ec109685

8 ms·

91.
by ec109685 1y ago
Why can’t bridge just convert the money into USD? What’s the point of the stable coins step?
92.
by ec109685 1y ago
Editing the commit at the bottom of the stack is part of the magic of jj since it makes it so easy. So not knowing how that works in a github workflow or what the recommendation is feels like a miss on a “jj for newbie” tutorial. And it’s b
93.
by ec109685 1y ago
If I use Anthropic’s api for search, but then send user traffic directly to websites after showing the user the link, there’s no way for cloudflare to attribute that search to Anthropic. That makes the ratios of crawl to referrals shown sus
94.
by ec109685 1y ago
Goal of every fantasy / gaming site is to have customers lose money sustainably. It’s not in their best interests to have whales chase everyone away. Parlays play in that space too where the allure of a high payday keeps people hooked.
95.
by ec109685 1y ago
This tutorial is excellent. It’d be helpful to also include how jj flows map to pull requests in the GitHub/GitLab model, since many new jj users collaborate via PRs with main protected by default. What’s the best way to represent a st
96.
by ec109685 1y ago
To the sibling comment that is too deep to reply to, this covers why delayed conflict resolution is advantageous: https://news.ycombinator.com/item?id=45084835 It’s for other branches that hang off the commit that introduce
97.
by ec109685 1y ago
Can’t the exploit just be encoded in files that are used when the npm module is actually used? It seems like not running it at package install time doesn’t afford that much protection.
98.
by ec109685 1y ago
Turned out my prediction was wrong that the big lab’s wouldn’t release a browser with cross-site access before it was safe: https://news.ycombinator.com/item?id=45004846 > Anthropic says it hopes to use this research pre
99.
by ec109685 1y ago
Given there was a complete implementation it was porting, the simplest thing possible has a greater chance of working.
100.
by ec109685 1y ago
The CEO of Perplexity hasn't addressed this at all, and instead spent all day tweeting about the transitions in their apps. They haven't shown any sign of taking this seriously and this exploit has been known for more than a month
101.
by ec109685 1y ago
A local LLM wouldn’t have helped at all here.
102.
by ec109685 1y ago
It still keeps you in the loop, but doesn’t ask to run shell commands, etc.
103.
by ec109685 1y ago
The fact that every single system prompt has been leaked despite guidelines to the LLM that it should protect it, shows that without “physical” barriers, you are aren’t providing any security guarantees. A user of chrome can know, barring b
104.
by ec109685 1y ago
That was my point about dropping privileges. It can still be exploited if the summary contains a link to an image that the attacker can control via text on the page that the LLM sees. It’s just a lot of Swiss cheese. That said, it’s definit
105.
by ec109685 1y ago
This statement on your post seems to say it would definitively prevent this class of attacks: “In our analysis, we came up with the following strategies which could have prevented attacks of this nature. We’ll discuss this topic more fully
106.
by ec109685 1y ago
It’s clear if what Comet was doing was safe, Chrome would already have implemented it. The browser is the ultimate “lethal trifecta”: https://simonwillison.net/2025/Jun/16/the-lethal-trifecta/ Giving an
107.
by ec109685 1y ago
The only safe CORS spec is CORS. Have to treat everything the LLM is doing as malicious. It’s actually worse than that though. An LLM is like letting attacker controlled content on the page inject JavaScript back into the page.
108.
by ec109685 1y ago
It’s obviously fundamentally unsafe when Google, OpenAI and Anthropic haven’t released the same feature and instead use a locked down VM with no cookies to browse the web. LLM within a browser that can view data across tabs is the ultimate
109.
by ec109685 1y ago
He gave up on that: https://www.independent.co.uk/news/world/americas/us-politic...
110.
by ec109685 1y ago
Anthropic has opportunities to optimize their models / prompts during reinforcement learning, so the advice from the article to stay close to what works in Claude code is valid and probably has more applicability for Anthropic models t
111.
by ec109685 1y ago
I wouldn’t get hung up on one shotting anything. Output to a format that can be machine verified, ideally in a format there is plenty of industry examples for. Then add a grader step to your agentic loop that is triggered after the files ar
112.
by ec109685 1y ago
This article is bad in a lot of ways, but all of sudden it jumps to running Linux in a browser: “After my failed custom image attempt, I found a promising XFCE-based Debian image on Docker Hub. I downloaded it in minutes and, with a few com
113.
by ec109685 1y ago
Actually call volumes increased: “ The union took CBA to the workplace relations tribunal earlier this month as the company wasn’t being transparent about call volumes, according to a statement Thursday from the Finance Sector Union. The na
114.
by ec109685 1y ago
Hilarious how little context that model has access to.
115.
by ec109685 1y ago
Do all ai agents only include a single AGENTS.md, closest in the hierarchy? Where do people put common instructions or do they just copy and paste?
116.
by ec109685 1y ago
Why did it move to Forgejo?
117.
by ec109685 1y ago
I used to be a meta moderator there. But you're right, you need to have a strong "hand" or the communities like that fall apart. Their original owners also sold the site.
118.
by ec109685 1y ago
The I is emphasized more in I scream than ice cream I think. But it’s great point that you need context to be sure.
119.
by ec109685 1y ago
Or at least more details. Very cool!
120.
by ec109685 1y ago
You’d think they’d use the better model for at least videos that have a large view counts (they already do that when deciding compression optimizations).

More ›