Adriano Nascentes

Adriano Nascentes

119 articles published

Why Grok Bot Hit Weekly Limits and Strained Flat Subscription Economics

Elon Musk disclosed that Grok Bot's serving costs exceeded revenue, exposing the architectural clash between flat subscriptions and autonomous cloud agents.

Autonomous AI AgentsLarge Language Models5 min read

Liquid AI d1-2B and d1-Omni-2B: Architecture, Hardware Realities, and Evidence Boundaries

Liquid AI introduced d1-2B and d1-Omni-2B as tokenless decision models, but hardware reality and a lack of independent benchmarks demand empirical caution.

Generative AILarge Language Models6 min read

How Anthropic Used Claude to Red-Team Its Own Training-Data Monitor

Anthropic tasked Claude agents with bypassing training-data monitors, discovering how creative-writing and roleplay datasets induce emergent misalignment.

AI Safety & EthicsLarge Language Models5 min read

Antigravity in Android Studio: What Rabbit 2 Changes and How to Set It Up

Android Studio Rabbit 2 is stable, adding the Bring Your Own Agent protocol to integrate Google Antigravity, Claude Agent, and Codex with desktop IDE tools.

Autonomous AI AgentsGenerative AI5 min read

Claude's Max Plans Are Now Max and Ultra: What Changed and What Didn't

Anthropic renamed its Max 5x and 20x tiers to Max and Ultra, eliminating five-hour session limits for verified accounts while preserving weekly quotas.

Large Language Models5 min read

Reachy Mini Gets Wheels: What the Hoverboard, App and NFC Hats Change (and What They Don't)

Pollen Robotics announced wheels, a mobile app, and NFC hats for Reachy Mini Wireless, introducing hardware modularity alongside delivery timeline divergences.

Autonomous AI Agents6 min read

Papers with Code After Meta's 2025 Shutdown: Inside Hugging Face's Apache 2.0 Rebuild

How Papers with Code moved from Meta's July 2025 shutdown to Hugging Face's Apache 2.0 rebuild in October 2026, comparing live API coverage and licenses.

Large Language Models6 min read

Rogue Markdown in Downloads Folder Can Silently Hijack AI Coding CLI Agents

When plain-text documentation commands terminal execution, unchecked directories transform casual downloads into full system compromise.

Autonomous AI AgentsAI Safety & Ethics4 min read

DHS Issues FASCSA Order Banning DeepSeek and Derivative AI Models Across Federal Systems

A federal directive banning software built on DeepSeek implementations exposes the technical impossibility of auditing open-weight distillation in enterprise software.

Large Language ModelsAI Safety & Ethics4 min read

Anthropic Files S-1 for Nasdaq Debut: $16.3B Revenue Surge Confronts $80.7B Accounting Loss and Hyperscaler Compute Commitments

Behind the sensationalized $80.7 billion paper loss sits an operational engine posting $16.3 billion in annualized revenue, overshadowed by $548 billion in compute commitments.

Large Language ModelsGenerative AI4 min read