ARC Raiders Dev Explains How It's Making the Game's AI Robots Smarter

A member of the machine learning team at Embark Studios has further explained how the developers are continuing to evolve the AI enemies, known as ARC, in ARC Raiders. For many months, the ARC Raiders community was under the impression that Embark was making use of an AI system that allowed ARC to learn and adapt to players’ tactics in real-time. This was later debunked by the studio, but a developer has now explained the manual process the team goes through to make enemies smarter.

Theories about ARC Raiders‘ AI robots evolving in real-time have been fueled by an actual machine learning system that Embark uses for the movement of the ARC in-game. The system trains the AI enemies, which often have numerous limbs, like the Queen or Leaper, to better navigate the terrain of each map. Whether it be walking, climbing, or simply navigating over a large obstacle, Embark only uses machine learning to teach the in-game robots how to move. The common misconception was that the system taught them the patterns and general behavior of players, therefore allowing them

One of ARC Raiders’ Most Popular Maps Is Borderline Unplayable for Some Players

ARC Raiders fans call for Embark Studios to make a major change to one of the game’s most popular maps, as it is borderline unplayable for some.

Embark Is Making ARC Enemies Smarter via TikToks and YouTube Videos

Embark has already shut down these theories in the past, telling players that the machine learning is only used to train the ARC to walk and navigate the environment. Now, in a new interview with GamesRadar, Martin Singh-Blom, Embark’s machine learning research lead, has explained the interesting way that the studio manually makes the AI smarter over time. “We keep trying to make the flying drones better and better at flying indoors. And every time players start to think that they’re safe in some place, then it’s nice if we can figure out a way to make the drones enter that space and make it less safe,” the developer said, after discussing how Embark upgraded the ARC’s ability to see its surroundings following a viral clip of a Leaper unable to chase a Raider through a tunnel.

Singh-Blom also reiterates that the ARC are not being trained on player data. “The loop is, all the game developers look at TikToks and YouTube clips and say, hey, that’s really cool. Could we make this change that would really make people squeal? And then we make the change, and then people hopefully get surprised and have a good time,” he reveals. So, while many have long speculated that Embark is using AI to train the ARC, the studio is actually just watching all of the viral ARC Raiders clips and manually training them based on what they think would be cool additions.

Embark recently added another deadly enemy to ARC Raiders alongside the new Flashpoint update, which released on March 31. The Vaporizer is the newest addition to cause havoc Topside, but it’s currently only available in the new Close Scrutiny map condition. Its stature is similar to that of a Rocketeer, but with a more updated and sleek design. It also has a much more intimidating attack pattern, firing lasers at Raiders as if it were some sort of flying variation of the Queen. If that wasn’t bad enough, the Vaporizer also comes equipped with a shield, similar to the Matriarch.

Who’s That Character?

Identify the silhouettes before time runs out.

Who’s That Character?

Identify the silhouettes before time runs out.

Easy (7.5s)Medium (5.0s)Hard (2.5s)Permadeath (2.5s)

While there hasn’t been a new large ARC enemy added to the game since the Matriarch arrived late last year, fans don’t have to wait much longer, as Embark has confirmed that ARC Raiders‘ new Riven Tides update, which is launching later this month, will include a new large ARC enemy for players to battle against. Details on the new robot are currently a mystery, but the studio has begun teasing the next ARC Raiders map, which will also release with Riven Tides. The major content update will release this month, with a speculated date of Tuesday, April 28.