snet0

5 ms·

1.
by snet0 14d ago
Correct, that's not how LLMs work [0]. [0] - https://arxiv.org/abs/1706.03762
2.
by snet0 5mo ago
Why are you taking the fall and not the PM who authored the change (and submitted a PR with an uninformative title and no comment) and, I'm assuming, plays a role in managing the project?
3.
by snet0 5mo ago
How are you handing the cube state to the model?
4.
by snet0 5mo ago
A fine way to start a house fire, sure.
5.
by snet0 7mo ago
So what? If I could take a sugar pill that guaranteed I feel comfier looking at my screen, nobody can tell me it "doesn't work". I'm not trying to optimise my life, I'm trying to have my eyes feel better.
6.
by snet0 7mo ago
But have you considered that it feels better?
7.
by snet0 8mo ago
Or: this is why you strictly regulate the storage of confidential/private/sensitive information. There were multiple failures here, but a single step could've prevented the entire hack: industry-standard encryption of the sen
8.
by snet0 8mo ago
Is there a perspective or analysis that you've read that does a good job, in your opinion?
9.
by snet0 8mo ago
What did happen, then?
10.
by snet0 9mo ago
I mean the fact is that it's easy to fake because the permissible space of interpretation is almost infinite. That will always be the case, and the only thing people demonstrate when they create fake analyses is that they can't be
11.
by snet0 9mo ago
I don't think you believe this, honestly. The point, in my view, of art is to form personal relationships with the artwork. I can read Notes From Underground with no background on the era or the author, and pass my own judgements on th
12.
by snet0 9mo ago
At some point we have to bound our terms, obviously if someone interprets The Great Gatsby to be making commentary on interplanetary space travel they are incorrect but if someone was to interpret The Great Gatsby as containing some meani
13.
by snet0 9mo ago
I hate this sentiment. The book isn't "about" a thing in particular, neither does it "mean" any specific thing. It may have been written with some ideas in mind, and there may even be overt indications as to those i
14.
by snet0 9mo ago
I haven't read much about it to understand what's going on, but the development of multi-modal models has also felt like a major step. Being able to paste an image into a chat and have it "understand" the image to a comp
15.
by snet0 9mo ago
Notably this doesn't match the current thread.
16.
by snet0 9mo ago
I never said C++ was inefficient, you don't have to prove anything. It's a hypothetical, try use your imagination. > Corporations gobbling up tax payer money to build power hungry datacenters so billionaires can replace worker
17.
by snet0 9mo ago
If a "C+++" was created that was so efficient that it would allow teams to be smaller and achieve the same work faster, would that be anti-worker? If an IDE had powerful, effective hotkeys and shortcuts and refactoring tools that
18.
by snet0 9mo ago
If you define "simple thing" as "thing an AI can't do", then yes. Everyone just shifts the goalposts in these conversations, it's infuriating.
19.
by snet0 9mo ago
This is conspiratorial nonsense.
20.
by snet0 9mo ago
I don't see why this would be the case.
21.
by snet0 9mo ago
To say that a model won't solve a problem is unfair. Claude Code, with Opus 4.5, has solved plenty of problems for me. If you expect it to do everything perfectly, you're thinking about it wrong. If you can't get it to do a
22.
by snet0 10mo ago
I've often seen people mention that one reason for games like Call of Duty being so enormous is optimising for performance over storage. You'd rather decompress textures/audio files at install-time rather than during run-time
23.
by snet0 10mo ago
> With their latest data measurements specific to the game, the developers have confirmed the small number of players (11% last week) using mechanical hard drives will witness mission load times increase by only a few seconds in worst ca
24.
by snet0 10mo ago
You can read the proposal and found out, if you're interested. > In the light of the more limited risk of their use for the purpose of child sexual abuse and the need to preserve confidential information, including classified inform
25.
by snet0 11mo ago
I've had this thought before, when seeing labels that talk about kWh/day. The answer is very simple: you pay per kWh. When people want to know power efficiency, what they really want to know is "how much will this cost me to
26.
by snet0 1y ago
Do you try trick your calculator by entering large sums that it might not have been programmed to answer?
27.
by snet0 1y ago
I've not followed any evolutions in this area, but there's a cool paper from 2014 about using WiFi channel state information to detect 87%(!) of falls in an experimental condition[1]. It's been a while since I read the paper,
28.
by snet0 1y ago
> Instead they believe model alignment, trying to understand when a user is doing a dangerous task, etc. will be enough. Maybe I have a fundamental misunderstanding, but I feel like hoping that model alignment and in-model guardrails are
29.
by snet0 1y ago
In cases like this, where it's used to "automate the boring stuff", I just see it as outsourcing. You don't want to spend time building the frustrating CSS/HTML or whatever that isn't your skill set, and the LL
30.
by snet0 1y ago
> That's a great question! Never forget what AI stole from us. This used to be a compliment, a genuine appreciation of a good question well-asked. Now it's tainted with the slimy, servile, sycophantic stink of AI chat models.

More ›