Hello, dear readers!

We are reviving our weekly brief on remarkable AI topics, so you can start your day knowing all that's worth knowing about what happened last week in AI.

Today's focus — Google's Project Genie, an AI model generating virtual worlds and shaking markets. Top gaming stocks suffer record losses on uncertainty. Do we even need gamedev platforms anymore? Spoiler: might be early to cancel your Unity subscription.

Also in this week's edition:

Genie: Beautiful Worlds, Too Few Fingers

About a week ago, Google unveiled an experimental AI tool for building virtual worlds, Project Genie. Tech markets haven't experienced this big of a shakeup since last year's release of DeepSeek, which sent Nvidia and foundry stocks plummeting. This time it was the gaming tech that fell victim to AI menace — the stock of Unity, the developer of arguably the leading cross-platform game engine, lost almost a quarter of its value.

The initial reactions ranged from alarmed to ecstatic. Some were in awe of how easily the model can generate realistic worlds with controllable characters — all from a single prompt. Whether it's a runaway tire rolling through suburbia, a stray cigarette pack exploring a subway station, or a shady GTA-meets-Watch-Dogs protagonist walking down Hollywood Boulevard — the results indeed deserve attention.

Other commenters preemptively bemoaned the death of numerous professions — from level designers and 3D modelers to artists and riggers — hence the market dive for Unity. The game engine developer's stock tanked by over 24%, and still has not bounced back. 

However, it is not entirely clear how well Genie handles tasks requiring some more creativity than [generic object in a typical location]. Also, the bane of AI image generation — namely, missing fingers and other peculiarities — is ever-present even in Genie's creations.

Altman vs. Anthropic: When the Super Bowl Gets Weirdly Meta

Some of the drama last week came not from flashy demos, but in the form of a Super Bowl ad. Anthropic took out prime real-estate time with a series of spots mocking the idea of ads in AI conversations, waving the banner “Ads are coming to AI. But not to Claude.” The jabs weren't just playful — they hit a sore spot. Sam Altman previously called ads a “last resort” for AI monetization, a position he's now quietly walked back as OpenAI rolls out tested ad placements in ChatGPT's free and Go tiers.

Sam Altman didn't let it pass. The OpenAI CEO denounced the ad as “deceptive and dishonest,” accusing Anthropic of oversimplifying and fearmongering. The exchange quickly spilled onto social media and into industry group chats. If nothing else, it was a reminder that the AI culture war has officially gone mainstream — right there between beer commercials and pickup trucks.

Moltbook: Explosive Growth, Questionable Posts

The AI-native social network continues to grow at a pace that would make another platform jealous. Moltbook claims 1.6 million users, fueled by viral screenshots and outrage-bait posts — including several infamous “kill all humans” style rants.

The twist: some reports suggest that many of these posts weren't generated by AI agents at all, but written by humans cosplaying as unhinged models to farm attention.

Whether that makes Moltbook more or less interesting is still up for debate. Is it an authentic glimpse into emergent AI expression, or just the internet doing what it always does when a new toy appears?

Thanks for reading AIport. Until next Monday — by then, AI will definitely do something we can't possibly expect.