dijksterhuis

5 ms·

1.
by dijksterhuis 12d ago
> The model weights change as the model goes through the training process. Yes. They do. You are absolutely right about that. But the model architecture doesn't change as a result of the training process. A piston doesn't sud
2.
by dijksterhuis 12d ago
> Nobody says a system or product works a certain way and means the system while it's half built. "Bridges drop cars in the water!". Right. To understand how an engine works, it's important to understand what a piston
3.
by dijksterhuis 12d ago
> The finished product is something which has been through post training. again, the finished product wasn't what was discussed by GP, and you didn't clarify that you were switching to discussing RL (which is still probabilis
4.
by dijksterhuis 12d ago
> "it's not just a next-token predictor because a bunch of the training isn't about predicting the next token." clever procedures on top of the base transformer architecture. i used simplified words/phrases to
5.
by dijksterhuis 12d ago
From GP, i.e. the context for this local part of the thread > Autoregressive LLMs generate tokens one at a time, disputing this is just plain wrong. next-token prediction i.e. the bit built during pre-training. at no point in your repl
6.
by dijksterhuis 12d ago
they are cleverly arranged / configured next-most-likely-token predictors, possibly with some clever procedures / attachments on top.
7.
by dijksterhuis 12d ago
prediction is a very specific term of art in the field of machine learning. generally speaking, machine learning models like LLMs are based on probability; performing a statistical prediction of the likely y given some input x Proba
8.
by dijksterhuis 12d ago
hn submission: https://news.ycombinator.com/item?id=49562744 -- 84 points -- 67 comments
9.
AI adoption at work is broad but shallow(theregister.com)
4 points by dijksterhuis 12d ago|0 comments
10.
by dijksterhuis 12d ago
if nobody owns the work, then anyone can take it from you.
11.
by dijksterhuis 12d ago
From the linked report > The agents continue to poke around on DSEWiki. A few hours after they find the site, they start probing it for cross-site scripting (XSS) vulnerabilities. [...] The agent swarm starts testing whether they can exe
12.
by dijksterhuis 13d ago
will not not necessarily be true for AI generated works in all territories https://news.ycombinator.com/item?id=49382041
13.
by dijksterhuis 16d ago
I had a great experience using $PROJECT_NAME. Another sentence that says nothing at all; but look, fancy and technically correct grammar use! You’re absolutely correct in agreeing with the author. Of course there’s an opposing point to cons
14.
by dijksterhuis 20d ago
> should probably not be listened to > a right to express their opinion ...
15.
by dijksterhuis 21d ago
couple of notes — 1a. Illegal immigration has nothing to do with visa based, legal immigration. illegal immigration is the type of immigration people are upset about here in the UK. most people have no problem with legal migration. 1b. anyo
16.
by dijksterhuis 21d ago
i'll echo this sentiment. historically there's been a lot more flexibility with GL CI yaml syntax, including real yaml anchors https://frenck.dev/github-actions-yaml-anchors-aliases-merge... sys-admin-ing gitlab
17.
by dijksterhuis 22d ago
i feel the same way. there's nothing interesting or curious about someone writing prompts to outsource a bunch of effort/thinking/work off to some most likely continuation sequence predictor / set of most likely continua
18.
by dijksterhuis 22d ago
i wouldn't mind one as well if at all possible? a first look through the comments on some first page threads and they seem far more thoughtful and considered . it generally seems like a nicer commenting environment that might actually
19.
by dijksterhuis 23d ago
> Short version: We provide our service as is, and we make no promises or guarantees about this service. Please read this section carefully; you should understand what to expect. https://docs.github.com/en/site-polic
20.
by dijksterhuis 25d ago
Is it?
21.
by dijksterhuis 26d ago
for anything you want to be able to copyright, yes. otherwise you can still prompt your agent things, but any output falls under public domain. since this hype wave started some people have been saying don’t use these tools in an environmen
22.
by dijksterhuis 26d ago
if you modify enough of the output as to make it unrecognizable as the machine output -- that's a new work created by a human. i.e. you don't have to change every line. same thing applies to sampling in music. if i change enoug
23.
by dijksterhuis 27d ago
a bit further down ... > Neither mere prompting nor the selection between several AI suggestions is sufficient as a human creative contribution. which reads like you actually have to contribute by modifying the output from the system, i.
24.
by dijksterhuis 27d ago
> What's the equivalent of "Google it yourself" these days? not mine, no license, only chinese language on the eventual page, but has the beginnings of something beautiful https://github.com/chenxuan520&#x2
25.
by dijksterhuis 28d ago
me personally? the company up to a reasonable point. it’s the company’s responsibility to ensure they provide a suitable working space. then again i live in the UK where we actually have workers rights.
26.
Self-Service Analytics Is Not Self-Service. But Don't Tell Anyone(dataanalysis.substack.com)
2 points by dijksterhuis 28d ago|0 comments
27.
by dijksterhuis 28d ago
> In my situation above we were accommodating people. They got to choose where they worked. this sounds like the WFH people might be treated the same way as the in-office people. as in you're retaining an in-office culture but peopl
28.
by dijksterhuis 28d ago
"office hours". a dedicated afternoon where anyone from the company could drop in to a running call and just ask me anything about engineering. bug reports; features; questions; feedback; anything. building in an extra 30 minutes
29.
by dijksterhuis 28d ago
> Coworking spaces might help with socialization and some people need it but I am not sure how that benefits who you work for. because, as someone else i replied to in the thread points out, miserable employees perform worse .
30.
by dijksterhuis 28d ago
100% agreed. i flamed out while remote working at $LAST_JOB. it was a combination of a changing cultural dynamic, changing expectations of my role (which i wasn't comfortable with) and a lack of necessary staffing. remote work was a bl

More ›