ESIC Commissioner Ian Smith on MSI’s AI “cheat monitor”: “It should be completely banned from esports because it has a radically unbalancing effect on the meta of any game...”

Written By Naim Rosinski Content Manager & Editor
Last UpdatedFebruary 14, 2024 at 08:47AM
share on
share on

CES 2024 put on display a ton of AI-oriented technology and seemed to be the focal point of this year’s event. With tech such as an AI mattress and pillow set, an AI assistant named Rabbit R1, or a baby cry translator that will aid parents in understanding their toddlers being some examples, one thing stood out to gamers in particular. The MSI MEG 321URX QD-OLED gaming monitor. Despite the obnoxiously long name, the monitor has been onboarded with its own AI, named SkySight.

At the event, the monitor is showcased with SkySight put to work in a game of League of Legends. SkySight actively scans the player’s minimap to then create pointers at the edges of the screen to where the enemies are located. This is done through the team’s cones of vision. In other words, when an enemy is spotted on the minimap by the player or their teammates, SkySight will provide pointers to those spotted enemies. While not cheating per se, additions as such were never intended by the developers of League of Legends (Riot Games).

This has led to a debate in the community about competitive integrity and the use of SkySight. One thing that seems to be crucial about SkySight is that MSI claims that it will be launched with “Trainable AI Models” upon a firmware update. This will allow the training of SkySight in video games to aid in whatever the AI will be trained to do.

AI SkySight description at CES with
AI SkySight description at CES with

This has led to speculation on the possibilities of SkySight. YouTuber “jackfrags” has released a video speculating that SkySight could potentially be trained to detect enemies in Warzone if they are for example playing sniper positions from far away. This would work similarly like in the League of Legends showcase where the player may not see the enemies directly on the screen, but SkySight will highlight them for you if they are just a couple pixels in the corner of the screen, barely visible to the average gamer's eye.

In essence, it is currently impossible to 100% pinpoint how trainable the monitor’s AI is, but there is an aspect to consider, and that is competitive integrity.

We have asked Ian Smith, the commissioner of Esports Integrity Commission (ESIC), that deals with cheating and match fixing in the esports environment, about his thoughts on the MSI monitor with the AI SkySight module.

Ian Smith on the MSI monitor, AI modules, and esports integrity

What are your initial thoughts on AI assistance in competitive titles?

Ian Smith: The balance of any competitive title is a very technical matter over which the game developer has spent a lot of time and in consultation with the community with the balance change. If an external factor radically changes the overall meta of the game as this AI module would, it affects the entirety of the competitive scene. Unless every player uses the same monitor, the competitive balance between the players in any particular match is obviously affected. That would both, not be good for the competitive title, would unbalance the entire game, but would also under the terms of service in 99% of games be considered cheating as an external aid to the playing of the game. It would be similar to stream sniping and other forms of cheating that are heavily sanctioned by any integrity organization or game publisher.

I have no doubt that any game publisher made aware of the potential of this monitor would outlaw its use in competitive gaming. So yes, it gives a significant competitive advantage so even if every player uses it, negating that competitive advantage it changes the meta of the game to such an extent that whilst it’s confined to one ecosystem of 10 players that may be ok, as it may have been agreed, it’s not the same game as played by everyone else.

Apparently, the monitor features training AI SkySight locally, allowing it to, for example, draw more attention to health bars when low HP, highlighting enemies on the screen, or even potentially highlighting sound direction. Is such hardware featuring AI something that should be integrated and normalized, or should this be completely banned from esports?

Ian Smith: It should be completely banned from esports because it has a radically unbalancing effect on the meta of any game where competitive balance has been carefully constructed both by the developers of the game and the community in conjunction of the game commenting on each aspect. So whilst it’s conceivable, but in conjunction with the community you could now balance the game, to include how the software, the new AI, affects it, it would be irrelevant if everyone could use that AI software, that monitor, which we know is unlikely for a lot of reasons. So my view is that it should be completely banned in the same way that you would ban a hack, or external software, or stream sniping, or any measure that takes the players away from the core quality of the software un-enhanced by any external factor.

2023 saw a rise in AI being more accessible and usable by consumers. With AI inevitably creeping into the gaming sphere, do you think there should be any boundaries made to draw the line what is and is not allowed in esports as far as AI-assisted hardware software is concerned? 

Ian Smith: The obvious answer in my view what should be allowed and not allowed is to do with anything that potentially provides a competitor with an advantage over their opponent that is not part of the original game software or meta. So any external factor, that includes a second screen for stream sniping. Anything that allows you to see what you are not intended to see as part of the game should be outlawed and for the most part it already is. I’m most certain in every competitive game title that is used in esports there will be terms of service that cover the misuse or abuse of any external software or hardware. Of course, these AI modules, allied to hardware-like monitors, constitute both hardware and software. I don’t think they’d be allowed under current rules but I do think that it’s a significant factor that these sort of aids are going to become more common and should be addressed by the game developers.

Do you see a place for PC (or gaming in general) hardware with built in AI that may not be detectable by the PC’s operating system as a potential danger to competitive integrity? Especially at online qualifiers?

Ian Smith: The same thoughts apply. When game developers provide the software, especially for competitive titles, there is an assumption that the vast bulk of factors that enable the playing of that game are equal as between the competitors. So one of the primary tenants of that is that no software gives any player an advantage over another player. There are of course factors like ping or stuff that has to do with the internet but in your question, you highlighted the particular danger here, which already exists with things like stream sniping in particular. And of course with cheat detection, which is during online qualifiers, or any online competition, the ability of the tournament organizer or the publisher, game developer, to detect the external factors that any player may be abusing to give themselves an advantage is extremely difficult. Which is why you make nice, broad rules, regarding the integrity of competition, both in terms of betting integrity and just basic competitive integrity to ensure that if you detect something you can do something about that. You can take action against the player because it is cheating. The very clear intention of competitive gaming is to have the players compete on a level playing field that allows their skill as a gamer to purely determine whether they win or lose. Whilst in RNG-related games, there are elements of luck, you don’t want to tilt the table in favor of one player or team by providing external factors such as AI, stream sniping, bots, cheats, or hacks. I’m absolutely certain, especially from an ESIC point of view, these AI enhancements would constitute external cheats.

But in light of everything that you’ve laid out in your questions, my recommendation to games publishers and tournament organizers, particularly of online competitions, would be to specifically to address these developments in hardware and software in their terms of service and their rules and regulations in competition.

Note: Questions and answers may have been modified slightly for the sake of readability and clarity.

A former ESL teacher with a Master’s degree in English Studies that found his drive in esports. Merging a passion for the craft of writing with a never-ending need to out-aim others in the game of VALORANT and Counter-Strike.

Naim began his esports-writing and reporting journey at DailyEsports (now Upcomer), where he was a contributor for CS:GO. More and more enamored with esports, he quickly received a chance to attend IEM Katowice 2019 as press and managed to interview Cath, Lekr0, FalleN, and jkaem, his first dive into covering an event on-site.

With the release of VALORANT, Naim saw potential in Riot Games’ take on a 5v5 competitive shooter and joined Run It Back as a writer. After 3 months, he got promoted to a Managing Editor, managing a team of contributors and held regular brainstorming meetings to get content pumping for 3 years straight. In that time he attended Masters Berlin 2021, as well as VALORANT Champions 2022 as press on-site press.

Now, Naim is a Content Manager & Editor at THESPIKE. He keeps tabs on news, features, and evergreen content production, as well as contributing himself from time to time.

View More

Latest News

Comments

You must be logged in to be able to submit comments
Login

Top Predictors

Username
Points

Recent Activity

THESPIKE
© 2024 THESPIKE.GG | All Rights Reserved | Not affiliated with Riot Games
18+ Bet Responsibly | BeGambleAware.org