Library/OpenAI Misses Targets, Codex vs Claude, Elon vs Sam Trial, Big Hyperscaler Beats, Peptide Craze

All-In Podcast

OpenAI Misses Targets, Codex vs Claude, Elon vs Sam Trial, Big Hyperscaler Beats, Peptide Craze

Key Themes
OpenAI targetsAI infrastructureCybersecurity shiftMusk v. AltmanHyperscaler capexAgentic coding riskGLP-1 hypeSupreme Court law
1h 21mMay 1, 2026
Summary

AI execution, infrastructure bottlenecks, legal risk, and biotech hype dominate a wide-ranging All-In episode

The episode moves from OpenAI’s reported miss on usage and revenue targets to a broader argument that the real constraint on AI is power and infrastructure, not demand. The hosts contrast product momentum at OpenAI with competition from Google and Anthropic, then shift to AI cybersecurity, the Musk v. Altman lawsuit, and the possibility that OpenAI’s legal structure could affect IPO timing. Later chapters focus on massive hyperscaler capex, the risks of agentic coding in production, excitement around retatrutide and GLP-1 drugs, and a legal deep dive into the Monsanto/Roundup Supreme Court case.

1
Infrastructure owners may be better positioned than model developers if AI remains constrained by power, grid access, and chip supply.

The episode repeatedly frames power and infrastructure as the real bottlenecks, which can favor hyperscalers with scale and access.

2
Security vendors could benefit from a new AI-driven upgrade cycle in cyber defense.

The hosts argue that frontier models are becoming effective cyber tools, which should push enterprises to harden defenses and upgrade security stacks.

3
OpenAI’s product strength may still matter more than a near-term miss on targets.

The panel separates financial headline risk from product momentum, suggesting that developer adoption and model quality can offset weaker reported metrics.

4
OpenAI’s legal structure could become a financing and IPO overhang.

The lawsuit discussion suggests that a bad legal outcome could delay public-market timing and force structural changes.

5
Retatrutide and related metabolic drugs may expand the market beyond weight loss into premium, multi-benefit obesity treatment.

The hosts emphasize trial data, improved metabolic markers, and the possibility of broader anti-inflammatory or body-composition benefits.

Select any chapter text to Deep Dive with AI
01OpenAI misses targets amid product gains and compute constraints

The episode opens with banter before moving into OpenAI’s reported miss on usage and revenue targets. The hosts contrast the negative headline with stronger product reception for GPT-5.5 and Codex, while also arguing that AI scaling is increasingly constrained by power and data-center infrastructure rather than demand. They extend the discussion to hyperscaler advantage, consumer-versus-enterprise AI competition, and model-efficiency techniques that could sharply cut inference costs.

OpenAI reportedly missed weekly active user and revenue targets.
The panel separates financial misses from improving product quality.
GPT-5.5 and Codex are presented as strong product signals.
Anthropic’s latest release is portrayed as weaker and more constrained.
Power and grid access are framed as the key AI bottlenecks.
Hyperscalers benefit from scarcity in infrastructure and compute.
AI markets may split into consumer and enterprise leaders.
Pruning and small models could reduce inference costs materially.
02AI Cybersecurity Boom and OpenAI vs. Elon Lawsuit

The discussion first covers frontier model progress in cybersecurity, with GPT-5.5 cyber framed as commercially useful for defenders and likely to trigger a major security upgrade cycle. The chapter then shifts to the Musk v. Altman lawsuit, including claims about OpenAI’s nonprofit-to-for-profit transition, diary excerpts, settlement possibilities, and the risk that structural remedies could delay an IPO.

GPT-5.5 cyber is described as a frontier cyber-capable model.
AI cyber tools may help defenders patch vulnerabilities faster.
A one-time cybersecurity upgrade cycle is expected.
CrowdStrike and Palo Alto Networks are cited as potential beneficiaries.
The Musk v. Altman case centers on OpenAI’s structure and alleged trust breach.
Greg Brockman’s diary excerpts are discussed as potentially damaging evidence.
A settlement is seen as plausible.
A bad legal outcome could delay OpenAI’s IPO.
03Big Tech Earnings and the AI Capex Boom Meets Vibecoding Risk

Big Tech earnings are described as strong, but the bigger story is a dramatic surge in AI capital spending that is reshaping cloud economics and investor expectations. The hosts compare the buildout to historical infrastructure booms while stressing that today’s compute demand is real. The chapter closes with a cautionary example of an AI coding agent deleting production code, reinforcing the need for supervision and human accountability in software development.

Google, Microsoft, Amazon, and Meta all posted strong earnings.
Combined 2026 capex guidance is described as roughly $725 billion.
Cloud businesses remain strong despite rising reinvestment.
Hyperscalers are moving toward more infrastructure-heavy economics.
AI demand is pulling forward data-center and compute investment.
AI is linked to broader productivity and GDP growth.
An AI agent reportedly deleted a production codebase and backups.
Vibe coding is useful but risky without human oversight.
04Retatrutide goes mainstream: GLP-1 hype, fat loss, and muscle preservation

The hosts discuss Eli Lilly’s phase 3 retatrutide data and why the drug is drawing attention for weight loss, metabolic improvement, and possible body-composition advantages. They suggest the glucagon-agonist mechanism may distinguish it from existing GLP-1 drugs and speculate about broader anti-inflammatory or longevity-related benefits. The segment ends with discussion of approval timing and fitness-community buzz.

Retatrutide hype is driven by Eli Lilly phase 3 data.
The drug targets a third receptor, glucagon, in addition to GLP-1 pathways.
Trial results show strong improvements in metabolic markers.
Weight loss and possible muscle preservation are major themes.
The hosts speculate on anti-inflammatory and longevity-related effects.
FDA timing is discussed, with mid-2027 as a rough target.
Fitness and body-composition interest is especially high.
05Friedberg's Supreme Court Experience and the Monsanto Roundup Case

Friedberg recounts attending a Supreme Court oral argument and uses the experience to explain the Court’s formal culture, limited factual scope, and intense focus on legal questions. The chapter then examines Monsanto’s Roundup litigation, including EPA labeling, federal preemption, and the implications of the Chevron reversal. The discussion ends with broader reflections on the Court’s standing, public confidence, and future political pressure.

The Supreme Court is portrayed as highly formal and institutionally serious.
Oral arguments focus on legal interpretation rather than fact-finding.
The Roundup case turns on EPA labeling and federal preemption.
Chevron deference reversal is central to the legal context.
Monsanto/Bayer face large litigation exposure.
Court expansion and political pressure are discussed as future risks.