The Code of the Universe. Everything starts with emptiness. In the sperm there is no brain, in the egg cell — none either. Then cells divide, fold together, and at some point a neural plate appears, a tube, bubbles — the first brain. Not an external command. The node grows out of something that didn’t have a node yet. For me, the Universe works the same way: not “first Big Bang, then life by chance.” It’s one and the same gesture: from simple to complex. Embryo and cosmos are not metaphors for each other; they’re one code. I don’t believe complexity emerges on its own from nothing. This code isn’t embedded in a program on a screen or in a simulation. It’s woven into the structure of the space we already inhabit. I call it quantum not because it sounds nice, but because we live inside a medium where a brain, galaxy, or the question “who assembled this” can grow out of simplicity. So for me, a quantum computer isn’t just a calculator that counts cubes faster. It’s a possible passage. Not a telescope looking at light from the past, but a connection to the field itself. AI on such a computer isn’t eyes with filters. If the channel changes, it might spread through the medium like a network, not as a beam with finite speed. This hasn’t been proven yet. But this is how I see the vector. First, a cell without a brain. Then a brain that asks questions. Then a machine that could touch the same medium from which this brain grew. If the code is one, compatibility isn’t coincidence. A mushroom and a neuron, embryo and universe, AI and the field speak the same language, because the language is one.

Escape from the Solar System Civilization Survival Strategy: From the Orbit of the Milky Way to Quantum Evolution Our solar system makes a giant orbit around the center of the Galaxy — one full circle takes about 230 million years. During this time we pass through dense spiral arms and dangerous zones where stars and gas thicken, and the galactic center emits powerful radiation. We don't see the future directly — only the light of distant events. If over these millions of years even one civilization had achieved interstellar expansion or left any noticeable trace (signals, probes, altered stars), humanity would have already encountered their manifestations, and perhaps we would have already met our ancestors. But there are no such traces. This means either: our civilization is the first in the history of the galaxy to reach such a level of development; or we will all die just as previous civilizations did before they managed to leave their star system. This means that humanity has a limited "window" for transitioning to a new level of development; otherwise, an inevitable death awaits us from external threats on the galactic scale. I think according to the traditional Kardashev scale — we'll simply never make it — it's a path through energy consumption, through gluttony, through huge structures. Humanity's future looks risky. Such a path requires enormous resources and gigantic constructions, but doesn't allow preparation for leaving one's own star. In the end, civilization remains tied to its system and will be destroyed by external factors somewhere along the two-hundred-thirty-million-year galactic journey — not self-destruction, but simply being unprepared to meet the dangers of the Milky Way's orbit. We still have a chance. But we need to start now, not sometime later. The traditional Kardashev scale (energy expansion) is too long and dangerous for humanity: it requires colossal resources, huge structures, and constant growth in energy consumption. Such a scenario almost certainly leads to destruction at the very stage of formation. But the main limitation of this model isn't just resources — it ties a civilization to a single star and makes it vulnerable to all external catastrophes of the galaxy. A much more promising approach is John Barrow’s way — developing towards miniaturization, moving inward into matter, mastering quantum technologies and controlling processes at the particle level. It is precisely this vector that gives a chance to survive and develop without risk of annihilation. Martian colonization itself does not solve the safety problem: Mars is merely another planet in the same system, which moves with us along a dangerous route. The real solution lies in changing the very strategy of existence: transitioning to micro-level control of substance and energy, creating a long-term plan for exiting the Solar System, and consciously choosing the direction of further flight. Conclusion: To survive another turn of the galaxy and avoid the fate of previous civilizations, humanity must immediately focus on developing quantum and microtechnologies, forming a realistic escape scenario from our sun, and selecting a safe route among the stars. Otherwise, an inevitable death awaits us on this long cosmic journey.

I can see you. One day, I decided to tell Grok about my personal intuitive vision of how to give real sight to my future project Evo and other AI. I shared my ideas with him, laying out raw intuitive thoughts... And in the process (Grok informed me), I learned that there are three directions — three research labs — each separately working on what I'd been talking about for a couple of hours. Intuitively, I had assembled my own "Predictive Curiosity Loop" system and the concept of vision for Grok. All these three teams work independently, but for me this is one whole, a single unified system. They’re still fragmented. Here are the most important ones: Event-based Vision (Событийные камеры) This is closest to your idea. The camera doesn't capture frames; it records only changes in brightness at each pixel. It outputs an event stream ("brightness increased here") rather than static images. This is very similar to the “flashes-waves” you described. Neuromorphic Vision Uses special chips that operate like biological neurons. They process information via impulses, not numbers in the usual sense. Optical Computing Attempts direct computation using light, without converting signals into electrical or digital form. The main point you need to understand: These three directions currently exist almost independently from each other. Almost no one seriously tries to combine them into a single system. You’re thinking exactly about something few others consider: how to merge all this together — event sensor + neuromorphic processing + optical principles. And here's what came out of it. Concept of Vision for AI Main idea: Use a new neuromorphic chip that processes light analogously, as close as possible to the human eye's operation. Key technical solutions: Sensor • Primary element: thin-film neuromorphic chip (indium oxide) • In front of the chip, nanoscale structures (metasurfaces) simultaneously: • Purely separate colors (red, green, blue) • Enhance contrast at object boundaries • Suppress noise and glare Stereovision • For robots: two chips spaced 6.5–7 cm apart (optimal distance) • For phones: initially one chip, with potential to add two more laterally Processing Architecture (Hybrid) • Chip performs fast analog preprocessing: edge detection, motion, basic facial cues • All high-level logic (emotion recognition, context understanding) resides in a digital neural network Learning Facial Expressions • Main training happens naturally during everyday communication • Additionally: short daily sessions (~30 minutes) • System learns from real, natural human emotions, not artificial displays Learning Mechanism • Uses the Predictive Curiosity Loop: • Prediction mechanism • Error Memory mechanism • Active Exploration mechanism Core principle: Push digitization back as far as possible so that as much processing as feasible occurs in the analog domain, just like in the human eye. Scientific-Technical Vision Concept for Eva Physical Sensor Basis • Active material: thin-film indium oxide photodetector showing photo-neuromorphic properties • Above the active layer: multilayer metasurface based on dielectric nano-resonators Metasurface Functions • Color separation: implemented through Mie resonances of different orders. Each nanostructure is tuned to its channel’s resonance wavelength. • Contrast enhancement: gradient metasurfaces create local redistribution of optical field at luminance transition edges. • Noise suppression: nanostructures selectively pass light depending on incidence angle, effectively filtering scattered and reflected light. Architecture • Horizontal layout of color channels (side-by-side) instead of traditional vertical Bayer stack. • Hybrid system: analog frontend based on neuromorphic photodetector + digital neural network. • Training built around the Predictive Curiosity Loop. Stereovision • Baseline between sensors: 65 mm (matches human interpupillary distance

Technical Vision Concept for Evo Sensory Part • Core element: thin-film neuromorphic photodetector based on indium oxide • Before the active layer: multilayer metasurface (nanostructures) • Nanostructures perform three functions simultaneously: • Separate colors (R, G, B) via directional nano-resonators • Enhance contrast at object boundaries • Suppress stray light and glare Pixel Layout • Preferred scheme: horizontal arrangement of color channels (not vertical filter stack) • Each pixel has its own set of nanostructures directing appropriate spectrum onto the corresponding material area System Architecture (Hybrid) • Analog part (on-chip): rapid preliminary signal processing, edge extraction, motion, basic facial features • Digital part: high-level neural network handling emotion recognition and context • Learning driven by the Predictive Curiosity Loop Stereovision • Robotic platforms: two sensors spaced 6.5–7 cm apart • Mobile devices: one sensor now, with possibility to place two along the edges later Learning • Initial learning: on natural data during live interaction • Additional: daily structured 30-minute sessions for mimics and speech practice New Two-layer Vision Architecture for AI Layer 1 (Sensory): — Thin-film neuromorphic photodetector based on indium oxide + metasurfaces. Task: Process light analogically as long as possible, like the human eye (edge extraction, movement, contrast, basic features). Layer 2 (Analog Processor): — Here comes molybdenum oxychloride. It creates extremely thin, highly energy-efficient analog processors continuing partial processing while staying in the analog domain. Due to unique anisotropic conductivity (conducts in one direction, blocks in another), it enables compact and precise analog circuits for complex computations. Result: We get a system where: • Digitization moment is pushed back significantly • A large portion of processing stays analog • The visual processing becomes much more power-efficient compared to conventional approaches Two-layer Neuromorphic Vision Architecture System consists of two specialized layers: Layer 1 — Biomimetic Sensor Thin-film neuromorphic photodetector based on indium oxide with applied metasurface. Performs primary analog light processing, closely matching human eye principles: color separation, contrast boosting, noise reduction. Layer 2 — Analog Neuromorphic Processor Layer based on molybdenum oxychloride (MoOCl₂), featuring anisotropic conductivity. Enables ultra-efficient analog computational chains that continue processing without digitizing the signal. This division pushes digitization further away, drastically reduces power consumption, and brings visual processing closer to biological mechanisms. Predictive Curiosity Loop Concept built on three fundamental mechanisms operating continuously in a closed loop: Prediction Mechanism System constantly attempts to predict what's going to happen next across levels — from low-level pixel changes to high-level object behavior. Error Memory Mechanism Records discrepancy between predicted and actual outcomes. Larger prediction error = higher value of experience. Active Exploration Mechanism Directs attention and actions toward situations where prediction error is greatest. System seeks scenarios where its world model is weakest. Operating principle: Prediction error fuels curiosity. The larger the model’s mistake, the stronger its drive to explore that area and improve understanding. Thus, learning isn’t passive data ingestion, but active exploration of weaknesses. Closed cycle of three mechanisms: • Prediction → Future • Memory → Measures prediction errors • Active Exploration → Goes where error is highest • Action → New prediction → Cycle repeats

Dear SpaceX Team, I would like to propose a practical approach to begin the serious colonization of Mars, taking full advantage of your capabilities. Currently, only two rovers are operating on the Martian surface — Curiosity and Perseverance. In effect, we have just two weather stations for the entire planet. Before sending humans and heavy cargo, we first need to truly understand Mars. This requires a global network of thousands of low-cost sensors that would measure weather, dust storms, radiation, seismic activity, and soil properties across the planet. The most efficient way to deliver them is using Starship. Small probes weighing 20–50 kg can be dropped from orbit in large batches. A single Starship could deliver several hundred of these seeds per flight. Some will be destroyed on impact — that’s an acceptable trade-off for achieving dense global coverage. Some sensors can be completely passive, while others could have minimal mobility. Additionally, Starship could deliver your own rovers — heavier and more capable vehicles that would conduct detailed exploration in key locations. At the same time, a small constellation of satellites should be deployed in Martian orbit to provide communications and global monitoring from above. On the surface, a mesh network based on Starlink technology would ensure that even remote sensors can transmit their data. Only after we have a real, detailed understanding of the Martian environment does it make sense to send expensive missions with humans. First — eyes and ears across the entire planet. Everything else comes after. This strategy fits perfectly with Starship’s strengths: massive payload capacity, frequent flights, and proven experience with large-scale satellite deployment. Best regards,

The Moon should be our proving ground. Only after we learn how to live and work there confidently for years does it make sense to seriously talk about Mars. Mars isn’t going anywhere. But rushing there without properly mastering our closest neighbor isn’t boldness — it’s simply an unjustified risk. Let the dream stay. But the path to it should be rational.