Summary
- Razer’s Project AVA uses AI to level the playing field in games by providing real-time tips.
- AVA’s AI can make games both easy to learn and master, bridging skill gaps and cutting the need for easier difficulties.
- AVA can customize guidance based on a player’s skill level, potentially making games more accessible.
Earlier this year, Razer revealed a future of gaming driven by AI. Artificial intelligence aid Project AVA is designed to provide real-time tips and tricks, making it a tool that could finally be used to end the debate surrounding gaming difficulty.
The Conversation Around Game Difficulty
The discussion surrounding video games, their difficulty level, and accessibility is not a new one. Yet, it seems with the continued popularity of harder genres and subgenres, particularly Soulsborne and Roguelike games, the debate has become more contentious. Taking a broad overview, it’s evident that two sides have formed: one championing a standard easy difficulty level in all games and the other backing a developer’s right to create their vision without concern for player skill level. Ultimately, one wants to lift the difficulty roadblock while the other thinks it’s an essential part of the creative process.
Related
What exactly is a “Soulsborne” game? Here are a few defining characteristics of the subgenre.
During the 2025 Consumer Electronics Show (CES), I had the opportunity to stand in a small room with the ghost of gaming future: Razer’s Project AVA. Large processors stacked atop one another whirred loudly through countless permutations to handhold a Razer representative through a round of League of Legends. I’ll get into the potentially horrifying specifics of a gaming future where we no longer have to learn a game’s mechanics later. On a slightly more positive and optimistic note for now, I didn’t realize it then, but I was staring at the potential answer to the ongoing back-and-forth over video game difficulty.
At the core of the debate is a question that technically both sides are asking: Should games be built with a generous learning curve?
Do All Games Need an Easy Difficulty?
Atari founder Nolan Bushnell once said, “All the best games are easy to learn and difficult to master.” It was a concept that made sense when games were controlled by two buttons and a joystick. Bushnell’s Law feels less and less relevant the further we stray from retro gaming, especially when you have a developer like FromSoftware, that really enjoys challenging players from the get-go.
From Dark Souls to Sekiro: Shadows Die Twice, the Japanese developer doesn’t seem too concerned about Bushnell’s Law. Rather, it serves the opposite purpose, creating games that are difficult to learn and even harder to master. The developer wants to, whether it’s explicitly stated or not, test players at every turn. Anyone who’s booted up Bloodborne knows that even the early-stage enemies are ruthless and require patience, skill, and an understanding of attack patterns.
FromSoftware doesn’t cater to the button-mashy hack ‘n slash games of yore. There’s no reward for spamming a light attack button or rushing into mobs without assessing the situation. Unfortunately, that kind of gameplay precludes some players from even remotely making any progress. I know this, not just because I was one for a spell, but because just about every time there’s an argument over game difficulty, FromSoftware’s name comes up as the biggest perpetrator of gamer-unfriendly experiences.
Of course, it’s not just the Elden Ring developer that challenges players. In 2004, Team Ninja faced criticism for Ninja Gaiden’s difficulty. Thirteen years later, Canadian developer Studio MDHR infamously stumped players with Cuphead (and no, I’m not talking about that review). Virtually any Roguelike game is inherently difficult from the opening moments. Despite their difficulty levels, though, these games bring up a fact that’s important to the discussion: They are all quite popular.
Related
Don’t Have the Time to Play an MMORPG? Try These Roguelikes Instead
Cut out the fluff and get to the fun without committing to the MMO grind.
From a business standpoint, no, games aren’t beholden to the skill level of the general audience. Developers can create for their intended market and still garner success. Not every gamer, particularly those who can’t enjoy these experiences, will agree. However, when it comes to “needs,” the overall entertainment factor trumps an inclusive gaming experience.
So, how does all of this tie into Razer’s Project AVA? After a little contemplation, and recalling exactly what the soothing feminine voice directed the player to do in League, it hit me. Developers won’t need to change their vision if AI can bridge the skill gap that prevents more novice players from enjoying intentionally difficult games.
Can AI Put an End to the Difficulty Discourse?
A game can be deemed too hard for many reasons. Maybe enemy movesets are hard to read. Perhaps resource management is in-depth and core to success. From not knowing the best build to tackle a boss to reading a computer opponent’s next move on a map of possibilities, challenges present themselves in a number of ways. Even the best players can only process a certain amount of information at once, though they may do so a little more intuitively than more inexperienced gamers.
An AI like Project AVA serves to put both on the same playing field by learning and adapting to the game’s mechanics. While it’s not meant to physically take control of a player’s inputs, it does what no detailed online guide can: It reads a situation and provides players with guidance on how best to proceed. Even for the toughest titles, artificial intelligence completely wipes Bushnell’s Law off the map, making games both easy to learn and easy to master.
It’s like having someone in your ear during a school exam, feeding you the answers. Best of all, it doesn’t compromise the developer’s vision. AI can just boost players who can’t grasp more complex gameplay, leaving others who enjoy the challenge to tackle the experience without any assistance. Dev teams still get to create the game they want at the difficulty level that best fits the intended experience. No one has to feel left out, but the integrity of the developer’s vision remains intact.
How AVA Can Level Out the Learning Curve
During the brief demo I saw of AVA, Razer presented a few scenarios where the AI was most helpful. For those unfamiliar with League of Legends, players control a hero and are ultimately vying for control of a map. It’s a little more involved than that, and I’m sure LoL diehards are typing out long scolding emails at this moment, but that’s the very basic gist.
While playing an AI opponent, the Razer representative was being fed suggestions to achieve the cleanest victory. What route to take, when to push the enemy, what consumables to buy: AVA was there for the ride, firing off suggestions at a surprising pace. Now, that’s just the League of Legends loop. Apply the same type of learning to a game that thrives on players dying. Instead of spouting off the best routes to take on a map, AVA can warn of an enemy’s attack pattern, literally guiding players on what actions to take to get a leg up and advise on the best loadout, the most effective spells, and so much more.
That harder level is suddenly less of a problem when you have someone advising, in real-time, that you need to dodge, perform a heavy attack, cast a debuff, and then finish with a flurry of quick attacks. It’s a difficult concept to illustrate because it is limited only by what the player can do in the game. In a sense, it’s the equivalent of having your older brother take over to get you through that obnoxious second level in Ninja Gaiden (1989).
Since AVA caters its guidance by combing through performance, it can tailor the tips to ensure a fully customized experience that caters to a player’s current skill level. Essentially, lower-skilled players will receive much different guidance than someone who exhibits a higher skill level. While it may not allow a beginner to match the proficiency of someone who considers themselves an expert, it does allow the beginner the opportunity to make strides in a game they may have otherwise stumbled through.
Skirting Ethical Boundaries With Assistive AI
While we’ve been focusing primarily on simplifying the single-player experience, surely any avid gamer reading this has questioned the ethics of integrating AI into a multiplayer environment. That’s an entirely different discussion, but one we’ll touch on briefly to at least quell concerns.
According to Razer, Project AVA is being developed to “fall within game developers’ terms of service.” Meaning, Project AVA can’t be used to cheat as it is perceived by the game creators. There is surely going to be a lot of back and forth on this, but it’s worth noting that virtually every multiplayer game has a “no cheating” clause. In the League of Legends code of conduct, it essentially bars the use of “third-party exploits, hacks, [and] bug abuse.”
For now, Project AVA could fall under a “third party exploit,” that is until Riot Games outwardly expresses otherwise.
Could artificial intelligence change the multiplayer gaming landscape? Even if it can’t be directly used in a multiplayer match, AVA’s core purpose is to make gamers better. The guidance the AI offers outside of the online environment will undoubtedly affect one’s skills, potentially making them a bigger threat in online gaming.
Related
As AVA is still in development and hasn’t entered a beta environment, any concerns over the AI’s direct role in multiplayer gaming remain largely speculative.
