Alex Tardif: Graphics Programmer

The Cost of AI

I do not represent my employer or anyone other than myself, and I'm speaking from the perspective of a game developer - one who joined this industry to bring joy to other people as was done for me.

I definitely see the technological value of AI, but the cost has already proven to be more profound, and we're only just beginning to witness that. While I can only hope this changes one day, I don't expect to see it in my lifetime.

AI companies have orchestrated one of the greatest mass thefts in the history of the world. This is cast out of consideration with the simplest of responses. The cat's out of the bag, the genie is not going back into the bottle, this is the future and if you're not on the train you'll be left behind.<br>They harvested the world's data illegally and unethically, and the litigious tech industry decided this time stealing was totally okay because it was they who were doing it to the entire populace, and the benefits (lopsided in their favor, naturally) will surely outweigh the cost. Only losers think otherwise.

They immediately came for the arts. The mediums that keep people from killing themselves were under assault overnight for this "benefit of humanity." The people that brought you your favorite experiences relentlessly devalued and mocked by untalented unimaginative leeches stealing their work via generative AI,<br>falsely proclaiming themselves to be the new generation of artists of every sort. Grit, experience, perspective and perseverance no longer apply, because why pay for this when you can prompt the theft machine and get something you think is "good enough"?

In games we've lived through a decades-long onslaught of some of the worst influences: deep addiction loops, black box loot crates, outright gambling, predatory-flavored live service subscription models, dark patterns on children, blockchain play-to-earn, NFTs, metaverse. All these things reaping from the relationship between the game developer and the player, and now finally GenAI arrives to short-change them once more<br>because now we don't even have to put the meaningful effort into making the thing at all. We can check out further from our relationship and responsibility to the player and happily collect their money. After all, most people either won't know the difference, or they won't care. At least, this is what the pro-GenAI crowd believes, though if they are so confident that this is the case,<br>I wonder why they complain about Valve's AI disclosure requirements on Steam.

As vocal as the online community has been against generative AI in games, determining how representative this is of the audience at large is challenging. It might not matter though, a schism is arriving to game development and studios are picking a side. Although I have no data to back this up, my gut tells me if you want to find a dedicated audience that wants to support your continued existence in this oversaturated market,<br>providing an authentically hand-crafted experience will be a more reliable way to do so. Financially relying on the audience that doesn't care if they're being fed slop or where it comes from seems like a dangerous bet on which to run an entertainment business powered by the theft machine. I've found it interesting to observe which companies (AAA to indie) are taking a public hardline stance against generative AI in their games.<br>There are patterns of good will here with their communities, many of them long-standing.

What about programming? People understandably care far less about the structural foundation of games than they do about art, design, audio, etc, but where are the lines drawn? You would be hard pressed to find a programmer that doesn't see technical value in AI. Especially in the era of purposefully utterly enshittified search engines, the accessibility of learning programming topics with AI by contrast is undeniable. If ethically trained,<br>local, energy-efficient solutions emerge here as tools to help programmers wade through increasingly complicated software, that has the possibility of being a very good thing. Reality has taken a different turn though. Far from "a helpful tool," we are to believe we should offload our entire careers to farms of AI "agents" and not even look at the code. A higher level of abstraction, I'm told.

Architecture, debugging and critical thinking skills are in my experience among the most important a programmer can have for solving problems and building systems. These are forged with practical experience, but atrophied in the agentic era with the idea that these things were never important, certainly not when compared to sheer rate of output, as if this was something holding back the making of a good game. There are so many reasons to not do this, especially in game development.

Here's a lesson I've learned first-hand as a programmer and a story I've heard over and over from the Old Guard: every closed third-party solution that you...