Dear Google, Please Don’t Ever Mention Doom Again


Google I/O 2026 was deathly boring. Speakers talked about AI — of course — tokens, agents and other things that put me to sleep. Until I was shocked awake by Google’s Varun Mohan. He took the stage and dealt me a psychic blow I wasn’t prepared for when he included Doom in his portion of the presentation.

Mohan showed Google using its AI agent-infused development platform Antigravity to build an operating system and asked it to play Doom as a sort of QA test. It couldn’t because it lacked video and keyboard drivers, he explained. Mohan then prompted Antigravity to add the necessary drivers, which it did, allowing him to launch Freedoom, the free open-source game based on one of the most influential video games ever made.

Doom is a foundational first-person shooter game, full of demon-slaying action and over-the-top fun, while Google’s I/O presentation was a lifeless parade of unasked-for AI features. As CNET’s Lori Grunin wrote in a roundup of her favorite things from I/O, there was “a lot … that looked problematic at best and dystopian at worst.” 

Doom and Google’s AI ambitions are so far apart conceptually that hearing them in the same sentence side by side hurt. Those AI tools could never have the ingenuity or creativity needed to build something as lasting as Doom.

When Id Software released Doom in 1993, its complex level design was a major achievement. The game featured varying landscapes, locked doors, hidden chambers and even different types of elevation within a level. Years before GoldenEye 64 left its mark on the first-person shooter genre, Doom introduced mazes of death and destruction with tight hallways and narrow hiding spaces, and the lighting helped make these levels, filled with plenty of places for enemies to hide, feel truly scary. The ingenuity of its level design is still studied today and continues to inspire communities of players and creators who design levels and modify the game.

A Google I/O presenter playing Freedoom

Google’s Varun Mohan.

Google

Doom’s art direction was also unique. It combined sci-fi, fantasy and heavy metal influences in a way that made it feel wholly original. It’s essentially a cross between Dungeons & Dragons, Ridley Scott’s Alien and the band Slayer. That creative vision helped shape countless games that followed. I can’t imagine films such as Event Horizon or games like Halo ever being made without Doom.

The current slate of AI tools, however, lacks the creativity needed to push any boundary forward or create something that feels lastingly influential. These tools rely on chainsawing past works into mincemeat and reassembling them into a simulacrum of novelty. There’s no forward thinking involved, only derivatives of what has already been made.

That lack of creativity was especially apparent while I watched presenters try to hype up Gemini’s ability to [checks notes] write emails and come up with ideas for family activities. There wasn’t anything particularly inspiring about the technology or the ways Google told us we could use it.

And sure, Doom was an original game that borrowed from earlier works, but it built on those influences to create something new and lasting. These AI tools don’t have — and never will have — the ingenuity to make something with Doom’s staying power. Instead, they just give us back versions of things that have already been made, covered in a weird, creepy sheen.

For more from Google I/O 2026, here’s what to know about Google’s Project Aura and what to know about Ask YouTube.





Source link

Leave a Reply

Subscribe to Our Newsletter

Get our latest articles delivered straight to your inbox. No spam, we promise.

Recent Reviews


Researchers in South Korea developed a wearable system that uses seven smart rings to read finger and hand motions to translate American Sign Language and International Sign Language into text. The purpose is to make communicating easier between those who sign and nonsigners without needing a separate human interpreter. 

AI Atlas

According to the study, published Friday in the journal Science Advances, the system reliably recognized 100 ASL and ISL words during testing. It also performed well with users the system had not seen before, and it didn’t require recalibration for each person. Because the system detects words in sequence, it can produce sentence-level translations without extra training on grammar. 

ASL and ISL are the everyday languages of more than 72 million deaf and hard-of-hearing people. However, most hearing people do not know any words in these languages or have a very basic understanding. That gap makes certain tasks, like ordering at a restaurant or asking for help, much more difficult. 

A graphic shows two illustrated people talking in sign language, ASL and ISL. The graphic also shows the different components of the ring as well as pictures of hands modeling the rings.

A concept of how the rings work in the real world. 

American Association for the Advancement of Science (AAAS)

Existing sign language translator prototypes often rely on bulky gloves that can distract from or block natural hand movement or feel uncomfortable for the wearer, which limits real word adaption. Camera-based technologies can work well in controlled environments but are often limited to those places where a camera can be set up with a clear line of sight, the researchers wrote. 

To solve these problems, the researchers designed sensing rings for each finger that can capture precise motion and finger position while letting the hands move naturally. The rings can detect both signs that involve movement, like the words for “dance,” “fly” and “sun,” and signs that are held still, like “I” and “you.”

“These advances suggest that [the device could enable] barrier-free public translation systems for unseen users and unrestricted daily assistive interfaces,” the authors wrote in the study. 

The authors are affiliated with Yonsei University, Hankuk University of Foreign Studies and the Korea Institute of Science and Technology, among others. While the technology is still experimental, the authors wrote that the technology has the potential to ease communication difficulties. The underlying idea could also help improve controls for other systems, like virtual or augmented reality.

“Beyond sign language translation, the ring-type, wireless, and modular architecture of (wirelessly connected, ring-type sign language translators) may also be extended to other gesture-driven applications such as virtual or augmented reality control, touchless device interfaces, or rehabilitation monitoring systems where fine-grained hand movement tracking is essential,” they wrote.





Source link