Tech Heresy: Using AI to make Army Lists

Forum adverts like this one are shown to any user who is not logged in. Join us by filling out a tiny 3 field form and you will get your own, free, dakka user account which gives a good range of benefits to you:

No adverts like this in the forums anymore.
Times and dates in your local timezone.
Full tracking of what you have read so you can skip to your first unread post, easily see what has changed since you last logged in, and easily see what is new at a glance.
Email notifications for threads you want to watch closely.
Being a part of the oldest wargaming community on the net.

If you are already a member then feel free to login now.

**Lathe Biosas**

Has anyone attempted to let AI create their GW army lists for them?

I asked the Google AI to make me a viable 10th Edition Adeptus Custodes army list at 2,000 points.... and this is what it gave me:

Based on current 10th edition tactics and units, here is a 2,000-point Adeptus Custodes army design using the popular Shield Host detachment.

This list focuses on a durable infantry core supported by elite deep strikers and powerful allied support.

Detachment: Shield Host

Characters

Trajann Valoris (160 pts): As your Warlord, Trajann is an essential choice. His leadership abilities and superior melee prowess make him a force multiplier and a dominant combatant. He is best attached to a large squad of Wardens.

Blade Champion (145 pts): Armed with Vaultswords, the Blade Champion is a potent and fast melee threat. He is an ideal leader for a Warden squad, providing enhanced combat effectiveness and mobility. He is an excellent candidate for the Veiled Blade enhancement to increase his damage output.

Shield-Captain (170 pts): This adaptable character provides additional combat punch and aura support. Give him the Ceaseless Hunter enhancement to let a unit rapidly redeploy, offering tactical flexibility.

Battleline

5x Custodian Guard (225 pts): A solid, durable unit for holding critical objectives. They are excellent when paired with the Shield-Captain for increased resilience and combat options.

5x Custodian Guard (225 pts): A second unit of Guards provides more objective control and a sturdy front line for soaking up damage.

Other Datasheets

5x Custodian Wardens (250 pts): The core of your army's melee punch and durability. Attach Trajann to this unit for a nearly unkillable centerpiece that can destroy enemy units and control the board.

5x Custodian Wardens (250 pts): A second squad of Wardens, led by the Blade Champion, provides another highly dangerous and mobile combat unit that can counter-attack or push aggressively.

2x Allarus Custodians (130 pts): This small, fast, and hard-hitting unit is perfect for deep-striking to apply pressure, eliminate key targets, or seize undefended objectives.

2x Allarus Custodians (130 pts): A second squad offers more deep strike flexibility, allowing you to pressure multiple parts of the board simultaneously.

Allied Units

Callidus Assassin (100 pts): This Officio Assassinorum operative provides valuable tactical depth. Her ability to mess with your opponent's stratagem costs and redeploy makes her a useful utility piece.

Strategic breakdown

Core Battle Plan: The army is built around two powerful, character-led Warden bricks that will dominate the midfield. Trajann's unit is the main anvil, a nearly unbreakable wall that can absorb immense punishment, while the Blade Champion's unit is the hammer, a faster and more aggressive unit for flank attacks.

Flexible Deployment: The Allarus Custodians and Callidus Assassin offer excellent deep strike and harassment options, keeping your opponent on their toes and allowing you to respond to threats across the board.

Objective Control: The two Custodian Guard units are your foundation. They are tough enough to hold objectives, even under fire, freeing up your more mobile units to deal with specific threats.

Winning the Attrition War: Adeptus Custodes can no longer simply stand on objectives and win. This list is designed to engage the enemy with its durable core, remove key targets with deep strikers, and win the game through a war of attrition.

I was under the impression that this was going to be a terrible idea... what do you think?

**JNAProductions**

It is a terrible idea.

First off, Trajann is 140.
Blade Champ is 120.
Shield Captain is 120.
Five Guard are 200.
Two Allarus are 120.
Five Wardens are 260.
It did get the Callidus's points right, at least.

But even if it DID make an accurate list (1,785 by its points, 1,640 by real points) there's the issue that AI is so power intensive that it's wrecking the environment.

I get using AI to try to further science by automating tasks that are difficult/tedious and easy to mess up for people. Using it for something you can easily do on your own and do better than it? No. Don't.

**Tyran**

What the AI is doing is looking at a lot of competitive lists and giving something similar to that.

It will look fine and could be used as the start of a list, but it will get things wrong like point costs, so double check everything it.

JNAProductions wrote:

I get using AI to try to further science by automating tasks that are difficult/tedious and easy to mess up for people.

If it is a task that is easy to mess up for people then the AI cannot be trusted with it.

**JNAProductions**

Tyran wrote:
What the AI is doing is looking at the overall competitive lists and giving something similar to that.

It will look fine and could be used as the start of a list, but it will get things wrong like point costs, so double check everything it.

JNAProductions wrote:

I get using AI to try to further science by automating tasks that are difficult/tedious and easy to mess up for people.

If it is a task that is easy to mess up for people then the AI cannot be trusted with it.

Yeah, generative AI as commonly used doesn't have much purpose that I can see, not when it's so damn bad for the world.

**tauist**

I hear business is booming for people whos job is to clean up messes made by AI evangelists. Their accuracy is not great. However, no harm in asking for list suggetions I'd say, just take em with appropriate levels of salt and compare notes after the first few battles.

Problem with copypasting tourney winner lists verboten, you might not have a clue how those players utilized the inherent synergies and combos of such lists. Therefore, your performance with identical lists might vary greatly