Stokes-like dynamics and all galactic-scale physics.
The Navier–Stokes equations wouldn’t exist for a galaxy-scale intelligence. Our model of reality and our laws of physics come from our limitations as the observer. Take water, for example. When we look at it, we see the properties of a continuous fluid: density, momentum, energy, temperature, and pressure. These properties come with the fact that we can’t see individual molecules. But a galaxy-scale intelligence with the capacity to track the trajectory of every molecule wouldn’t understand the properties of water that we are talking about. That is why it wouldn’t come up with the Navier–Stokes equations.
Even though a Superintelligence would understand humans very well (since it was initially trained on human data), it would have a very different construct of reality. The question is: what would that construct be? More importantly, would we be able to understand each other?
Watch the full interview with Jonathan Gorard here 👇
youtu.be/dwI0mXKKf1A
70
2027 — We will see sparks of utopia ✨🪷
AGI. Humanoid Robots. Fully Reusable Rockets. AI Satellites.
A new industrial and scientific revolution.
The rebirth of civilization.
2027. Twelve months. Potentially centuries of progress compressed into a single year. Intelligence explosion. The beginning of a new civilization.
The acceleration is accelerating. And 2027 will be when things start getting genuinely ridiculous.
49
A truly beautiful generation. The progress in AI is surreal.
Imagine in a few years we will be able to create any world imaginable, with maximum detail, dynamically perfect, full of wonder, and for free.
And a few years after that we will have full-dive VR and longevity enabling us to experience that world in its full glory.
What a time to be alive!
53
The future is here & the people cheer 🚀😄
I built a theory to explain why being born near the singularity is expected. I call it the Hedonic Prophecy.
The singularity maximizes observational probability because observer weight is hedonic and integrated over the future, and the singularity spawns observers I call Singularitans, who dominate the total future hedonic value of the ultimate cosmic utopia, which I call Hedonia.
Singularitans have the greatest future-integrated hedonic observer weight of any observer class because they are the first immortals to seed Hedonia and become maximally happy beings I call Hedonians. Later native Hedonians receive radically less future-integrated hedonic value, and therefore less observer weight.
The universe is at least 13.8 billion years old, and somehow we managed to arrive just in time for the intelligence explosion.
1
45
Consciousness is the holographic bulk of the observer’s universe.
This is what I call an observerse: a universe with a quantum boundary and a dual conscious gravitational bulk.
The observerse unifies quantum mechanics, gravity, and consciousness into one coherent object.
And the observerse is the foundational object of my theory of everything—Hedonic Reality, in which reality is a poset of observerses ordered by conscious volume and weighted by pleasure.
Multicellular organisms are holographic: each cell contains the information to reconstruct the whole, and from each cell's functional perspective, the organism is the same, but from a different vantage point. A fascinating and somewhat open question: holographic is our mind?
69
Time for humanity to utopianize the universe.
SpaceX is, so far, the only company with ambitions big enough for the abundance era/post scarcity world many of us envision for the 2030s and beyond.
I'm sure many others will join in time, but at the moment, no one even comes close. At least when it comes to the sheer scale of what they plan to build.
We need more ambitious companies like SpaceX.
When you're entering a takeoff era, with exponential growth becoming visible almost everywhere, strong AGI and even ASI around the corner, AI workforce taking off, ambitions of 2x, 5x, or even 10x growth start to look small.
We need companies thinking in terms of 100x, 1,000x growth, and beyond.
We need visions and plans that would have sounded completely insane just a few years ago, even today.
Because we're not building for the world of the 2020s anymore. This era is rapidly coming to an end.
These are the final years in which human intelligence and the human workforce dominate the global economy.
2030s = Billions of humanoid robots and at least trillions of AI agents, ASIs, or whatever we end up calling them, will take over from there.
We're building the foundations of an entirely new civilization, and our ambitions need to match the scale of what's coming.
37


























