capnrefsmmat

6 ms·

1.
by capnrefsmmat 22d ago
More recent surveys have not contradicted the trend. For instance, in 2015, > On average, students spent about 14 HPW preparing for class, 6 HPW working for pay, 12 HPW relaxing and socializing, 7 HPW participating in co-curricular activ
2.
by capnrefsmmat 23d ago
> He seems quite out of touch with the level of economic hardship involved in being a student; a huge number of students already work a couple of jobs and are navigating small businesses, especially here in the UK where he apparently now
3.
by capnrefsmmat 1mo ago
> Well, what do you expect? LLMs are trained on blithering, mostly from web sites. So you get blithering out. I realize this isn't entirely serious, but I can't resist pointing out that this doesn't seem to be a good expla
4.
by capnrefsmmat 3mo ago
Being invited to conference talks around the world is a completely normal part of being an active researcher in almost any academic field, so it doesn't register as pompous to other academics.
5.
by capnrefsmmat 3mo ago
From your link, > TerraPower must still complete construction, submit an operating license application, and satisfy all applicable safety and regulatory requirements before loading fuel and beginning operations.
6.
by capnrefsmmat 6mo ago
Discussion is included in the Dangerous Goods Panel report, agenda item 4.3 (pages 39-41) and Appendix E (beginning page 89). https://www.icao.int/sites/default/files/DangerousGoods/DGP%... Paragraph 4.3
7.
by capnrefsmmat 6mo ago
Several reasons: 1. The post mainly reiterates a single idea (Capsicum enumerates what the process can do, seccomp provides a configurable filter) in many different ways. There is not much actual depth, code samples notwithstanding. Nothing
8.
by capnrefsmmat 6mo ago
Probably. One common feature of LLM output is grammatical features that indicate information density, like nominalizations, longer words, participial clauses, and so on. Perhaps training tasks that involve asking the LLMs for concise explan
9.
by capnrefsmmat 6mo ago
Thanks for the links. You may be interested in the other LLM writing style studies I've been collecting: https://www.refsmmat.com/notebooks/llm-style.html
10.
by capnrefsmmat 6mo ago
I've heard the Kenya and Nigeria story, but has anyone backed it up with quantitative evidence that the vocabulary LLMs overuse coincides with the vocabulary that is more common in Kenyan and Nigerian English than in American English?
11.
by capnrefsmmat 6mo ago
I work on research studying LLM writing styles, so I am going to have to steal this. I've seen plenty of lists of LLM style features, but this is the first one I noticed that mentions "tapestry", which we found is GPT-4o&#x27
12.
by capnrefsmmat 7mo ago
No, that doesn't really work so well. A lot of the LLM style hallmarks are still present when you ask them to write in another style, so a good quantitative linguist can find them: https://hdsr.mitpress.mit.edu/pub&#x2
13.
by capnrefsmmat 8mo ago
The first sentence is a reference to prior research work that has found those productivity gains, not a summary of the experiment conducted in this paper.
14.
by capnrefsmmat 9mo ago
Most of the tedious formatting requirements do not match what the final typeset article looks like. The requirements are instead theoretically to benefit peer reviewers, e.g., by having double-spaced lines so they can write their comments o
15.
by capnrefsmmat 9mo ago
Outside of disciplines that use LaTeX, the ability of authors to do typesetting is pretty limited. And there are other typesetting requirements that no consumer tool makes particularly easy; for instance, due to funding requirements, many j
16.
by capnrefsmmat 1y ago
It didn't "survey" devs. It paid them to complete real tasks while they were randomly assigned to use AI or not, and measured the actual time taken to complete the tasks vs. just the perception. It is much higher quality evid
17.
by capnrefsmmat 1y ago
Sure, if you're learning to write and want lots of examples of a particular style, LLMs can generate that for you. Just don't assume that is a normal writing style, or that it matches a particular genre (say, workplace communica
18.
by capnrefsmmat 1y ago
I don't think the AI companies are systematically working to make their models sound more human. They're working to make them better at specific tasks, but the writing styles are, if anything, even more strange as they advance. Co
19.
by capnrefsmmat 1y ago
Probably because the article uses the Unicode right single quotation mark instead of apostrophes, due to some automated smart-quote machinery. I'll have to adjust the tagger to handle those.
20.
by capnrefsmmat 1y ago
In our studies of ChatGPT's grammatical style ( https://arxiv.org/abs/2410.16107 ), it really loves past and present participial phrases (2-5x more usage than humans). I didn't see any here in a glance through
21.
by capnrefsmmat 1y ago
OpenAI is the custodian of the user data, so they are responsible. If you wanted the court (i.e., the plaintiffs) to find specific infringing chatters, first they'd have to get the data from OpenAI to find who it is -- which is exactly
22.
by capnrefsmmat 1y ago
If the output is interpreting sources rather than just regurgitating quotes from them, you need to exert judgment to verify they support its claims. When the LLM output is about some highly technical subject, it can require expert knowledge
23.
by capnrefsmmat 1y ago
Courts have always had the power to compel parties to a current case to preserve evidence. (For example, this was an issue in the Google monopoly case, since Google employees were using chats set to erase after 24 hours.) That becomes an
24.
by capnrefsmmat 1y ago
For introductory problems, the kind we use to get students to understand a concept for the first time, the AI would likely (nearly) nail it on the first try. They wouldn't have to fix any non-working code. And annotating the code likel
25.
by capnrefsmmat 1y ago
> Well, if you’re a novice, don’t do that. I agree, and it sounds like you're getting great results, but they're all going to do it. Ask anyone who grades their homework. Heck, it's even common among expert users. Here&#x2
26.
by capnrefsmmat 1y ago
I agree with all of this. But it's already very difficult to do even in a college setting -- to force students to get deliberate practice, without outsourcing their thinking to an LLM, you need various draconian measures. And for many
27.
by capnrefsmmat 1y ago
Sure, you could learn about grammar, plot structure, narrative style, etc. and become a reasonable novel critic. But imagine a novice who wants to learn to do this and has access to LLMs to answer any question about plots and style that the
28.
by capnrefsmmat 1y ago
The argument seems to be that for an expert programmer, who is capable of reading and understanding AI agent code output and merging it into a codebase, AI agents are great. Question: If everyone uses AI to code, how does someone become a
29.
by capnrefsmmat 1y ago
I'm not sure about other intellectual property rights. I do know there's a similar dichotomy for privacy: Americans tend to view it in terms of property rights (ownership of your image, data, etc.) while Europeans view it as a mat
30.
by capnrefsmmat 1y ago
For what it's worth, this is a uniquely American view of copyright: > The ONLY reason to have any law prohibiting unlicensed copying of intangible property is to incentivize the creation of intangible property. In Europe, particular

More ›