Hot AI summer isn’t slowing down. Even as GPT-5 rumors swirl in a never-ending game of “any day now,” Google dropped something real: Gemini 2.5 Deep Think, its most advanced reasoning model to date. Deep Think is Google's first multi-agent model. Instead of one model following a single thread, Deep Think spawns multiple agents to explore different reasoning paths in parallel - a computationally expensive but logic-rich approach. A variant of this model famously won a gold medal at the International Math Olympiad. The full-strength version, which “takes hours to reason”, is being released to a handful of academics. The consumer-facing version still sets a new bar on Humanity’s Last Exam: 34.8% for Gemini 2.5 Deep Think vs 25.4% for xAI's Grok 4 and 20.3% for OpenAI’s o3. Gemini’s technical depth is impressive, but as always with Google, the gap between cutting-edge research and productization remains the strategic question. Meanwhile, when he’s not vagueposting about GPT‑5 being so powerful it scares him, Sam Altman is busy overseeing OpenAI’s rocketship of a P&L. – ARR jumped from $10B in June to $13B, pacing toward $20B by year-end – 700M weekly active users, up from 500M in March – 5M business subs – $8B annual burn – $300B valuation, with $8.3B raised months ahead of schedule, 5x oversubscribed But beneath the growth metrics lies the most dramatic subplot: OpenAI and Microsoft are locked in a high-stakes standoff over a clause no one thought would ever matter. Under their current agreement, Microsoft loses access the moment OpenAI achieves “AGI” - which no one can define but everyone is racing toward. With AGI suddenly feeling less hypothetical, Microsoft is now blocking OpenAI’s for-profit conversion until the clause is rewritten. The proposed trade: Microsoft keeps post-AGI access in exchange for a 30–35% equity stake, more board seats, less revenue share, binding safety terms, and greater autonomy for OpenAI. A throwaway line from 2019 has become the fault line for one of the most consequential partnerships in tech- a sign of how quickly the tech (and the stakes) are accelerating. And then there’s Anthropic, the quiet compounder. ARR hit $5B, up from $4B - earlier this month and they’re in talks to raise $5B at a $170B valuation, a stunning figure for a 4-year-old company that came up from behind in the model race. Anthropic has taken the lead in one of the few commercially proven GenAI use cases: code generation. – Claude Code is generating $400M ARR, up 2x in weeks – Top two customers (likely Cursor + GitHub Copilot) drive $1.2B – Even excluding them, revenue is 11x YoY - Unlike OpenAI, Anthropic is almost disciplined: 60% gross margins vs OpenAI’s ~40%. Anthropic is selling investors a story of efficiency, not just scale. All 3 labs are still pointing at the AGI endgame. But that horizon is fuzzy and they're scaling it in different ways.