• 0 posts
  • 50 comments
Joined 3 months ago
Cake day: August 22nd, 2026
  • I think its about how. Using LLMs to write entire game sub systems, is a recipe for bloat, poorly optimized/hard to maintain code. We call this vibe coding. Its a problem. Devs who forgot how to code is a real problem. Studies are finding devs spending MORE time coding than less, mostly debugging.

    But using LLMs to remind oneself how to most efficiently sort a list, or write a database query, its very handy. For our indie game dev team we have a strict rule: not allowed for full systems and classes, and all AI written code must be read and interpretable before we accept code merge requests.

  • Well put. My day job is training AI for major corporations and governments and my point has become the same: its an advanced autocorrect feature, built on a model and architecture that guesses the best answers based on language patterns and nothing more.

    For programmers to use them to remind themselves of the fastest way to iterate a list, or map a matrix, it is amazing. But anything more than a single function in size and its immediately risky for the same reasons it is a good guesser.

    Programmers have turned out to spend MORE time coding than before LLMs https://metr.org/blog/2025-07-10-early-2025-ai-experienced-os-dev-study/, from my experience that is due to not understanding the limitations of the architecture. But that so means admitting they aren’t that game changing as we were sold.

    Especially with art. LLMs are absolutely dog shit at art if you need to iterate for accuracy.

  • The US falling does not mean China will supplant them. China struggles to build alliances and have serious border disputes with nearly all of their neighbours due to Chinese imperial ambitions in the region, and how the CCP treats diplomacy.

    Also, the AI bubble is popping in China too which has shed hundreds of billions too.

    We can criticize and should criticize US and Israel, without glorifying another despotic regime.