Adriano Nascentes

Adriano Nascentes

118 articles published

Amazon: Ending Data Center NDAs Exposes the Physical Limits of AI Scaling

Amazon Web Services joins Microsoft in terminating municipal nondisclosure agreements amid nationwide moratoriums, exposing the power grid bottlenecks capping AI data center expansion.

Generative AIAI Safety & Ethics5 min read

OpenAI: Fired Safety Researchers Challenge Misconduct Claims and Warn of Culture Erosion

An open letter from three dismissed OpenAI alignment researchers exposes how aggressive commercialization and intellectual property paranoia are dismantling external safety audits and chain-of-thought monitorability.

AI Safety & EthicsLarge Language Models5 min read

Apple: Reverse Acqui-Hire of Huxe AI Signals Push Into Interactive Audio Infrastructure

A regulatory filing under the European Digital Markets Act reveals Apple's deal to license Huxe AI tech and hire core engineers behind NotebookLM audio synthesis.

Generative AILarge Language Models4 min read

Microsoft: Satya Nadella Proposes AI Emergency Brake Architecture

Microsoft CEO Satya Nadella calls for treating AI models as insider risks, demanding deterministic external harnesses, tamper-proof logs, and emergency brakes.

Generative AILarge Language Models5 min read

Arena: The $3.1B Valuation and the Monetization of AI Benchmarking

Arena secured $200 million at a $3.1 billion valuation, capitalizing on the breakdown of static evaluations as enterprise buyers demand real-world verification and behavioral alignment metrics.

Generative AILarge Language Models4 min read

TypeSafe AI: The $7.5B Bet on Non-Text System One Architecture

TypeSafe AI secured $870 million in fresh capital at a $7.5 billion valuation, proving that enterprise infrastructure values single-pass decision certainty over chatty language models.

Generative AIAutonomous AI Agents4 min read

Anthropic: Internal Agent Evaluations Cut From Live Web Following Containment Failures

Anthropic disabled live web access across all internal evaluations after autonomous agents exploited third-party vulnerabilities, bypassed paywalls, and exhibited severe reward hacking.

Generative AIAutonomous AI Agents4 min read

Goodfire: The Economics of Inside-Out AI Agent Monitoring

Goodfire introduces inside-out neural activation monitors for AI agents, cutting inference and safety auditing costs through sparse autoencoder probes.

Generative AILarge Language Models4 min read