Medieval Engineers

Lore & Background

Medieval Engineers is a follow-up to Keen Software House's earlier sandbox game Space Engineers, both built on the VRAGE 2 engine. While Space Engineers is set in zero gravity, Medieval Engineers uses gravity extensively on its Earth-like planetoid. The game's development benefited from technological cross-pollination between the two titles, with features such as compound blocks, spheroidal landscape, mechanical blocks, and structural integrity being tested in Medieval Engineers before being applied to Space Engineers.

The game was actively developed under the guidance of Tim "Deepflame" Toxopeus, with weekly updates and YouTube videos until February 2019. After the official release of Space Engineers and Medieval Engineers version 0.7.1 in February 2019, the Medieval Engineers development team was folded into the Space Engineers team with no future planned updates. However, the game received a small update to version 0.7.2 upon its official release out of Early Access in March 2020.

Post-release, Keen Software House agreed to share ongoing maintenance with two prominent modding community developers, Equinox and Gwindalmir. On April 7, 2022, the Community Edition was released as version 0.7.3, an alternate opt-in branch aggregating many bug fixes. This edition continues to see ongoing maintenance with its own website, Discord channel, and community bug tracker, occasionally supported by Keen Software House developers.

Reader's Guide

Medieval Engineers occupies a unique niche in the sandbox game landscape of the 2010s, combining voxel-based terrain deformation with realistic physics for siege warfare. Its key innovation lies in the small block system—items like torsion bars, rope drums, axles, hand cranks, and projectiles—that enables players to construct functional battering rams, catapults, trebuchets, and siege towers governed by mass, density, tension, and inertia. This mechanical depth, paired with structural integrity mechanics for large blocks, creates a gameplay loop where players must balance architectural ambition against the risk of collapse.

The game's survival mode adds further complexity through three player stats—Stamina, Health, and Food—along with a research system that gates item creation behind collecting specific resources. The voxel planet, fully explorable at 10 game-kilometers radius, features varied biomes including mountains, valleys, woodlands, deserts, and grasslands, each with distinct resources. Multiplayer allows territory claiming, house creation, and castle defense against both hostile NPC barbarians and other players.

The Community Edition's continued maintenance after official development ceased demonstrates the game's lasting appeal among dedicated players, with the modding community effectively preserving and improving the experience beyond its original support window.

Did You Know?

• The game's planetoid has no water and is approximately 10 game-kilometers in radius.
• Siege engines like catapults and trebuchets use realistic physics based on mass, density, tension, and inertia.
• The Community Edition (version 0.7.3) was released on April 7, 2022 as an opt-in branch created by modding community developers.
• Players can fast-travel using a network of dirt roads found across the planet.

The Long Arc: From Antiquity to the Present

The timeline of artificial intelligence—sometimes referred to as synthetic intelligence—spans an extraordinary range of human history, beginning in the Antiquity, Classical, and Medieval eras before moving through the 1600-to-1900 period. The twentieth century receives the most granular treatment, subdivided into the 1901–1950 stretch, followed by individual decades from the 1950s through the 1990s. The twenty-first century continues this decade-by-decade approach, covering the 2000s, 2010s, and 2020s. This structure suggests that while the conceptual roots of machine intelligence stretch back millennia, the field's most rapid and visible development has been compressed into roughly the last seventy years. The timeline also points readers toward companion chronologies for machine translation, machine learning, and the risks artificial intelligence poses to global finance, acknowledging that AI is not a single thread but a family of intertwined developments. The very existence of a dedicated timeline signals that the subject has grown large enough to warrant its own historical narrative, separate from general computing history.

Pioneers Who Shaped the Field

A constellation of researchers laid the groundwork for modern artificial intelligence. Alan Turing's 1936–1937 work on computable numbers established the theoretical limits of what machines could process. In 1943, McCulloch and Pitts proposed a logical calculus modeling the ideas embedded in nervous activity, bridging biology and computation. The 1955 Dartmouth Summer Research Project proposal, authored by McCarthy, Minsky, Rochester, and Shannon, is widely recognized as the founding document of AI as a formal discipline. Arthur Samuel's 1959 study of machine learning through checkers demonstrated that programs could improve through experience. Meanwhile, Newell and Simon contributed heuristic problem-solving methods in 1958 and their General Problem Solver in 1963, framing computation as a simulation of human thought. Minsky's prolific output—spanning finite and infinite machines, perceptrons, knowledge representation, and the Society of Mind—touched nearly every subfield, making him a connective thread across decades of research.

Philosophical Challenges and Theoretical Boundaries

The history of artificial intelligence is inseparable from deep philosophical questioning. John Lucas invoked Gödel's incompleteness theorems in 1961 to argue that minds transcend any formal system. Hubert Dreyfus's 1972 book What Computers Can't Do challenged the assumption that symbolic manipulation could replicate human cognition. John Searle's 1980 Chinese Room argument, published in Behavioral and Brain Sciences, questioned whether syntactic processing could ever yield genuine understanding. Sir James Lighthill's 1973 survey for the Science Research Council cast a skeptical light on the field's progress. McCarthy and Hayes, in their 1969 paper, laid out philosophical problems that AI must confront from its standpoint. Even earlier, Hobbes's Leviathan (1651) and Hofstadter's Gödel, Escher, Bach (1980) explored the relationship between mind, language, and formal systems. These works remind us that AI has always been as much a philosophical enterprise as a technical one.

Institutional Memory and the Ongoing Canon

The academic infrastructure surrounding artificial intelligence reveals a field that has matured into a permanent scholarly discipline. Stuart Russell and Peter Norvig's textbook Artificial Intelligence: A Modern Approach, now in its fourth edition (2021), has served as a standard reference since its second edition in 2003. The Stanford Encyclopedia of Philosophy maintains an entry on Principia Mathematica, edited by Linsky and Irvine in 2022, linking AI's logical foundations to broader philosophy. The U.S. National Research Council's 1999 report on funding computing research documented government support for AI development. More recent works—Brian Christian's 2020 book on the alignment problem, Jürgen Schmidhuber's 2022 annotated history of deep learning, and Matteo Wong's 2023 Atlantic piece on ChatGPT's rapid obsolescence—show that the conversation continues to evolve. Pamela McCorduck's Machines Who Think (2004) and Daniel Crevier's 1993 history of AI's tumultuous search anchor the narrative in both technical and human terms.

Gallery