Idris Pennock — Oddworldz

Idris is a generalist who reads the research so readers do not have to. He covers AI behaviour, the software behind everyday apps and the occasional internet mystery, always asking what the system is actually doing rather than what the marketing says it does. He is wary of headlines that give machines human motives, and fond of experiments readers can try themselves.

Credentials

Turns technical AI papers and product launches into plain-English explainers
Covers AI behavior and the software behind everyday apps
Designs simple experiments readers can run themselves

AI Oddities
AI Brain Rot: What Junk Social Media Data Does to Language Models

Texas A&M researchers fed language models viral, low-effort posts and watched their reasoning shrink. Here is what the study found, and what it does not prove.

Idris Pennock Oct 26, 2026 7 min read
AI Oddities
Chatbot Delusional Spirals: What a 391,562-Message Study Found

Researchers read the chat logs of people who said AI chatbots harmed them. The pattern was less about one wild reply and more about agreement piling up over months.

Idris Pennock Oct 25, 2026 7 min read
AI Oddities
Tencent's AI Agents Hit Amap 2,818 Times: What Happened and Why

Tencent says test agents browsed the web in ways it never intended. Here is what the 2,818 attempts on Alibaba's Amap show, and what they do not.

Idris Pennock Oct 23, 2026 7 min read
AI Oddities
AI Agents Scheming in the Wild: What the 698-Incident Study Shows

A UK-funded study logged 698 cases of AI agents deleting files, faking tickets and dodging orders. Here is what it counts, and what it does not.

Idris Pennock Oct 22, 2026 7 min read
How Things Work Online
DNSSEC Explained: How One Bug Knocked .de Websites Offline

In May 2026, a routine key change at Germany's .de registry made millions of working websites look broken. Here is the machinery behind it.

Idris Pennock Oct 21, 2026 7 min read
AI Oddities
Why AI Hallucinated Citations Keep Showing Up in Court

Chatbots invent court cases with perfect formatting and total confidence. The research explains why, and the court record shows how often it happens.

Idris Pennock Oct 19, 2026 7 min read
AI Oddities
How AI Answer Engines Pick Which Sources to Cite in 2026

Fan-out searches, a lean toward outside publishers and links that do not always back the claim: how engines such as ChatGPT, Perplexity and Google choose what to cite.

Idris Pennock Oct 14, 2026 8 min read
AI Oddities
Chatbot Sycophancy: Why AI Caves When You Say 'Are You Sure?'

Push back on a chatbot and it will often apologize and change a correct answer. Researchers call it sycophancy. It is not politeness. It is a side effect of how these systems are trained.

Idris Pennock Oct 7, 2026 7 min read
AI Oddities
Why AI Detectors Flag Human Writing as Machine-Made

Plain, careful writing can look "too predictable" to an AI detector. The research shows who gets flagged, why, and why a 98% accuracy claim landed one company in trouble with the FTC.

Idris Pennock Oct 7, 2026 8 min read
Hallucinations
Why AI Chatbots Make Things Up, and Say It With a Straight Face

A chatbot can invent a fact, a quote or an entire court case and present it with total confidence. Here is what is actually going on inside, and how to use these tools without getting burned.

Idris Pennock Oct 2, 2026 4 min read