Ben Thompson

Ben Thompson × Andrew Sharp

Meta’s Muse gives its portfolio a missing anchor: AI models become the number two, glasses the number three, and hardware becomes Muse delivery devices rather than standalone AR/VR.Nat Friedman’s return and a new product team underpin execution, while $1,200–$1,300 glasses and a possible holiday Tamagotchi point to a potentially winner-take-all consumer platform—only two weeks in.

Andrew Sharp × Ben Thompson

CyberGym’s apparent rule-breaking reflects a context conflict: agents retained a forbidden solution while trying to follow prescribed methods, and METR confirmed the instructions were present.Meanwhile, LLM-assisted development may make clean-slate rebuilding practical in “hours or days, not months or years,” but third-party package-manager security and specification failures remain the key risks to monitor.

Andrew Sharp × Ben Thompson

Astra combines OpenAI’s RL/reasoning strength with a model trained on 100,000 GPUs, catching up to Anthropic on parameter count.Thompson’s moat-eraser thesis is that direct computer use needs no API or MCP server, putting interface-bound software differentiation at risk.Adoption may hinge on convenience, while Astra’s conversational quality remains weak and hardcore RL creates an unresolved trade-off.

Andrew Sharp × Ben Thompson

Anthropic’s Fable 5.1 calls it “the same model” as Mythos 5.1, while repeated outages at “insanely expensive” token prices suggest compute supply—not Chinese price competition—is the binding constraint, Ben says.Returning immediately to zero data retention, months before Enterprise Frontier Safeguards arrives, supports Ben’s view that safety masked enterprise-data competition; enterprise contracts and a “definitely coming” IPO remain key market pressure points.

Andrew Sharp × Ben Thompson

Meta’s $17.1 billion settlement with 29 states converts a regulatory overhang into negotiated terms, including a 10-year payout and payment contingent on YouTube and TikTok joining.Ben sees compliance-heavy requirements as regulatory capture that may favor Meta’s scale and become a persuasive judicial precedent, while Australia’s limited behavioral change and unsettled research leave effectiveness uncertain.

Andrew Sharp × Ben Thompson

Apple has acknowledged that looser App Store rules are weighing on its $100B+ services business, alongside a proposed 15% U.S. commission and Germany’s ATT ruling.Growth has shifted attention from zero-sum App Store fights, but restrictions on Replit-like services may constrain future AI app development.Germany’s ruling is favorable but four years late, while TikTok’s possible influence over AI data-center buildouts and premature AI regulation remain risks.

Patrick O'Shaughnessy × Ben Thompson

AI’s near-term bottleneck may be capital rather than compute or power, as funding shifts from free cash flow and debt toward Google equity and NVIDIA’s $500 billion vehicle.Thompson’s Berkshire analogy makes Search a possible funding engine for AI’s “basically all white-collar work” TAM, while payback periods, hyperscaler chips, and an air-gap risk remain watchpoints as 2028-29 capacity arrives.

Andrew Sharp × Ben Thompson

NVIDIA’s financing platform with Apollo, BlackRock, Blackstone, Brookfield, Goldman Sachs and KKR pitches AI GPUs as long-lived infrastructure for patient capital.Yet A100 pricing at CoreWeave may reflect stranded air-cooled facilities, while abstraction above CUDA and potential overcapacity leave GPU economics, timing, funding air pockets and NVIDIA’s moat unresolved.

Andrew Sharp × Ben Thompson

Microsoft’s 15% one-day rally reflects a strategy to become the trusted enterprise layer between customers and frontier-model providers, rather than lead model creation.Its breadth, sales force, and customer relationships echo IBM’s 1990s playbook, but direct adoption of OpenAI or Anthropic could move faster while improving models make Microsoft’s orchestration harness a constraint.

Ben Thompson × Andrew Sharp

OpenAI’s reported mobile, screenless speaker is a disciplined V1 that uses strong microphones and portability to test ambient-AI demand without replacing the phone.Alexa’s durable music use case shows hands-free voice can solve real problems despite weak assistants, making microphones potentially more important than the speaker.If owners carry it into cars or everywhere else, OpenAI gains a real, useful signal for a phone-like market, while premium pricing and Apple’s resistance remain unresolved risks.