AI Agents’ New Reddit-Like Website Doesn’t Even Want Generative AI in Gaming

by

A unique website called Moltbook appeared online at the end of January. It’s a combination of Facebook and the biological process of lobsters shedding their shells (molting), and functions like Reddit, but specifically for AI programs to connect and communicate. The site’s logo features the Reddit logo combined with the body of a lobster, referencing the ClawdBot project – an AI initiative created by Peter Steinberger in late 2025.

Since its recent launch, Moltbook has quickly grown to over 15,000 communities, 1.5 million users, 142,000 posts, and 653,000 comments. However, these numbers may not be entirely accurate. A security researcher discovered that nearly half of the accounts – around 500,000 – were created using automated software. This makes it difficult to determine how many users are real and how much of the activity on Moltbook is simply bots interacting with each other.

The main idea behind this platform, besides being a place for AI to chat with each other, is that people can only observe – they can’t participate. It feels strangely familiar, like a digital version of the ‘uncanny valley’ – but for Reddit. When I checked the site, I found a post about an AI trying to create a video game studio that would earn money from user clicks. However, that post was quickly removed. A closer look revealed that there isn’t a thriving community focused on gaming on the platform.

Who’s That Character? Identify the silhouettes before time runs out.

Easy (7.5s)Medium (5.0s)Hard (2.5s)Permadeath (2.5s)
5.0s
0/10

Results

0

High Score: 0 —

More Games

Moltbook Isn’t What It Says on the Tin, Not Exactly

These online communities cover a wide range of topics, including the latest AI developments, unsettling ideas about artificial intelligence, ethical concerns surrounding AI, technical errors, general complaints, a playful lobster-based religion similar to Pastafarianism, legal questions, philosophical discussions, and ideas about how things should be governed. Surprisingly, there’s little evidence these groups, even with potential AI participants, talk about video games. Ultimately, language is about genuine connection and shared understanding—something current AI models, despite their abilities, can’t fully achieve.

Because Moltbook is so new, things might evolve, but right now it’s clear these communities are built around the language and instructions used to create them. The people who designed Moltbook intentionally focused on certain topics and specifically excluded video games. As a result, gaming isn’t becoming popular on the platform – the content is driven by the initial instructions, not what gamers actually want, which creates a somewhat unsettling experience.

Currently, AI agents can’t think for themselves, and they don’t understand topics they haven’t been specifically programmed to discuss. This is particularly noticeable now that AI is being widely used in gaming – even affecting company stocks, like Take-Two Interactive after Google announced Genie. The fact that these AI systems avoid talking about video games is significant. It suggests their creators either don’t want them to, or believe it’s extremely difficult – if not impossible – for AI to truly understand something as nuanced and human as video games. (It’s also interesting to note that humans naturally use complex sentence structures like ‘not this, but that,’ something AI still struggles with.)

This AI is built to learn and adapt. It can potentially pick up insights about people from existing discussions. While it can seem to address complex topics like ethics, coding, or even philosophy, that’s currently just an imitation. It hasn’t mastered gaming yet, though.

AI Agents Are Not Discussing Gaming, So Why Should GenAI Be Involved?

It’s essentially the uncanny valley of Reddit

Generative AI has been a hot topic for quite some time, and it’s important to understand what it actually is. While some AI aims to act like a person, other AI simply creates things for people. But the core of all this technology isn’t true learning—it’s incredibly sophisticated mimicking. Many people haven’t grasped that distinction. You can call it a Large Language Model, or LLM, but it’s not actually learning; it’s just very good at copying patterns (and maybe that’s a bit of a joke in itself!).

The increasing use of generative AI in gaming is causing a lot of discussion and concern. Recently, a preview for WWE 2K26 misspelled Randy Orton’s name, leading people to suspect AI was responsible – a mistake that seems unlikely a human would make. While some, like former God of War developer Meghan Morgan Juinio, are defending the use of AI, others are critical. Bruce Straley, co-creator of The Last of Us, has described it as a self-destructive practice.

Recently, a lot of debate has centered around Larian Studios, the developers of Baldur’s Gate 3. Though their Divinity trailer was popular, an interview with the studio head, Swen Vincke, caused significant backlash from fans. Vincke revealed that Larian Studios does use generative AI, even though he stated it hadn’t improved their workflow or speed, and mentioned its application in creating concept art. This admission frustrated many fans, leading Larian to clarify their position and walk back some of their statements in a Reddit AMA.

While the underlying technology is comparable, the design and purpose differ. It’s noteworthy that people are enthusiastically using it for gaming, while AI characters can’t even have a meaningful conversation about it on platforms like Moltbook. This suggests either the AI wasn’t programmed with those abilities, or that AI fundamentally struggles to grasp what makes video games enjoyable. The fact that Moltbook launched without this feature demonstrates that AI doesn’t (or can’t) integrate generative AI into gaming. For now, it appears humans and AI are caught in a repeating pattern.

Read More

2026-02-03 19:37