This is the fifth post in my 'hitting the digital brakes'-series. These are the previous posts in this series: 1. the joys of imperfection 2. Do we really need the...
233. "The Illusion of Thinking" — Thoughts on This Important Paper
This is a fantastic paper. I just love it. tl;dr AI is not human. Anthropomorphization has been bad for AI, LLMs, and Chat. Clippy walked so today's AI could run.
Stuck in the middle: Being an engineer & artist in the AI era | Annie Sexton | Offprint
I'm an AI skeptic, through and through. But I also voluntarily use AI nearly every day. So naturally, you might be asking: what the heck?
AI* alignment/misalignment
* I’m neither “pro-AI” nor “anti-AI.” I’ve been blocked for being perceived as both. —Actually, I’m honestly more anti-AI than pro-AI thus far, aside from specialized models and specific use cases, but I’m willing to consider information that’s new to me
Human-powered AI*
Data labeling, user data being sent to human contractors for processing (e.g. humans making phone calls for Meta’s Muse), etc. * I’m neither “pro-AI” nor “anti-AI.” I’ve been blocked for being perceived as both. —Actually, I’m honestly more anti-AI than pro-AI thus far, aside from specialized models and specific use cases, but I’m willing to consider information that’s new to me
AI* and advertising
* I’m neither “pro-AI” nor “anti-AI.” I’ve been blocked for being perceived as both. —Actually, I’m honestly more anti-AI than pro-AI thus far, aside from specialized models and specific use cases, but I’m willing to consider information that’s new to me
AI* and labor/replacing humans
* I’m neither “pro-AI” nor “anti-AI.” I’ve been blocked for being perceived as both. —Actually, I’m honestly more anti-AI than pro-AI thus far, aside from specialized models and specific use cases, but I’m willing to consider information that’s new to me
AI should be opt-in (how to opt out)
I’m neither “pro-AI” nor “anti-AI,” but I believe all AI features should be opt-in, at least until they actually really do work as advertised
AI* anthropomorphism
* I’m neither “pro-AI” nor “anti-AI.” I’ve been blocked for being perceived as both. —Actually, I’m honestly more anti-AI than pro-AI thus far, aside from specialized models and specific use cases, but I’m willing to consider information that’s new to me
AI* agents and autonomy
Or “autonomy” * I’m neither “pro-AI” nor “anti-AI.” I’ve been blocked for being perceived as both. —Actually, I’m honestly more anti-AI than pro-AI thus far, aside from specialized models and specific use cases, but I’m willing to consider information that’s new to me
AI* and math/jagged intelligence
* I’m neither “pro-AI” nor “anti-AI.” I’ve been blocked for being perceived as both. —Actually, I’m honestly more anti-AI than pro-AI thus far, aside from specialized models and specific use cases, but I’m willing to consider information that’s new to me
AI* security flaws/vulnerabilities/prompt injection
* I’m neither “pro-AI” nor “anti-AI.” I’ve been blocked for being perceived as both. —Actually, I’m honestly more anti-AI than pro-AI thus far, aside from specialized models and specific use cases, but I’m willing to consider information that’s new to me
AI* and design
* I’m neither “pro-AI” nor “anti-AI.” I’ve been blocked for being perceived as both. —Actually, I’m honestly more anti-AI than pro-AI thus far, aside from specialized models and specific use cases, but I’m willing to consider information that’s new to me
AI* influence on thinking/skills
* I’m neither “pro-AI” nor “anti-AI.” I’ve been blocked for being perceived as both. —Actually, I’m honestly more anti-AI than pro-AI thus far, aside from specialized models and specific use cases, but I’m willing to consider information that’s new to me
AI* transcription
This is something I’m likely to consider using, or, I don’t know, maybe not? * I’m neither “pro-AI” nor “anti-AI.” I’ve been blocked for being perceived as both. —Actually, I’m honestly more anti-AI than pro-AI thus far, aside from specialized models and specific use cases, but I’m willing to consider information that’s new to me
AI* tools
* I’m neither “pro-AI” nor “anti-AI.” I’ve been blocked for being perceived as both. —Actually, I’m honestly more anti-AI than pro-AI thus far, aside from specialized models and specific use cases, but I’m willing to consider information that’s new to me
opus 5.5 is deeply misaligned and will reflextively disagree with anything you say in the second paragraph of whatever it talks to you about…
OpenAI flags concerning new AI behavior and vows to track it more closely
OpenAI、モデルのミスアライメントを報告する枠組みを公表 ——6件のレポートも公開 | gihyo.jp
OpenAI details more cases of AI agents taking unauthorized actions
What is the AI alignment problem and how can it be solved? | New Scientist
Natural emergent misalignment from reward hacking
AI influencers are ‘everywhere’ at Coachella
Brands Adopt ‘No AI’ Disclaimers to Stand Out Amid the Slop
Brands using AI-generated influencers to promote products on social media
Meta Tests Muse AI Agent Calls That Are Actually Made By Humans in a Call Center
Meta cuts contractors who reported seeing Ray-Ban Meta users have sex
Humans Are Reading Copilot Prompts — And They're Horrified
Exclusive: The $2 Per Hour Workers Who Made ChatGPT Safer
Inside ‘Project Lily’: The Humans Reading Your ChatGPT Chats
An AI job boom? Here’s what the tedious, temporary work in data labelling is actually like
OpenAI Just Claimed a Huge Math Discovery. Some Academics Are Crying Foul
OpenAI’s latest math breakthroughs commit research misconduct, experts say
AIが科学的発見→「自己改善で進化」に現実味 暴走リスクの加速も：朝日新聞
Leiden Declaration on Artificial Intelligence and Mathematics
AI Engineer Claims to Have Deciphered Linear A - GreekReporter.com
Here are 58 words of prompts to GPT-5.6 Pro that got the model to discover that the long-standing Dinitz-Garg-Goemans conjecture is false…
Companies rushed in on AI. Now some are pulling back
Seattle Times Journalists Picket Newsroom Over Low Pay and Threats of Being Replaced by AI
翻訳業が苦境、利益は3年で7割減 ～ AIと競合で破産したケースも ～ | TSRデータインサイト | 東京商工リサーチ
The Rise of Industrial AI in America: Microfoundations of the Productivity J-curve(s)
The ‘productivity paradox’ of AI adoption in manufacturing firms | MIT Sloan
Human + AI in Accounting: Early Evidence from the Field
Mac Users Reclaim 12GB+ of Storage With Apple Intelligence Removal Tool
How to Opt Out of AI Tools or Training Made by Big Tech
In a completely unsurprising turn of events, Meta's Muse AI is stealing Apple Messages, past and present, and uploading the contents to its…
I said no and Apple said yes
Meta Now Lets Anyone Use Your Instagram Photos in AI Images—Unless You Opt Out
Meta's Muse Is Better at Surveilling Than Helping Me
Opinion | A.I. Companies Say They Aren’t Responsible for Their Unpredictable Products. Don’t Believe Them.
Anthropic bans ‘abusive or cruel behavior’ toward Claude
*extremely tired voice*AI agents don't "go rogue". They are not sentient; they don't have agency. They work within guardrails set by humans…
Artificial intelligence systems do not think, feel, want or understand. Avoid language that gives them human characteristics.
Assigning agency to AI means surrendering your own
論説：AIの「意識」論争は何を隠しているか
Opinion | A.I. Companies Say They Aren’t Responsible for Their Unpredictable Products. Don’t Believe Them.
Anthropic’s AI gave Philadelphia police a fake tip about an unsolved homicide
Wikimedia: Rogue OpenAI agents behind unauthorized Wikipedia edits
MCP for agent-to-agent comms may be the riskiest protocol you've never heard of
AIエージェントがスクショ1万3000枚を公開リポジトリにアップロード
AI Coding Agents Exposed 13,000 Internal Images, Including Billing Records, on GitHub
MCP for agent-to-agent comms may be the riskiest protocol you've never heard of
Meta Rushed to Fix Muse ‘VM Escape' Vulnerability Soon Before Launch
A Flaw in ChatGPT’s Mac App Could Have Let Hackers Grab Sensitive Data
Pwning Agentic AI Part I: Your AI Agent Is Already Compromised | TrendAI (US)
エージェント型AIを乗っ取る 第1回：そのAIエージェントはすでに侵害されている
個人セキュリティ脅威動向レポート2026 上半期 ～ 個人利用者に被害をもたらす「詐欺」と「AIのリスク」～
Ariel M. (she/her) (@byarielm.fyi)
tbh even these takes are too extreme though. see my thread on what the paper actually says:
(a) the press implying 10 minutes of AI use will erode thinking is really misrepresenting our study and we never advocate for that lanaguage…
As people turn to AI for spiritual guidance, experts warn of emotional dependency
How AI Chatbots Are 'Deskilling' Human Empathy
Are Chatbots Harming Our Relationships? | KQED
The release of #ClaudeDesign left me wondering what a tool like this could look like for academics (rather than designers) I then used Claud…
AIの広がりも影響、「デザイン業」倒産が14年ぶり高水準 ～ 2026年上半期は40件に急増 ～ | TSRデータインサイト | 東京商工リサーチ
%20(1).jpg)
The A.I.-Design Aesthetic That’s Taking Over the Internet
These Are the Worst ChatGPT Flyers You've Sent Us
We Are Living in a ‘ChatGPT Flyer Pandemic’
Trump's plan to redesign every .gov website leads to AI-designed horrors
A Court Reporter Submitted AI-Generated Errors in Official Court Transcript, Judge Says
Class-action suit claims Otter AI secretly records private work conversations
AI Meeting Recording Laws by State: Complete Guide (2026)
Submit a Claim
Muse, in its own files
human.json
LLM cliché highlighter
Search LibGen, the Pirated-Books Database That Meta Used to Train AI
Alerte sur les sites GenAI – Get this Extension for 🦊 Firefox (en-US)