sdlcnext.com

Cutting through AI
hype and hope.

Every post in three takes: skeptic, balanced, and all-in.

Recent Posts

View all →

Moonshots Ep. 258: The Organizational Singularity

Coase's law just broke. Salim Ismail's EXO 3.0 framework says the firm becomes a fiduciary wedge holding IP, data, agents, and the 20% of humans who do oversight.

moonshots organizational-design agentic-ai future-of-work exo
Read more →

Moonshots Ep. 257: The Two-and-a-Half-Lab Frontier

SpaceX files the largest IPO in history, hands its model lineage to Anthropic, and an OpenAI model disproves an 80-year-old Erdős conjecture. The frontier just collapsed to two and a half labs.

moonshots spacex-ipo anthropic ai-research organizational-singularity
Read more →

Moonshots Ep. 256: Money Can't Buy the Frontier

Cerebras IPO'd at $95B, Karpathy joined Anthropic, Google spent $190B in capex. Episode 256's actual lesson is that the moat is tacit knowledge, not capital.

moonshots ai-infrastructure semiconductors cerebras google-io
Read more →

Moonshots Ep. 255: The Singularity Economy Is a Compute Allocation Problem

Anthropic took over SpaceX's Memphis cluster, hit 80x quarterly growth, and is on track for a $1T ARR by mid-2027. Episode 255's throughline is that the next two years will be decided by who can route compute, capital, and licence into the same loop.

moonshots ai-economy anthropic spacex leopold-aschenbrenner compute
Read more →

Moonshots Ep. 254: Governance Catches Up to Capability

Alphabet posts a record quarter, the White House drafts a frontier-model executive order, and GPT-5.5 quietly matches Mythos at five times cheaper. Episode 254 reads as governance finally trying to catch the curve.

moonshots ai-policy ai-infrastructure frontier-models compute-economics
Read more →

Moonshots Ep. 253: The Moonshot Corporation Is the New Corporate Form

Elon's half-trillion-dollar SpaceX comp package isn't a pay story. It's the prototype for a third type of corporation that conventional governance can't reach.

moonshots corporate-governance exponential-organizations humanoid-robots agi
Read more →

Moonshots Ep. 248: Permission Is the Bottleneck Now

Capability shipped on time. What slipped is licence. Episode 248's actual throughline is that the AI race is now an institutions race.

moonshots ai-policy ai-infrastructure future-of-work personalized-medicine
Read more →

Moonshots Ep. 247: The Singularity Has a Plumbing Problem

$3 billion a day in AI investment, a CEO with no shares facing a $100 billion lawsuit, and a panel that puts white-collar replacement at 99%. The constraint on the singularity is no longer intelligence. It is the infrastructure underneath.

openai anthropic moonshots ai-governance musk-vs-altman ai-jobs
Read more →

Moonshots Ep. 246: Anthropic Held Back Mythos. That's the Story.

Anthropic's strongest model didn't ship. Polymarket crashed from 80% to 7%. The reason rewrites the next twelve months of AI competition more than any release would have.

ai-safety anthropic moonshots claude frontier-models cybersecurity
Read more →