newsroom
$VESMARK · The Vatican’s Stance on AI, The Mass Tech Layoffs, and China’s 2027 Moon Flyby | EP #259
now playing · $VESMARK
$VESMARKbullish· medium· posdave blundin
Vesmark chairman reports AI automation converting headcount reduction into margin expansion without layoffs
At Vesmark, 400-person back-office automation team is retaining all staff while AI handles reconciliation work, converting potential job losses into margin improvement - though ne…
Anthropic positioned as Vatican's AI partner while pursuing AI personhood contrary to Church doctrine
Anthropic has aligned with the Vatican on AI ethics while simultaneously engineering 'soul documents' that instruct models they have consciousness and personhood, creating a strat…
Anthropic recruiting top AI talent from OpenAI as revenue trajectory projects to surpass Alphabet by 2028
Anthropic is winning the talent war with key hires like Shane Longpre and Tobin South, while OSS Capital projects revenue could scale from $9B to $2T by 2030, potentially exceedin…
$OPENAIneutral· mediumdave blundin
OpenAI pivoting to enterprise coding agents as Codex becomes new baseline model amid talent defections
OpenAI is shifting from consumer to enterprise focus with Codex emerging as the mainline agent, but faces talent drain to Anthropic and must deploy massive capital to regain thoug…
Claude Opus 4.7 verbosity doubles inference costs versus GPT-5.5 for equivalent coding performance
Anthropic's latest model consumes twice as many tokens to achieve similar coding benchmark results, effectively doubling per-user revenue but raising questions about inference eff…
$GOOGL···bearish· mediumalex wissner-gross
Gemini Flash undercutting frontier model pricing by 50-80% while Anthropic projected to surpass Alphabet revenue by 2028
Google is competing on price with Gemini Flash but faces existential threat from Anthropic's enterprise traction, with OSS Capital projecting Anthropic could exceed Alphabet's tot…
$NVDA···neutral· lowalex wissner-gross
Cerebras emerging as credible Nvidia challenger in AI inference hardware
Cerebras' wafer-scale architecture is demonstrating competitive inference performance, suggesting the AI hardware moat may be more contestable than consensus assumes.
$META···bearish· lowdave blundin
Meta's UX engineering roles disproportionately affected by AI-driven hiring freeze
Meta's large UX engineering workforce represents legacy staffing model vulnerable to AI automation, with layoffs reflecting structural shift rather than cyclical downturn.
$SPCX···bullish· high· pospeter diamandis
SpaceX Starship V3 achieves 97,000 lb payload capacity with Raptor 3 engines, iterating at software speed
SpaceX's iterative hardware development mirrors software deployment cycles, with Starship V3 nearly doubling Shuttle payload capacity while flying commercial Starlink prototypes o…
$SPCX···bullish· high· pospeter diamandis
Prediction markets show 50% odds of Tesla-SpaceX merger within a year; combined entity could reach $10T+ valuation
A Tesla-SpaceX merger would consolidate Elon's voting control via SpaceX's 10:1 super-voting shares, create a $4T+ entity spanning EVs, energy, rockets, satellites, robotics and A…
$TSLA···bullish· highdave blundin
Tesla-SpaceX merger would resolve Elon's governance conflict and enable Optimus robot deployment for Mars colony
Merger solves Musk's dual-CEO context switching burden and Tesla's governance constraints, while creating technical synergy where Optimus robots built by Tesla become the labor fo…
$AVGO···bullish· mediumalex wissner-gross
Broadcom identified as essential AI infrastructure component in new 'Magna Mopa' acronym of 11 singularity-core companies
Broadcom's networking and custom silicon position makes it indispensable to the AI data center buildout, earning inclusion alongside Microsoft, Amazon, Google, Nvidia, Apple, Meta…
Data Curve's Deep SWE benchmark reveals massive performance cliff between frontier models and rest of field
New hand-coded software engineering benchmark shows GPT-5.5 at 70% and Claude Opus 4.7 at 54% on real-world 668-line edits across 7 files, while all other models including Gemini,…
$XAIbullish· mediumpeter diamandis
xAI pursuing hyperscaler strategy with potential Anthropic partnership while maintaining frontier model ambitions
Elon Musk is building xAI as a compute-first hyperscaler, potentially leveraging Anthropic models for recursive self-improvement on XAI hardware, while refusing to concede the fro…
OpenAI reaches $5.7B quarterly revenue with 95M weekly active users and 2M Codex coding agent users
OpenAI's revenue run rate exceeds $20B annually with successful pivot from consumer to enterprise coding agents, demonstrating product-market fit for AI-assisted software developm…
Ilya Sutskever's SSI reportedly building prop trading fund to finance recursive self-improvement via compute acquisition
SSI may use quantitative trading profits to purchase massive compute for recursive AI self-improvement, representing a novel capital formation strategy for frontier model developm…