ironrabbit

8 ms·

1.
by ironrabbit 2y ago
Which part of SB1047 do you think is too onerous or not "slow regulation" enough?
2.
by ironrabbit 3y ago
This is a forum post from a student with 18 total karma
3.
by ironrabbit 3y ago
Zero chance private github repos make it into openai training data, can you imagine the shitshow if GPT-4 started regurgitating your org's internal codebase?
4.
by ironrabbit 3y ago
Automatic kernel fusion (compilation) is a very active field, and most major frameworks support some easy-to-use compilation (e.g. jax's jit, or torch.compile which iirc uses openai's triton under the hood). Often you can still do
5.
by ironrabbit 3y ago
The median OAI employee with a 900k comp is probably L5, not L6 or L7
6.
by ironrabbit 3y ago
New hires' comp is much higher than existing employees', especially if you've hit your cliff. 7 figures for E6 can happen if you joined recently, have good counter-offers, and negotiate. It's not super uncommon but it&#x
7.
by ironrabbit 3y ago
All engineers and researchers, even junior, are "Member of Technical Staff" at OAI
8.
by ironrabbit 3y ago
> the company pushed back against a proposed amendment to the AI Act that would have classified generative AI systems such as ChatGPT and Dall-E as “high risk” if they generated text or imagery that could “falsely appear to a person to b
9.
by ironrabbit 4y ago
Character has their own models, and anecdotally I've heard they have one of the better LM training codebases out there.
10.
by ironrabbit 4y ago
Openai is probably the highest status employer among my network, but I don't know any university students
11.
by ironrabbit 4y ago
Andrej's "Building Makemore" series is exactly this, it includes a wonderful lecture where he computes all the gradients for a simple network by hand and compares them against the values produced by torch's autograd.
12.
by ironrabbit 4y ago
minGPT prioritized being understandable above all else, and was not very fast. This repo includes several optimizations, but it still much more understandable than probably any other open source implementation.
13.
by ironrabbit 4y ago
Anecdotally, chatgpt seems much worse to me than Google for getting correct answers. Like orders of magnitude worse. Tells me the wrong timezone for a city kind of bad. No doubt it will be much better in the future, and they've definit
14.
by ironrabbit 4y ago
This looks very nice but no android support is unfortunately a dealbreaker
15.
by ironrabbit 4y ago
Signal boosting in case anyone from notion is reading: please address performance, it is practically the only thing I care about. I love the product but it is almost unusably slow with my tiny amount of data. Cache more stuff on my device,
16.
by ironrabbit 4y ago
The "scandal" that prompted this article is referenced with sources under the heading "Why XCheck is in the news". Neither the article nor the comment you're replying to says anything about data breaches.
17.
by ironrabbit 4y ago
It's a position paper, academics publish them all the time, they're very much a part of scientific discourse. Just different than an experimental results paper.
18.
by ironrabbit 4y ago
are there any continuous calendars like the linked one that allow sync or import from google cal?
19.
by ironrabbit 4y ago
The author is a skilled research scientist in a very competitive space with some high-profile publications (e.g. Gumbel Softmax). He is absolutely an outlier, but not a unicorn -- AI researchers with good publications and reputation will at
20.
by ironrabbit 4y ago
Curious, are there benefits of greenway connections compared to allocating the same amount of green space elsewhere?
21.
by ironrabbit 4y ago
Is there any evidence that GPT-3 responses are edited/filtered before being returned to users? My understanding is that some GPT-3 responses are annotated post-hoc, and this data is used to fine-tune later versions of GPT-3 (InstructGP
22.
by ironrabbit 5y ago
As of a few years ago, lidar added at least $7500 to the cost of a car. That's a huge price difference for a consumer.
23.
by ironrabbit 5y ago
Not even mentioning Pluribus is just bad journalism
24.
by ironrabbit 5y ago
As someone who has had to reproduce others' research results: it is much much better to release the unclean, unorganized code that actually produced your results than it is to release nothing. Even if it doesn't run (e.g. it depen
25.
by ironrabbit 5y ago
I always assumed we'd keep the summer hours year round. I don't even think about DST during the summer, it's during the fall where we suddenly lose an hour of evening sunlight that I complain. Unless I'm misreading, your
26.
by ironrabbit 5y ago
It seems like this method outperforms those model-based approaches, why should we care that it's "just image analytics"?
27.
by ironrabbit 5y ago
I don't know much about meteorology -- what is "true weather prediction" then?
28.
by ironrabbit 5y ago
Slightly off-topic, but is there a good overview of machine learning research being done at Spotify?
29.
by ironrabbit 5y ago
I think it's very reasonable to hold that view (I do as well). But to propose a solution, it's necessary to understand the facts of the current situation, which is what I'm trying to learn.
30.
by ironrabbit 5y ago
Interesting. Resetting cost basis upon inheritance does seem like a massive tax avoidance loophole, I can't quickly think of a reason why this should be allowed. I agree also that this would probably be a popular (and sensible) modific

More ›