Forward Guidance hosts say a durable bottom likely needs a slowdown and then a policy pivot
The hosts frame the Anthropic IPO as a heavy supply event into a fragile AI trade and argue rising token usage has not translated into proportional spend.
The hosts frame the Anthropic IPO as a heavy supply event into a fragile AI trade and argue rising token usage has not translated into proportional spend.
Boockvar argues the Fed may be leaning too hawkish just as AI capital spending growth is set to decelerate, in a market heavily dependent on data-center and hyperscaler capex.
The teardown describes a CPU die of roughly 168 mm² paired with a GPU die of about 157 mm² or 300 mm², all on TSMC's N3P, joined by TSMC SoIC hybrid bonding on a silicon interposer.
An AI-native marketing automation company builds with frontier closed models, then moves repeatable tasks to open-weight models fine-tuned on proprietary or customer data to improve cost per completed task and vendor leverage.
A venture investor who backed both Anthropic and OpenAI at early stages describes the foundation-model opportunity as still early, arguing that most value accrues at the frontier of capability and that several players can coexist given the size of the addressable market. He says his firm sold nothing personally, distributed shares to its own investors, and that he added to his position around the public offering even as the stock traded below its offer price. He frames Anthropic's draw as unusual founder retention, a security-first worldview, and enterprise traction, while calling OpenAI's consumer position strong. He also treats the leadership turmoil at Nvidia as a temporary disruption rather than a structural break for AI infrastructure. Frontier AI labs remain early-inning, long-duration bets where a differentiated team and enterprise trust can compound value even in a multi-winner…
No publications match your search.