Imperium: Reviving Ascendancy (Ep.101)

Al and Drexy sit down with Dan, solo developer of Imperium, a spiritual successor to the 1995 cult classic Ascendancy, to discuss its EVE Online-style ship customisation, a diplomacy system built from scratch to fix the original’s biggest flaw, and how generative AI tools made a six-and-a-half-month solo passion project possible.

https://criticalmovespodcast.com/listen

This interview features Dan, a solo developer building Imperium, a spiritual successor to the 1995 4X space game Ascendancy. Dan explains how a wish to replay Ascendancy after switching to a Mac, and not knowing about DOSBox, turned into a working recreation in about six weeks and then a much larger project over the following five and a half months. The conversation covers Imperium’s modular ship building, its layered real-time and turn-based structure, its 21 alien species and four alignments, and a brand new diplomacy system. It closes with a candid discussion of AI’s role in the project and Dan’s direct response to community criticism of AI-assisted development.

Critical Moves Podcast: Episode 101 Show Notes

Episode Title: Imperium: Reviving Ascendancy with Dan (Solo Dev Interview)
Hosts: Al, Drexy
Guest: Dan, solo developer of Imperium
Episode Length: ~67 minutes

Episode Summary

This interview features Dan, solo developer of Imperium, a spiritual successor to the 1995 cult classic 4X space game Ascendancy. Dan traces the project to a simple wish to replay Ascendancy after moving to a Mac and not realising DOSBox existed. He decided instead to rebuild the game himself with AI tools despite having no professional game development background. The conversation covers Imperium’s core systems in depth, including its modular ship customisation, layered time structure, 21 alien species and an entirely new diplomacy system. It closes with an extended discussion of AI’s role in making a solo project of this scope possible, and Dan’s response to criticism of AI-assisted development.

From Wanting to Replay Ascendancy to Building Imperium

Dan explains that Imperium began because he wanted to play Ascendancy again and, on a Mac, assumed he couldn’t. Having already used AI tools for other work, he found an online resource documenting the original’s buildings, components and rules, and built a working copy in roughly six weeks. He then spent the following five and a half months refining it, summing up the experience with the project manager’s line that the first 90% of a project is nowhere near as scary as the second 90%.

Why Dan Says the Hardest Problems Disappeared

Asked about his main challenge, Dan says he struggled to name one, crediting AI tools and a few helpful members of the Explorminate community. Basing Imperium closely on Ascendancy also removed the usual risk of getting the core game design wrong, since he already had a proven game to replicate.

From Bad 3D to 2D and Back Again

Dan describes an early setback in which his 3D ship and planet models looked poor, leading him to follow a collaborator’s advice that good 2D assets beat crude 3D. About two weeks before recording, improved tools and his own growing skill let him try 3D again with far better results. He rewrote the asset pipeline, so all ships, planets and celestial bodies are now 3D, while the colony management screen stays in 2D isometric.

What Ascendancy Is for Newcomers

For listeners who never played it, Dan describes Ascendancy as a single-player 4X space game in which you pick one of 21 species, build up a home planet, research space flight, colonise nearby worlds and eventually meet rival species to cooperate with or fight. He deliberately excluded human and humanoid species from Imperium to keep the player immersed in the alien setting. The original’s appeal, he says, lies in its nuanced ship building and its atmosphere, including striking 1995-era visual effects and music, rather than in mechanical complexity on the level of Master of Orion or Stellaris.

How Far Imperium Strays from the Original

Dan puts Imperium at roughly 90 to 100% Ascendancy plus 10 to 20% of his own additions, including a fifth ship hull size, tech tree changes and a far better AI. He explains the restraint as a project management decision: stick close to something known to be balanced, finish it, then expand. Two expansions’ worth of ideas are already in the pipeline, including megastructures, ground combat and galaxy events.

Turn-Based or Real Time with Pause

Responding to the mismatch between the Steam page (turn-based) and the website (real time with pause), Dan explains there are two layered time systems. The galactic timeline runs year by year and flows continuously unless paused, much like a Paradox grand strategy game, with options to auto-resolve battles or pause on chosen events. Combat inside star systems is genuinely turn-based. Al likens it to a reverse Total War, and Dan says that if forced to choose one label, turn-based is closer.

Energy as the Limit on Combat Turns

Combat turns are not capped by a fixed counter. Each ship’s power, generated annually by its onboard generators, determines how many actions it can take. A ship with many generators and few weapons can act almost indefinitely, while a heavily armed but underpowered ship runs out quickly, so the ship design trade-offs carry directly into battle.

The Ship Designer: Five Hulls and 88 Components

Imperium offers five hull sizes, from 5 to 25 slots, and around 88 components. These include power generators, engines, weapons, shields, scanners and a large category of special modules such as self-destruct devices, long-range jump modules, tractor and repulsor beams, shield disruptors and effects that neutralise crew veteran status. Scanners are a major upgrade over Ascendancy’s largely useless originals, now governing real detection range, so enemy ships can be present in a system without being visible.

Fleet Composition and Emergent Roles

No hull carries a built-in role. Specialisation comes entirely from how players fit components, which prompts Drexy and Dan to compare it to EVE Online, where Dan spent years. Players can simply stack weapons and shields, or build tools that drain enemy power, pull fleets together for a kamikaze strike or disable specific modules. Dan says he is spending considerable effort making the AI design and use varied ships rather than defaulting to one optimal build, which Al contrasts with Stellaris combat that often reduces to a numbers game.

How a Fleet Battle Plays Out

Battles take place on a fully 3D star system map that players can orbit and zoom. Each side takes turns, and on your turn you may pick any of your ships, including the same one repeatedly, to move or use a module, limited only by that ship’s remaining energy for the year. Drexy notes the energy draining and transfer mechanics feel like a turn-based version of EVE’s capacitor warfare, with Imperium adding multi-ship, multi-system control.

The 21 Species and Four Alignments

All 21 species share the same ship components, so differentiation comes from special abilities and alignment. Examples include the crab-like Throck, whose colonies can become immune to invasion for years, and the ruthless, nomadic Jakk, who need only one invasion module to take any planet. Four alignments, warmongers, peacekeepers, defence-minded “leave me alone” species and unpredictable chaos species, shape AI behaviour in diplomacy and ship design.

Rebuilding Diplomacy from Scratch

Dan calls Ascendancy’s diplomacy atrocious, limited to occasional tech trades that were useless once the galaxy was explored. His biggest design challenge was enabling diplomacy in a game with no tangible resources that can be exchanged between empires. His answer is Imperium Points, a non-stockpilable diplomatic currency generated through embassies and interactions and split across governance, espionage and surveillance. Points can fund reputation campaigns, improve or damage relations between third parties, sabotage rival industry or research, create causes for war, smooth over border disputes and treaty violations, and influence elections for galactic emperor, including the option to reject a result and trigger galaxy-wide war. Dan credits Vay, an experienced 4X player from the Explorminate community, with helping design the depth of the system.

Trade-Offs, Roll-On Effects and a Light-Touch Option

Because planets have limited build slots, diplomacy competes with research, industry and prosperity, and Dan stresses that points cannot be stockpiled to force a single decisive move. Attacking a species without cultivating a proper cause for war damages relations with its friends too. For players who dislike micromanagement, the system can be run with a few sliders on a light-touch basis. Al observes that diplomacy is the area where Imperium has moved furthest from Ascendancy.

AI as an Essential Development Tool

Dan says plainly that without AI there would be no Imperium. Despite an Adobe Photoshop certification and past experience teaching 3D modelling and writing some OpenGL code, he could not have built a game of this scope in six and a half months alone. He describes AI as a force multiplier that rewards skill and effort rather than replacing them, comparable to a project manager briefing an expert and iterating on the result.

Answering the “AI Slop” and Theft Criticisms

Dan pushes back on the label of lazy AI slop, describing 12-hour days on top of a full-time job and hundreds of hours of Photoshop work fixing AI-generated isometric assets that never come out consistent. He also rejects the argument that AI art is theft, comparing AI training to how human artists learn by studying, imitating and eventually diverging from existing work, while acknowledging that professional artists may understandably feel threatened.

Rapid Prototyping: The Shield VFX Example

To illustrate AI’s biggest practical benefit, Dan describes prototyping a shield visual effect. He suggested enlarging the ship model slightly, applying a holographic material and sending a wave from the point of impact, and had a recorded test in about 20 minutes and a polished version in production within a day. He argues this iteration speed lets a part-time solo developer test far more ideas than would otherwise be possible, and that lazy single-prompt use is where AI slop comes from.

Why Long Solo Development Cycles Can Hurt Design

Drexy raises two solo projects that have run for five or six years while their developers learned coding, art and 3D work. Dan suggests such developers are forced to cut corners and cannot afford to test ideas, whereas cheap experimentation lets him cut features for good reasons rather than guesswork.

Responding to Players Who Reject AI Outright

Al notes that some studios use “we do not use AI” as a marketing position. Dan says he cannot reason with people who did not reason their way into that stance, and believes many objections come from people who feel personally threatened, with rationalisations about water use, the environment and corporate greed following the initial emotional reaction. His message to experts in affected fields is to use the tools to become far better at what they already do. Drexy adds that AI can also help talented people who lack the funds to hire artists and musicians.

Avoiding Publishers and Funding Pitches

Dan says a commenter told him indie publishers such as Hooded Horse would not work with him because of AI use, but that he never wanted to answer to a corporate overlord about his game. He has no pitch, did not want to sell the idea before making it, and sees AI as opening opportunities for developers without funding.

Release Roadmap

The Imperium demo had been out for about six days at recording. It is capped at 1,500 in-game years, compared with 3,000 to 5,000 for a typical full game, with a smaller galaxy and fewer species, but is otherwise fully featured. Steam NextFest runs from 19 to 26 October, with a full release (not early access) set for 2 November. Dan is finishing localisation into Chinese, Japanese and Russian, is focusing on bug fixing and AI competence rather than new features, and plans a free content expansion around January or February. Al mentions that Explorminate founder Rob has been trying Imperium and may be starting a project of his own.

Contact & Links

About | Contact | Meet the Team | Get Involved | Forum | Episodes
Discord | Reddit | Twitter / X | Facebook
Instagram | Twitch | Steam Group | Steam Curator
YouTube | Spotify | Apple | Amazon
Email: [email protected]

Episode Verdict

This interview works as a detailed feature breakdown for prospective players and as a candid case study in the AI-in-game-development debate the show tackled the previous week. Dan’s specifics, from exact module counts to the energy-based turn mechanic to the precise workflow behind one shield effect, give his claims credibility that generic enthusiasm for AI could not. His reply to the “AI slop” criticism, grounded in the manual correction work AI output still needs, offers a useful counterpoint to the more abstract ethical argument from the previous episode without dodging the genuine tension. The diplomacy segment is the strongest design discussion, addressing a complaint common across the genre with a system that sounds considered rather than bolted on, and it gives listeners a clear reason to try the demo before NextFest and the November launch.