Manoj Kasichainula

@headmold
100 Followers
87 Following
203 Posts
A lemon gives by taking and cares by yelling. Former security at Google, Asana, Apache. If you're holding a snake right now, press 4.
Blueskyhttps://headmold.bsky.social
Profile photo by@bdowney
Manoj Kasichainula16h ago
https://simonwillison.net/2026/Jun/11/anthropic-walks-back-policy/ "We made the wrong tradeoff and we apologize for not getting the balance right." I think someone at Anthropic had to snip out "Honestly," from the statement before sending it over.
Anthropic Walks Back Policy That Could Have ‘Sabotaged’ AI Researchers Using Claude

Big scoop for Maxwell Zeff at Wired: “We’re changing Fable 5’s safeguards for frontier LLM development to make them visible.” Anthropic said in a statement to WIRED. “We made the …

Simon Willison’s Weblog
Manoj Kasichainula2d ago
Damien Miller2d ago
Show threadManoj Kasichainula2d ago
To be clear, I don't see much if any information on how this works still, so Apple devs could surprise me with their cleverness (or my lack) as I learn more.
Show threadManoj Kasichainula2d ago

With a bit less jargon: On some sites, Apple's agent might need to read pages full of user-generated text to find the "change password" link. The text could trick Apple's agent into letting an attacker hijack your account.

If Apple wasn't careful, the "confusion" could even spread to other sites.

Manoj Kasichainula2d ago
ahahaha uhhhhh, I'd *like* to think smart security people at Apple were on this and cut off one of the legs of the lethal trifecta, but uhhhhh... https://www.macrumors.com/2026/06/08/apple-passwords-can-now-automatically-fix-passwords-with-agentic-ai/
Apple Passwords Can Now Automatically Fix Weak and Compromised Passwords With Agentic AI

Apple today announced that the Passwords app can now automatically update weak and compromised passwords using Apple Intelligence and Safari to take...

MacRumors
Manoj KasichainulaMay 11

"Humans check the work after an LLM does it" can't be reliable. Humans will just check the LLMs' work less and less carefully, because that's how human brains work.

The NYT should only deceive us through omission, bad framing, and bothsidesism, not hallucinations:

https://simonwillison.net/2026/May/10/new-york-times-editors-note/

A quote from New York Times Editors’ Note

This article was updated after The Times learned that a remark attributed to Pierre Poilievre, the Conservative leader, was in fact an A.I.-generated summary of his views about Canadian politics …

Simon Willison’s Weblog
Manoj KasichainulaApr 26
Show threadMatt BlazeApr 26
And am I the only person here who's seen season five of The Wire? The president was not evacuated. Well, maybe he was, but that's not what they're trying to say. Rather, he was removed from the room.
Manoj KasichainulaApr 22
jcoglanApr 17
you know you've been computering for too long when you can recognise what this is
Manoj KasichainulaApr 11
I hadn't been to Alamo in a while. QR code menus? My god, what have they done? https://www.indiewire.com/features/commentary/alamo-drafthouse-sucks-now-1235188253/ https://drafthouse.com/sf/news/our-shift-to-qr-ordering
The Absolute Hell of Watching a Movie at the Alamo Drafthouse in 2026

Once America's most promising movie theater chain, the Alamo Drafthouse has become a QR-coded symbol of corporate enshittification.

IndieWire
Manoj KasichainulaApr 8
With people gearing up to rapidly update software because of Claude Mythos's bugs, it seems like a great time to attack some supply chains.