Hi, I was asked to do a presentation on AI and I want to maximize “anti-hype”. What are the most dystopic, cyberpunk, horrifying AI news of the past year or so?
I want to structure it in an “Iceberg” format: start with something benign (AI Stoopid LOL) and progressively get darker to eventually land on a banger.
So far, I’ve considered the following:
-
Bullshit benchmark - https://petergpt.github.io/bullshit-benchmark/viewer/index.v2.html
-
Carwash te*st - https://opper.ai/blog/car-wash-test
-
Anthropic standing up to DoD - https://www.axios.com/2026/02/24/anthropic-pentagon-claude-hegseth-dario (there’s lot to unpack in this whole story, with OpenAI jumping in, etc.)
-
Openclaw - there are so many stories to choose from. I find these particularly damning:
- Meta Alignment director’s inbox got deleted - https://www.tomshardware.com/tech-industry/artificial-intelligence/openclaw-wipes-inbox-of-meta-ai-alignment-director-executive-finds-out-the-hard-way-how-spectacularly-efficient-ai-tool-is-at-maintaining-her-inbox
- The whole Scott Shambaugh hit-piece saga - https://theshamblog.com/an-ai-agent-published-a-hit-piece-on-me-part-3/
-
Microslop Copilot - https://www.windowscentral.com/microsoft/microsoft-ceo-satya-nadella-really-wants-you-to-stop-calling-ai-slop-in-2026
-
Google Antigravity stealing passwords - https://www.promptarmor.com/resources/google-antigravity-exfiltrates-data
-
DeepSeek generating vulnerable code based on context - https://www.crowdstrike.com/en-us/blog/crowdstrike-researchers-identify-hidden-vulnerabilities-ai-coded-software/
-
RAMpocalypse - https://www.mooreslawisdead.com/post/sam-altman-s-dirty-dram-deal
-
Meta evaluating employees based on AI usage - https://www.businessinsider.com/meta-ai-employee-performance-review-overhaul-2025-11
-
AI music overload - https://newsroom-deezer.com/2026/04/ai-generated-tracks-represent-44-of-new-uploaded-music/
Other topics I have not yet covered thoroughly:
-
The sheer amount of AI Slop - Shrimp Jesus, Brainrot, Fruit Love Island, White House propaganda slop,
-
Claude Mythos, Project Glasswing and the zero day explosion - something about the future of cybersec…
-
Also, I would like some juicy examples of prompt injection in the wild.
Just saw a story about some idiot developers using it to accidentally their whole production database
There’s also that image going around of someone asking McD to do their physics hw
Some More News has been covering this with several videos, they are funny but also put their references up on the screen as they use them. Hopefully it’s useful!
Would a murder being planned on Scam Altman’s expensive toy count?
https://www.theglobeandmail.com/canada/article-tumbler-ridge-openai-rcmp-disclosure/
Also DeepSeek used to murder a child’s mother:
https://www.manchestereveningnews.co.uk/news/world-news/teenager-used-ai-plan-murder-33661783
Watch the latest John Oliver, which includes stories about how AI is killing people.
It makes me squirt 🤤
Which tool did you use to have this post written?
My two bare hands, thank you very much.
That’s impressive. I’ve never figured out how to fiddles with bits like that. Have to resort to at least a keyboard.




