@t_blom: Some founders don't follow the arc of progress; they bend it through sheer force of will. Assil is one of them.
Summary
A startup, based on MIT research, uses thousands of needle-like structures to create a compact steam generator, reducing nuclear power plant volume by 80% and construction time to two years. It plans to launch a 10 MW prototype in 2027 and achieve grid-scale modules by 2031, cutting electricity costs by more than half.
View Cached Full Text
Cached at: 05/30/26, 12:21 AM
Some founders don’t follow the arc of progress; they bend it through sheer force of will.
Assil is one of them.
https://t.co/s7acG8cZ7m
TL;DR
By using thousands of needle-like structures to create a compact steam generator, the nuclear plant footprint is reduced by 80%, construction timelines drop from over a decade to two years. A 10 MW prototype is planned for 2027, grid-scale modules by 2031, ultimately cutting electricity costs by more than half.
Nuclear Energy’s Challenge and Opportunity
A single nuclear fuel pellet contains millions of times more energy than an equivalent volume of coal, but extracting it safely requires power plants costing over $20 billion and taking more than a decade to build. Shrinking the plant size can dramatically cut both time and cost. In fact, reducing one of its largest components—the steam generator—achieves this without compromising power output.
Innovation from MIT
Based on my PhD research at MIT, we built the first compact steam generator using thousands of needle-like structures, occupying just 5% of the space. This enables an 80% reduction in nuclear plant footprint.
Using approved reactor designs, existing fuel, standard uranium resources, and regulatory pathways, construction time shrinks from over a decade to two years.
Our technology is proprietary, patent-pending, and we’ve built a working demonstrator that operates under the exact same conditions as today’s commercial nuclear plants. This is our reactor demonstrator, replicating the temperature, pressure, and chemical environment of current large commercial reactors.
Roadmap
- 2027: First prototype will validate our factory-built 10 MW modular technology, transportable by truck to directly power data centers.
- 2031: Four grid-scale modules can replace a power plant capable of powering a city like Seattle, ultimately cutting electricity costs by more than half.
Building a Modern Nuclear Company
We’re building a modern nuclear company by using synthetic AI workflows to reduce soft costs and leveraging advanced manufacturing for capital efficiency.
This power plant holds an important place in American history: The U.S. invented atomic energy. Yet today, other countries are building more nuclear plants faster.
Now we have the opportunity to mass-produce energy infrastructure, reshape American leadership, win the AI race, clean the grid, and unlock unprecedented human progress.
All my engines are running. Let’s build. It’s a little intense, a little WWE wrestling.
Source: @t_blom: Some founders don’t follow the arc of progress; they bend it through sheer force of will. Assil is one of them. (https://www.youtube.com/watch?v=vy8y_t1KAK4)
Similar Articles
@LiuVaayne: This is a long article by yan5xu (former ManusAI): "From Prompt to Harness: How to Understand LLM Engineering" Core Framework: The Spiral Evolution of Engineering Paradigms First Loop: Prompt Engineering (2022-2024) ...
This article by yan5xu (former ManusAI) proposes a spiral evolution model for LLM engineering paradigms: from Prompt Engineering (2022-2024) to Context Engineering (2025), then to Harness Engineering (2026-), and discusses the bottlenecks and driving factors at each stage.
@justloveabit: https://x.com/justloveabit/status/2070338139441484053
The article declares Prompt Engineering dead, proposes Loop Engineering as the new paradigm for AI development in 2026, emphasizes designing autonomous loop systems (Plan-Execute-Verify-Iterate) to let agents autonomously complete complex workflows, and provides practical examples and a getting-started approach.
@bscholl: If you’re making a new gas turbine, how big should it be? Decades ago, the computing industry figured out that large ar…
The article argues that the future of energy generation, particularly for off-grid use, lies in small modular gas turbines around 40-50MW, drawing an analogy to the computing industry's shift from mainframes to blade servers, and mentions Superpower's 42MW turbine as an example.
@GoSailGlobal: https://x.com/GoSailGlobal/status/2068243415070826738
GPU utilization in the AI industry is generally below 50%. Former a16z partner Anjney Midha founded AMP, aiming to dispatch computing power like electricity to improve utilization efficiency. The article also discusses Anthropic's success strategy, DeepMind's paper hoarding problem, and the correct approach for non-NVIDIA chips.
@oragnes: One person, zero employees, relying on an AI agent system, hard-earned $30 million in funding, valuation of $250 million, annualized revenue hitting nearly $10 million. Today, Ben Cera, the founder of Polsia, is sweeping the Silicon Valley tech scene. The $30 million funding process was largely run by AI itself.
This post reports two AI-related news: First, Polsia founder Ben Cera, with one person and zero employees, secured $30 million in funding via an AI agent system, with a valuation of $250 million and annualized revenue nearing $10 million. Second, Panthalassa has launched floating power spheres that generate electricity from ocean waves, dissipate heat using seawater, run AI computations directly at sea, and transmit results via satellite.