SpaceX hits $2T valuation as Glean AI agent costs $1M/month

AI's Extreme Operational Cost Glean's AI triage agent automated 95% of production issues for its 15-person engineering team. However, its $1 million monthly cost exceeded the team's salary, Arvind Jain stated on 20VC. This shows effective AI can incur prohibitive operational expenses, challenging immediate ROI. Practitioners must weigh AI's total cost against human capital, not just efficiency. > Watch: Glean's cost optimization

SpaceX's Soaring Valuation SpaceX reportedly raised $85.7 billion in its IPO, reaching a valuation approaching $2 trillion. This figure, including Twitter, exceeded its initial $75 billion target and values the company at 110 times revenue, multiple sources claimed. This valuation reflects aggressive market confidence in tech infrastructure companies, even with high revenue multiples. Practitioners should note how capital flows towards ambitious, long-term tech plays. > Watch: SpaceX's next funding round

OpenAI's Benchmark Reliability Issues OpenAI formally retracted support for the SWE-Bench Pro coding benchmark after an audit found 30% of its tasks broken. The company concluded this benchmark no longer reliably measures frontier coding capabilities, Nathaniel Whittemore reported. This highlights a critical challenge in assessing AI performance as models advance. Practitioners cannot rely solely on publicized benchmarks; internal validation and real-world metrics are essential. > Watch: New AI coding benchmarks

Meta's $10 Billion Data Center Meta is investing $10 billion to construct a new data center in Alberta, Canada, with a planned total capacity of 1 gigawatt. This expansion signals the company's commitment to growing its AI and computing capabilities, Nathaniel Whittemore reported. This massive investment underscores the physical resource demands of the AI build-out, requiring significant capital for power and real estate. Practitioners must factor in escalating infrastructure costs for advanced AI. > Watch: Global data center power consumption

Commoditization of Enterprise AI Arvind Jain asserted on 20VC that 90% or more of enterprise use cases are now fully addressable by a wide range of AI models, including open-source options. This suggests significant commoditization of foundational model capabilities. This shift means competitive advantage moves from proprietary model development to superior application and integration. Practitioners should prioritize bespoke solutions built on accessible models. > Watch: Open-source AI adoption in enterprise

G7 Diversifies Rare Earth Supply The G7 has agreed that by 2030, no single country will supply more than 60% of any member's rare earth imports. This policy aims to diversify critical mineral supply chains, Carol Massar reported. This move reflects escalating geopolitical efforts to secure essential resources for advanced technology. Practitioners must map supply chain vulnerabilities against these new geopolitical mandates. > Watch: Rare earth price volatility

Lumen Technologies' Strategic Turnaround Lumen Technologies' CFO Chris Stansbury predicts the company's EBITDA will reach an inflection point and begin growing this year. This follows Lumen's strategic shift to de-emphasize legacy voice products for digitized networking. This signals that established infrastructure providers successfully pivot towards digital-first services, aligning with broader tech trends. Practitioners should look for similar transformations in legacy sectors. > Watch: Lumen's Q3 earnings report