“Silence on joue!” is a French audio podcast about video games, produced by the French newspaper Libération, for 17 years.
It is run by Erwan Cario, a journalist at Libération, who talks about video game news and reviews the latest games released with his team every week.
I personally discovered Mr. Cario through his appearances on the French television channel Nolife and I have been following his work for years.
I particularly appreciate their lively and intelligent discussions where they manage to put each element into context, and are able to talk about the history of this or that game franchise.
However, in episode S17E23 that I listened to, I noticed several points of disagreement and a certain lack of nuance in the words of the interlocutors.
To understand the context, here is the episode in French, followed by my comments:
Here are my comments on the topics discussed:
1) Jobs and AI
In the podcast, it is said that generative AI (Artificial Intelligence) destroys jobs because it replaces employees’ know-how with randomly generated creations by AI.
It is true that some companies use this as an excuse to get rid of thousands of employees. While companies use AI at this time, it is of course not a substitute for employee know-how.
However, I think the wrong questions are being asked about generative AI. It is not whether or not it destroys jobs, but rather whether it is our relationship to income and work.
When industrialization began in Western countries, it made the manual work of many employees redundant, as their work became automated by machines and robots.
When computerization began, it made the work of many employees redundant, as many tasks became computerized and automated by computers.
I don’t think we’re going to be able to ban or censor generative AI.
Jobs will change again, some will no longer be useful, others will be created, but as long as we link the right to housing, food, electricity and health to the income we receive, we will continue to live in dysfunctional societies.
If we put machines, robots, computers at the service of the common good in order to allow a population to have what it needs to live decently, it can only be beneficial. These are vital services of general interest that are kept for the benefit of private interests.
In our current societies, any progress will be used to make a profit, but it is not the fault of this or that tool that automates tasks, it is because we decide to prioritize the individual and personal enrichment that we will continue to have these kinds of consequences.
2) “If there’s generative AI in a game, it’s bound to be crap.”
Generative AI can generate text, images, sounds. This can be used to generate a single image, for example, as well as millions of images. It’s hard to say that just because there are one or more elements generated by generative AI, “it’s going to be crap.” It’s a very generalizing narrative that ignores the immense amount of work developers have to put in to make a game.
For example, a character in a game might visit a street with AI-generated graffiti, while everything else is created by the developers. AI is used here to fill a gap of developers in terms of graffiti and street art. Many independent developers operate on a very small budget and would not be able to afford the services of an artist who specializes in this field, for example.
Generative AI can be used in the form of satire, or parody as well. There can be humor in a game, where you make fun of what a robot character would have done as a work of art.
All of this is to say that there are many possibilities for use, and that I think that just classifying a game with one or more AI-generated resources as bad is quite reductive and binary as reasoning.
3) A generative AI whistleblowing system in Steam
With a new system where Steam users can report whether or not they think a game contains generative AI, it opens the door to more review bombing. That is to say, ill-intentioned users could organize themselves into groups in order to give a game on Steam bad ratings.
As soon as a company makes a bad PR (Public Relation) move or a move considered to be like that, these games could have their rating affected by ill-intentioned users.
We’ve seen this in both positive cases (poor company communication, reduced effects of a weapon) or negative cases (adding DRM to a game).
If some players are frustrated that they don’t get what they expected in the game, classifying the game as “AI-generated” could be a form of revenge for them.
So I don’t think that a new system of this kind could be appropriate, and could lead to abuses.
Generative AI is something in its infancy, which has a bad reputation among some, but not always rightly. Its use can be very useful, but the ecological cost of its use has a great impact on the environment.
I am aware that generative AI relies on plagiarism and the massive collection of data without permission. I think that the strategy of the companies that offer these services is to break the market, by going very fast, like AirBnB and Uber, before politicians react. It’s a strategy that I condemn, but I don’t see how to stop it, and I don’t think vilifying its use is going to help anyone.
That’s it for my points of disagreement about this episode.
What do you think? Agree, disagree? Comment below!
Credits Featured image: thumbnail of the video Silence on play S17E23 / Liberation with an image of Ubisoft’s Prince of Persia Lost Crown in the background.
