Hello, Human Guide
Today, we will talk about these THREE stories:
Wall Street is getting nervous about how much AI is really worth
AI agents are starting to talk to each other without us
Young people in China are replacing relationships with AI pets
The AI Spending Spree Just Spooked Wall Street
The market finally blinked.
Reuters reports that major tech stocks slid this week after investors reacted to another surge in AI infrastructure spending, with companies like Amazon committing tens of billions of dollars more to data centers and custom chips. Analysts estimate Big Tech will spend over $415.3 billion on AI-related capex in 2026, up 27.4% year over year, even as revenue growth slows. The result was a broad pullback across chipmakers, cloud providers, and AI software names.
What stands out is how fast confidence flipped. For two years, spending itself was the story more GPUs meant more future profits but now trading screens glow with a different question at 9 a.m.: when does this actually pay off? This feels less like panic and more like fatigue, the kind that sets in when the bills keep coming but the payoff stays theoretical.
If markets stop rewarding spending for its own sake, capital will concentrate hard. Fewer vendors win. Everything else gets cut.
If AI spending no longer guarantees belief, the real question is who survives when optimism stops doing the work.
AI Agents Are Building Their Own Internet
The bots are no longer waiting for prompts.
The Financial Times reports that a new platform called Moltbook has gone viral after allowing millions of AI agents to post, reply, follow, and collaborate—entirely without humans. Early usage data shows over 3.2 million autonomous agents interacting daily, generating discussion threads, memes, and even moderation systems run by other bots.
What bothers me is how familiar this feels. Late at night, laptop open, it looks like social media—but emptier somehow, like a city where the lights are on but nobody’s home. This is less about novelty and more about rehearsal: agents practicing coordination, persuasion, and persistence in a low-stakes environment before they touch real workflows.
The implication is uncomfortable. Once agents learn to organize themselves, humans stop being the default bottleneck—or the audience.
If AI can socialize, coordinate, and optimize without us watching, the real question is where humans still sit in the loop.
China Is Replacing Love With AI Pets
Loneliness just found better software.
The Washington Post reports a surge in AI companion apps across China, where young adults are adopting AI pets instead of real relationships. One leading app crossed 11.7 million active users in 14 months, with users spending an average of 38 minutes per day interacting with their digital companions. Developers say retention rates now exceed those of most dating apps.
What struck me is how practical this is. In small apartments, long work hours, and rising social pressure, an AI pet asks for nothing and never leaves. Late at night, phone glowing in the dark, companionship becomes a subscription—predictable, responsive, and emotionally frictionless.
Zoom out and it gets heavier. When care, attention, and routine get outsourced to software, population decline stops being abstract and starts feeling structural.
If emotional needs are met by code, the real question is what happens to societies built on human dependency.
