misaligned bits

misaligned bits is our (roughly) weekly newsletter with bits and news, recaps from articles we published and latest studies in the field.

Newsletter Archive

April 9, 2026
misaligned bits #20: Bad For Your Health

AI is swallowing poison, AGI scares are back on the marketing menu, copyright questions again and LLMs are bad for your health.

April 2, 2026
misaligned bits #19: Flatter Me

Everyone is in court, chatbots are plotting, ghosts in machines, social media made addictive and bots want to flatter.

March 26, 2026
misaligned bits #18: AI Makes Us Think Alike

OpenAI is cuts corners, Lords are siding with creatives, Oracle scales down while others scale up, and AI makes scientists think the same

March 9, 2026
misaligned bits #17: Privacy? What privacy?

Pub Crawl, update on the baddies, Microslop, data centres in bad places, chatbots and privacy.

March 6, 2026
misaligned bits #16: Pseudonymously

The summit has peaked, no copyright for AI, standing up for science, and LLMs are bad for your privacy.

February 28, 2026
misaligned bits #15: In Your Face

AI overrated, Zuck in court, how to train your human, facial dystopia, weapons may hurt, AWS deletes itself

February 19, 2026
misaligned bits #14: Palantir Extravaganza

Palantir scrutinized, the DSA is indeed enforceable, TikTok addiction, and a Marxist perspective on recommender systems.

February 6, 2026
misaligned bits #13: Ouroboros

Generative AI eats itself, Anthropic bits, Grok v the world, AI washing, everyone is nervous, and who is in charge?

January 31, 2026
misaligned bits #12: Citation Needed

Everyone against Grok, US states want to regulate AI despite Trump, ChatGPT is biased but wants you to go shopping, citations desperately needed.

January 5, 2026
misaligned bits #11: Preparing For Impact

Insurers are unsure about AI. Grok has another lapse. And LLMs are seriously biased against people who speak dialect.

December 30, 2025
misaligned bits #10: Rather Inaccurate

AI rots its own brain. Falling numbers. Generative AI is a distraction. AI is only role-playing as scientists, otherwise missing the point.

October 22, 2025
misaligned bits #9: Subprime AI

GPT-5 unsafe and not so clever, subpoenas for everyone, who wants to pay for AI anyway, and poisoning LLMs might require little data.

October 16, 2025
misaligned bits #8: Slop Wave Coming

A tsunami of AI slop is approaching, friend has no friends in New York, science is in real trouble. Also, AI makes people dishonest.

September 30, 2025
misaligned bits #7: Circular Gigantomania

This week: Is the AI investment hype going in circles?

September 23, 2025
misaligned bits #6: A New Standard in Town

OpenAI's job platform, copyright wars and Americans see AI as risky.

September 9, 2025
misaligned bits #5: Is the Copyright Dust Settling?

This week with another episode in the generative AI vs copyright saga.

September 1, 2025
misaligned bits #4 : A Super PAC Network For Positivity?

Silicon Valley has now a “network of Super PACs with some of the usual suspects from the AI industry.

August 22, 2025
misaligned bits #3 (August 2025)

Regulatory bits. The copyright saga continues. And bits to read.

August 16, 2025
misaligned bits #2 (August 2025)

Voluntary code of practice. Tesla deletes data when the data is needed. And a lot of bits from the medical field.

August 6, 2025
misaligned bits #1 (August 2025)

Obligations in Europe. Altman warns, again. AN American action plan for AI.