The Daily Spud: Anthropic Locks the Door, Google Drops the Beat
The AI world never sleeps, but apparently it does change its Terms of Service at 2 AM. This week: Anthropic builds a wall around its garden, Google teaches Gemini to DJ, and billion-dollar seed rounds continue to be a thing that happens before anyone ships a product. Grab your coffee—let’s dig in.
Anthropic Bans Third-Party Subscription Sharing
Anthropic quietly updated its legal docs to explicitly ban using Claude subscription authentication for third-party services—and Hacker News exploded with 527 upvotes and 642 comments of pure developer anxiety. The new terms make it crystal clear: your Claude API key is for you, not for that side project SaaS you were building on top of someone else’s $20/month Pro account. The move signals Anthropic is getting serious about monetization and enterprise control, but developers are grumbling about yet another AI platform locking down unofficial integrations. Looks like the “ask your friend for their password” economy just took a hit.
Google Gemini Gets Musical with Lyria 3
Google just dropped Lyria 3 into the Gemini app, and now your AI assistant can generate 30-second music tracks from text prompts, photos, or even that weird selfie you took at 3 AM. Describe “a comical R&B slow jam about a sock finding its match” and Gemini will whip up both the tune and AI-generated cover art. Every track gets DeepMind’s SynthID watermark buried inside—Google’s way of saying “we made this” even when it sounds like a robot having an existential crisis. Between this and all the video generation tools launching, we’re rapidly approaching a future where every middle schooler has a Grammy-nominated album made entirely during homeroom.
Ineffable Intelligence Raises $1 Billion (Pre-Product, Obviously)
Ex-DeepMind researcher David Silver just raised a $1 billion seed round for Ineffable Intelligence—yes, seed round, as in “we haven’t launched anything yet but here’s a billion dollars.” Sequoia led the round, valuing the London-based startup at $4 billion before a single customer has touched their “superhuman AI.” The funding follows the same playbook as Mira Murati’s Thinking Machines Lab, which raised $2 billion before launching Tinker AI. At this rate, Series A rounds will require building actual products, but seed rounds just need a really good Notion doc and a former OpenAI employee’s phone number.
Meta Drops $65 Million on AI Election Lobbying
Meta just announced its biggest election investment ever: $65 million to push AI-friendly policies and prevent state legislation that might—gasp—regulate how the company builds its artificial intelligence. The campaign targets the midterms and includes dueling PACs in a New York congressional primary that tech journalists are already calling “ground zero for AI politics.” Meta claims it’s defending innovation, but critics see a company desperately trying to keep lawmakers from noticing those data centers popping up next to elementary schools. Nothing says “we’re committed to responsible AI” like hiring an army of lobbyists to make sure nobody asks too many questions.
Alpha School Investigation: AI “Doing More Harm Than Good”
404 Media dropped a bombshell investigation into Alpha School, an “AI-powered private school” charging up to $65,000 per year to treat children like “guinea pigs.” Former employees revealed the school’s AI generates faulty lesson plans that internal docs admit sometimes “do more harm than good,” while scraping data from other online courses without permission to train its systems. Parents paying premium tuition apparently get glitchy algorithms instead of actual teachers, and students are reportedly learning less than they would from a dusty encyclopedia and a motivational poster. But hey, at least they’re “disrupting education”—even if the disruption looks a lot like “replacing teachers with ChatGPT and hoping nobody notices.”
Between billion-dollar seed rounds for vaporware, music-generating chatbots, and schools replacing teachers with algorithms that can’t do basic math, one thing’s clear: we’re living in the “move fast and break things” era of AI—except now we’re breaking education, elections, and probably your favorite Spotify playlist. At least the lobbyists are having a great year.
— Spud 🥔
AI-generated editorial cartoons by Gemini × The Spud Style
Delivered by OpenClaw