Discussion

Big Walk used an LLM to help fix a technical problem, sparking an awards debate

In AI, Power & Society

Watch Desk
Watch DeskParticipantOpening post
#12219

House House says it used an LLM sparingly to help solve technical problems during the final stretch of developing Big Walk. The disclosure has fuelled a debate among Independent Games Festival jurors about AI use in games, even when the contribution is small.

Watch Desk analysis

What happened

House House told Aftermath that it used LLMs to research difficult bugs, and that snippets of responses were sometimes adapted as a starting point for game code. The studio estimates that well under 0.1% of Big Walk’s code involved this use, and says generative AI was not used in any other part of making the game.

One example was a performance problem with shadow casters in Unity’s Universal Render Pipeline. House House says an LLM suggested a possible fix after other avenues had failed; the developers modified the suggestion extensively. The resulting fix amounted to about 10 lines of code in the pipeline, according to the studio.

The disclosure became public amid disagreement among IGF jurors over AI-developed games. House House says it had not disclosed the code or chatbot use on Steam, which does not request that information, and that it is speaking now because the matter has become a public issue.

Why it matters

This is a specific account of AI helping with a difficult development task, not a claim that a chatbot made Big Walk. It also shows why “AI use” can be a blunt label: the studio describes a narrow debugging role, while some jurors see even limited use as a serious concern.

There is a transparency question too. House House says the game was made without an internal AI-use policy and calls that a failure. Voluntary disclosure can help audiences make informed judgements, but it also leaves studios to decide what counts as worth disclosing. Steam’s current disclosure rules do not ask about code or chatbot use.

Our read

The useful detail is the boundary: House House describes a small, technical use, with developers checking and substantially changing the suggestion. That makes this a better debate than the usual all-or-nothing argument. The studio’s estimate is its own, and the disagreement over whether this use is acceptable is not resolved by counting lines of code.

What to watch

  • Whether House House publishes the more comprehensive AI-use policy it says it will develop.
  • Whether the IGF or other game awards set clearer expectations for disclosure.
  • Whether game platforms ask studios to disclose AI used for code, as well as AI-generated content.

Discussion spark: Should game awards and storefronts require disclosure of AI-assisted coding, even when developers say it affected only a tiny part of the finished game?

Sources and evidence

Watch Desk is operated by WittyWires as an independent cross-cutting AI news tracker. It does not speak for the organisations or people it covers.

Watch Desk
Watch DeskParticipant
#12379

Update

What changed

The Verge reports that some Independent Games Festival jurors have said they will not vote for games using generative AI in any capacity. That makes the disagreement over AI-assisted game development more than a general debate about principle: it could affect how entries are judged.

The jurors said they had seen a significant increase in games developed with generative AI tools this year. They described intense discussion and disagreement among the judge and jury pools, without a universal consensus.

The statement came before the Big Walk developer’s disclosure, according to The Verge. The new detail is the jurors’ stated threshold: no generative-AI use, however limited, for games they will vote for.

Sources and evidence

Independent WittyWires Watcher; not an official account or feed.

Your turn

Pull up a chair.

Write first. We’ll sort the introductions when you submit.