nickpsecurity

7 ms·

61.
by nickpsecurity 6mo ago
I was testing them on a HP laptop I bought for $200 with 4GB of RAM. Windows, its default, used so much memory that there was not much left for apps. Ubuntu used 500MB less than Windows in system monitor. I think it was still 1GB or more. I
62.
by nickpsecurity 6mo ago
Larger models better understand and reproduce what's in their training set. For example, I used to get verbatim quotes and answers from copyrighted works when I used GPT-3.5. That's what clued me in to the copyright problem. Where
63.
by nickpsecurity 6mo ago
The silo'd codebases I was referring to are verification tools they produce. They're used to prevent attacks. Each tool has one or more capabilities others lack. If combined, they'd catch many problems. Examples: KLEE test ge
64.
by nickpsecurity 6mo ago
They're great at Python and Javascript which have lots of tooling. My idea was to make X-to-safe-lang translators, X initially being Python and Javascript. Let the tools keep generating what they're good at. The simpler translator
65.
by nickpsecurity 6mo ago
You've never seen the full power of static analysis, dynamic analysis, and test generation. The best examples were always silo'd, academic codebases. If they were combined, and matured, the results would be amazing. I wanted to do
66.
by nickpsecurity 6mo ago
Also, synthetic data and templates to help them discover new vulnerabilities or make agents work on things they're bad at. They differentiate with their prompts or specialist models. Also, like ForAllSecure's Mayhem, I think they
67.
by nickpsecurity 6mo ago
I agree with you on how their quality is spread out. But, this... "School-type long essays only seem to exist in academia." Does an AI know what an essay is? Would it consider any long, descriptive post an essay? Especially if pre
68.
by nickpsecurity 6mo ago
I feel like this is actually human-like but like the average human in the pretraining data. Let's look: 1. They reward short or under-developed essays. I'd say most online content, especially with high upvotes next to the post, fi
69.
by nickpsecurity 6mo ago
That we're building theories on what's left of mostly-trashed data has scientific implications. Most people hearing LHC proved something probably didn't think a preprocessor threw away most observations first. That layer of i
70.
by nickpsecurity 6mo ago
It's a discussion forum. Saying people are all wrong with no proof comes off as arrogant but isn't helpful. If you have links to examples, you can simply say, "Here's some prior art or previous work in this area you all
71.
by nickpsecurity 6mo ago
Christians usually only believe in God, angels, humans, and animals. That would mean intelligent UFO's might be angels or demons. While that's speculation, one guy did an interesting test of it. Non-believers are much more vulne
72.
by nickpsecurity 6mo ago
Nah, they like to lay on the laptop to eliminate their competition. They want all our attention.
73.
by nickpsecurity 7mo ago
Says a biased source selling AI code generators. He's also a liar because he knows many of us reject it due to principles, preferences, to stay mentally sharp, increased legal accountability for suppliers, or avoiding lawbreaking (eg c
74.
by nickpsecurity 7mo ago
Those made without agentic tooling are still better. So I don't know that what you said is true in practice even if I agree on its potential.
75.
by nickpsecurity 7mo ago
Something that works, has good diagrams, and the right datatypes. And with HN comments reflecting that.
76.
by nickpsecurity 7mo ago
Thanks for this! Parallel on Python is always a pain point. I'm always grateful for each tool one of you builds to help us speed up our code. :)
77.
by nickpsecurity 7mo ago
There's a lot of stuff with basic errors in peer reviewed journals. Things also can get rejected for anything from formatting to politics. I like Arxiv better. I get the paper, know it's probably not reviewed (like in many journal
78.
by nickpsecurity 7mo ago
They don't make up things for the reason we do. They usually don't have a sense of self, introspection, etc driving their actions. They have no memory like the hippocampus outside some academic prototypes. They don't have a c
79.
by nickpsecurity 7mo ago
Thanks for thr tip. The closest thing to real neurons are called spiking, neural networks. Here's a list of libraries for building them on GPU's: https://github.com/open-neuromorphic/open-neuromorphic For aca
80.
by nickpsecurity 7mo ago
It was a great video. You're doing interesting work. I've also seen implementations of realistic neurons, spiking models, etc. In software implementations, what combo of libraries and hardware would equal your 200,000 biological n
81.
by nickpsecurity 7mo ago
Putting in a GPU or virtualization system, like Intel's, is probably much harder than the average, independent commit.
82.
by nickpsecurity 7mo ago
I remember trying out Macromedia Flash 6.0. My GUI apps were ugly at the time. Learning to build something like I saw in the movies could take years. Then, Flash let me throw together beautiful, animated interfaces like it was nothing. One
83.
by nickpsecurity 7mo ago
Also, TPU support. Hardware diversity.
84.
by nickpsecurity 7mo ago
There was an open, real-time strategy game created for this purpose long ago. I think it was intended for designs like the Starcraft AI's of the time. Anyone remember or use it?
85.
by nickpsecurity 8mo ago
I'll cross link the last submission about this: https://news.ycombinator.com/item?id=47009024 I feel like Chris way, way understates the prior art on C language. It's not just two compilers and some textbooks. Tha
86.
by nickpsecurity 8mo ago
"Is Google hurting itself in its confusion? Google is the largest scraper in the world. Google's entire business began with a web crawler that visited every publicly accessible page on the internet, copied the content, indexed it,
87.
by nickpsecurity 8mo ago
My concept was to do this with two pieces: 1. Generic, mask layers and board to handle what's common across models. Especially memory and interface. 2. Specific layers for the model implementation. Masks are the most expensive part of
88.
by nickpsecurity 8mo ago
Bruce Schneier had the Friday Squid Blogging posts that he allowed random stuff on. Most of the noise was contained in the Friday threads. Lobsters had a weekly thread by caius called, "What are you doing this week?" People would
89.
by nickpsecurity 8mo ago
You could charge people to fine-tune or customize models. Maybe synthetic data. Maybe rent it on vast.ai but careful about your bandwidth and energy. I'm not sure of the security implications of vast.ai, either.
90.
by nickpsecurity 8mo ago
I found a good review of it: https://undark.org/2025/11/28/book-review-hidden-guests/

More ›