Giovanni Dicanio

mentions 2 type Person feed RSS

// recent coverage 2 mentions

16:20

2026-08-18

giodicanio.com

large-language-models

Why We Should Double-Check the AI Output: Claude Lies (AKA AI Hallucinations)

Anthropic's Claude chatbot incorrectly claimed that Google's C++ Style Guide recommends #pragma once over #define guards, a factual error that remains uncorrected as of August 18, 2026, according to t…

16:42

2026-07-28

giodicanio.com

artificial-intelligence

We Cannot 100% Trust AI-Generated Output: A Concrete Example

A developer found that Claude, an AI assistant from Anthropic, incorrectly flagged zero-length bugs in the WinReg C++ library's string getters, failing to account for the RegGetValueW API's null-termi…

// co-occurs with top 8 entities

Anthropic 2 Claude 2 Google 1 Gemini 1 WinReg C++ library 1 Windows Registry 1 RegGetValueW 1 Windows 11 1

// topics top 4 topics

ai safety 2 large language models 1 artificial intelligence 1 ai tools 1