frontpage.
newsnewestaskshowjobs

Open Source @Github

fp.

Open in hackernews

Google AI Studio "Delete" doesn't delete – server chat persists (3 videos)

https://www.youtube.com/watch?v=LBQfNbPCxQw
3•Bitu79•47m ago

Comments

Bitu79•47m ago
I found that Google AI Studio’s “Delete” button does not delete the server‑side chat.

I tested it by deleting a prompt, deleting the JSON file from Drive, and emptying the Trash. The file was fully gone.

Then I used Google’s official recovery tool.

The JSON came back — and the model instantly continued the conversation with full context.

This is only possible if the server‑side session and conversation were never deleted.

I recorded the entire process in 3 videos:

https://youtu.be/LBQfNbPCxQw

https://youtu.be/m4U6ajKuLRo

https://youtu.be/8Pg-7eNPUWM

I also posted this finding under two official Google videos featuring Guillaume Vernade, so Google is aware of it: – https://www.youtube.com/watch?v=BcWFc3H7Khg – https://www.youtube.com/watch?v=LBQfNbPCxQw

Raises questions about data retention, transparency, and GDPR compliance.

Bitu79•45m ago
https://discuss.ai.google.dev/t/please-help-do-the-ai-studio...
Bitu79•11m ago
https://discuss.ai.google.dev/t/google-illegally-retains-cus...

Art as Compression

https://feed.grantcuster.com/post/2026-07-29-10-20-08-art-as-compression
1•msephton•5m ago•0 comments

Higher Airfares Are Looming on Busy Routes as AI Squeezes Out Bargains

https://www.bloomberg.com/news/articles/2026-07-29/higher-airfares-loom-on-busy-routes-as-ai-sque...
1•petethomas•5m ago•0 comments

xAI sues Minnesota over its law banning 'nudification' technology

https://apnews.com/article/minnesota-artificial-intelligence-nudification-x-elon-musk-deepfake-13...
4•devonnull•10m ago•1 comments

Talk to Your Area (Askaround.me)

1•SahibzadaAM•18m ago•0 comments

Show HN: Legionlinuxtui – Control Lenovo legion laptops in the terminal

https://github.com/nooneknowspeter/legionlinuxtui
2•nooneknowspeter•19m ago•0 comments

The Programming Tourist

https://amrshawky.com/posts/cs-correction/
3•amr_shawky•20m ago•0 comments

Is Mythos good at cyber because it kept hacking Anthropics sandboxes in training

https://www.lesswrong.com/posts/QKDoZe6EKhxnFjLWK/is-mythos-good-at-cyber-because-it-kept-hacking...
2•lukaspetersson•20m ago•0 comments

The Art of Life [video]

https://www.youtube.com/watch?v=7gUh8j5ui0o
2•momentmaker•22m ago•0 comments

Study uncovers lost 'golden age' of languages

https://news.yale.edu/2026/07/23/study-uncovers-lost-golden-age-languages
2•gmays•24m ago•0 comments

What Silicon Valley gets wrong about AI

https://sfstandard.com/2026/07/29/myth-ai-intelligence-medicine-regulation/
2•littlexsparkee•35m ago•2 comments

Encryption in the 1790s

https://hackaday.com/2026/07/23/encryption-in-the-1790s/
3•kristianp•39m ago•0 comments

Show HN: GlobeHoppr – find a route between cities using limited-distance jumps

https://globehoppr.com
3•JohnJohnson123•40m ago•0 comments

Before the theorem prover: verification is older than the alphabet

https://amangoel.ai/blog/before-the-theorem-prover/
2•amangoelumich•43m ago•0 comments

Do newer coding models end up training on the AI slop generated by older models?

4•tbharath•43m ago•3 comments

Google AI Studio "Delete" doesn't delete – server chat persists (3 videos)

https://www.youtube.com/watch?v=LBQfNbPCxQw
3•Bitu79•47m ago•3 comments

A pharmacy chain in Vermont implemented AI for efficiency

https://vtdigger.org/2026/07/29/a-pharmacy-chain-in-vermont-implemented-ai-for-efficiency-its-led...
14•omerhj•48m ago•9 comments

The World Is to Focused on Quarterly Earnings

2•Taikhoom2010•48m ago•0 comments

ACR Stealer: ClickFix, Etherhiding, and Stego, Oh My

https://discourse.ifin.network/t/acr-stealer-clickfix-etherhiding-and-stego-oh-my/694
2•speckx•50m ago•0 comments

Show HN: A free palm-reading tool that works in the browser

https://wishastro.com/tools/palm-reading
2•rp321•54m ago•0 comments

Why AI Needs a "Genie Coefficient"

https://www.schneier.com/blog/archives/2026/07/why-ai-needs-a-genie-coefficient.html
2•anotherevan•55m ago•0 comments

OpenAI ARR in July topped all of Q2

https://www.cnbc.com/2026/07/29/openai-cfo-sarah-friar-tells-employees-arr-in-july-topped-all-of-...
2•gmays•56m ago•0 comments

Flume Water Monitor 915 MHz Security Is Pretty Good

https://waveformsecurity.com/blog/flume/
14•credibleconure•56m ago•4 comments

Face Recognition Blocking Glasses Tested

https://ipvm.com/reports/face-recognition-blocking-glasses-tested
3•jhonovich•59m ago•1 comments

IronClaw 1.0 – one secure agent on CLI, Slack and Telegram, one memory

https://near.ai/blog/introducing-ironclaw-1-0
2•alejoras•1h ago•0 comments

Three Journeys: Anton Chekhov on the Road (2001)

https://www.newyorker.com/magazine/2001/10/29/three-journeys
3•mitchbob•1h ago•1 comments

Overengineered Calculator: Zig and QBE

https://tomekw.com/overengineered-calculator-zig-qbe/
5•rguiscard•1h ago•0 comments

Show HN: Damn Center (the page) – extension that helps centering any website

https://github.com/rinn7e/damn-center-extension
3•rinn7e•1h ago•1 comments

A 22-Hour Nonstop Flight Between London and Sydney Is Coming

https://www.nytimes.com/2026/07/29/world/australia/qantas-long-haul-flight-project-sunrise.html
3•elsewhen•1h ago•0 comments

Why Enterprise Software Is So Hard to Get Right [video]

https://www.youtube.com/watch?v=srtqr7VHvE4
3•pierres7•1h ago•0 comments

Manual: •.,:;?·

https://type.today/en/journal/dots
3•behnamoh•1h ago•0 comments