croemer

9 ms·

181.
by croemer 7mo ago
> you’re reading an AI expansion of someone’s smaller prompt, which contained the original info you’re interested in This got me thinking: what if LLMs are used to do the opposite? To condense a long prompt into a short article? That tak
182.
by croemer 7mo ago
Related to this discussion: https://news.ycombinator.com/item?id=47673350
183.
Don't Write Like AI (1 of 101): "It's Not X, it's Y"(blakestockton.com)
1 points by croemer 7mo ago|1 comments
184.
by croemer 7mo ago
Pangram doesn't reliably detect individual LLM-generated phrases or paragraphs among human written text. It seems to look at sections of ~300 words. And for one section at least it has low confidence. I tested it by getting ChatGPT to
185.
by croemer 7mo ago
It's not just terseness, it's the rhythm and "it's not x, it's y". In fact, the latter is the opposite of terseness. LLMs love to tell you what things are not way more than people do. See https://www
186.
by croemer 7mo ago
I doubt you write like that. Where can I find your writing other than your comments which IMO don't read like the blog post?
187.
by croemer 7mo ago
In theory, wouldn't be too hard be to settle the question if whether he used ChatGPT to write it: get Olang to write a few paragraphs by hand, then have people judge (blindly) if it's the same style as the article. Which one sound
188.
by croemer 7mo ago
I have nothing against em dashes. As long as your writing is human, experienced readers will be able to tell it's human. Only less experienced ones will use all or nothing rules. Em dashes just increase the likelihood that the text was
189.
by croemer 7mo ago
These are just some of the good examples I found. My hunch that this is substantially LLM-generated is based on more than that. In my head it's like a Bayesian classifier, you look at all the sentences and judge whether each is more or
190.
by croemer 7mo ago
Here's one tell-tale of many: "No alarm, no program light." Another one: "Two instructions are missing: [...] Four bytes." One more: "The defensive coding hid the problem, but it didn’t eliminate it."
191.
by croemer 7mo ago
The AI figures of speech make for painful reading. I don't want to read dozens of examples of what something is not. Just say what it is.
192.
NPM: Require 2FA confirmation to change email, issue token, alter 2FA methods(github.com)
1 points by croemer 7mo ago|1 comments
193.
by croemer 7mo ago
Currently, as long as you're logged in, npm will not ask for password nor 2FA token for anything. You can change email, generate tokens, add/remove 2FA at will.
194.
by croemer 7mo ago
Incidentally, email bombing was likely used in the axios attack to bury some important security notifications: https://github.com/axios/axios/discussions/10612
195.
by croemer 7mo ago
Maintainer replied here https://github.com/axios/axios/discussions/10612 > nope this was just someone bombing the repo throught the API it seemd > i then just closed and deleted them with a script.. see
196.
by croemer 7mo ago
`npm install` might be enough to pull it, unless you pin down to the patch?
197.
by croemer 7mo ago
You mean you vibe coded something. "Zero deps. One file." People prefer hand-written comments over LLM-written ones. "Already detects the hijacked maintainer email on the current safe version." You simply flag all proton
198.
by croemer 7mo ago
Has anyone else noticed there was a recent sudden flurry of 3000 deleted issues on axios/axios? The jump happened on March 23. Was this a first sign of compromise? Or just coincidence of an AI agent going rogue. There are pretty much e
199.
by croemer 7mo ago
I'm impressed how this was caught as a network anomaly in a GitHub actions monitoring tool. This might have taken a lot longer to discover, otherwise.
200.
by croemer 7mo ago
NextDNS has a setting to block newly registered (<30d) domains.
201.
by croemer 7mo ago
I lost respect for Axios when they made a breaking change in a patch release. Digging into the root cause, I found the maintainer had approved an outside PR with an obvious AI slop PR description: https://github.com/axios&#x
202.
by croemer 7mo ago
(2024)
203.
by croemer 7mo ago
It's just a tournament, the winner goes to next round. Play it fast, all the way to the end and you'll see how it works. It's some 16 or 32 rounds all the way.
204.
by croemer 7mo ago
It does show you on the left. Just not on the certificate.
205.
by croemer 7mo ago
Played it twice to see if it's reproducible. First time, Fira Code; second time Source Code Pro. Source Code Pro came in second first time round as well. Been using Fira Code until now.
206.
by croemer 7mo ago
When using ChatGPT Android app with some NextDNS block lists, I get an error modal in app saying "security misconfiguration blah blah". Clearly I'm blocking some tracker and it's upset about that. I allowlisted a sentry
207.
Android’s new sideload settings will carry over to new devices(androidauthority.com)
150 points by croemer 7mo ago|240 comments
208.
by croemer 7mo ago
Google clarifies that this status can carry over to new devices, so you only ever have to go through it once.
209.
by croemer 7mo ago
Does anyone have an archive.is? Page hasn't been loading the 4 times I tried over the course of an hour. Found one: https://web.archive.org/web/20260328080015/https://hn-trusts...
210.
by croemer 7mo ago
It's not littering. It shows why fax is stupid and they should accept email. Littering has no such benefit.

More ›