Y
HN Search
Hacker News Search
new
|
comments
|
top
|
jobs
zetalyrae
searching PlanetScale…
1.
▲
2.
▲
3.
▲
4.
▲
5.
▲
6.
▲
7 ms
·
31.
▲
by
zetalyrae
4mo ago
But the problem is people are not just delegating formulaic procedural prose to AI. They're using AI to write entire scientific papers, so now reviewers have to use Pangram[0] to screen submissions. Literary magazines have the same pro
32.
▲
by
zetalyrae
4mo ago
> This is a very good essay when you get past the arrogant tone evident especially at the beginning (is that a form of engagement bait?…) We have reached the point where you have college professors who defend[0] using AI to write scienti
33.
▲
by
zetalyrae
4mo ago
But writing, calculators, search engines etc. are specific tools, you drop a skill to gain effectiveness in another. While AI is a general tool whose builders intend it to be able to perform all cognitive work. Past a certain point, what is
34.
▲
by
zetalyrae
4mo ago
> The reason why people use AI to manifest their ideas is probably because they have no other way communicate otherwise. Isn't this a bit circular? They're not communicating to the AI through a BCI.
35.
▲
by
zetalyrae
4mo ago
Translation is fine.
36.
▲
by
zetalyrae
4mo ago
If you can't even be bothered to write the README without AI, why should I read it?
37.
▲
The History of Neural Networks, 1900-1990 [video]
(youtube.com)
2 points
by
zetalyrae
4mo ago
|
0 comments
38.
▲
Starfish by Peter Watts (1999)
(rifters.com)
145 points
by
zetalyrae
4mo ago
|
65 comments
39.
▲
The Weather Machine (2008)
(events.foresight.org)
4 points
by
zetalyrae
5mo ago
|
0 comments
40.
▲
The World, the Flesh and the Devil (1929)
(marxists.org)
4 points
by
zetalyrae
5mo ago
|
0 comments
41.
▲
by
zetalyrae
5mo ago
Author says he likes his boss and calls him genuine. Maybe he's more able to look on the good side than I. But I can't imagine working with someone like this and having positive thoughts about them. This: A month or two into
42.
▲
by
zetalyrae
5mo ago
Arguably this is what retirement is, no? Productivity gains did create extra leisure time. It's just we save all the leisure for retirement.
43.
▲
by
zetalyrae
5mo ago
> no level of AI-enabled automation will help me with my inherent ADD I have another post you might find useful :) https://borretti.me/article/notes-on-managing-adhd
44.
▲
by
zetalyrae
5mo ago
This may be some saltiness on my part. I wish I could curate the wiki of late Soviet military hardware!
45.
▲
by
zetalyrae
5mo ago
The font looks like the same font in every AI-generated visualization like this.
46.
▲
by
zetalyrae
5mo ago
Why some kinds of information processing and not others? As I wrote elsethread: why are some types of information processing "privileged" to create phenomenal experiences, while others run "in the dark"?
47.
▲
by
zetalyrae
5mo ago
Very well explained.
48.
▲
by
zetalyrae
5mo ago
For two reasons: 1. This requires explaining why only some kinds of information processing are privileged to be conscious, which seems rather arbitrary. 2. There's the question of levels of abstraction. Which information processor is c
49.
▲
by
zetalyrae
5mo ago
> Question: who says math cannot result in consciousness? Which math? Why some kinds of information processing and not others? If all information processing leads to consciousness: why does consciousness stop at the boundary of the brain
50.
▲
by
zetalyrae
5mo ago
> And yes, a spreadsheet simulating every neuron in your brain would also simulate it. But why a spreadsheet simulating the brain, and not just a spreadsheet doing normal financial math? In other words: why are some types of information
51.
▲
by
zetalyrae
5mo ago
> The hope for resolving this, I think, is that once we understand all processes in the brain, there will be some process that clearly is the self-referential "person" that is produced by the brain in normal operation. Yes. I t
52.
▲
by
zetalyrae
5mo ago
The first point (analogizing the hard problem to the reaction to Darwinism) is a very common rhetorical move: an analogy and history of ideas, which is convincing to many people, but what does it prove? > A philosophical zombie would cla
53.
▲
by
zetalyrae
5mo ago
Do people not realize what they do to their reputation when they publish bottom-of-the-barrel, unedited AI slop under their own name?
54.
▲
by
zetalyrae
5mo ago
I remember reading this as a teenager, and despite being already so steeped in transhumanist ideas that I should have found it very ordinary, I was so excited by it, having never read something like it before, I raved about it to my mother.
55.
▲
by
zetalyrae
7mo ago
> Instructions are mere suggestions — linters are law Written by AI.
56.
▲
by
zetalyrae
8mo ago
I remember! I also used Pidgin OTR over the Facebook XMPP gateway. At some point Facebook started recognizing it, but not banning it: you could go to the web interface and you'd see "encrypted message" instead of noise.
57.
▲
by
zetalyrae
8mo ago
I remember using Pidgin in ~2009. A dozen chat networks, all on one app. Desktop software built with a native GUI toolkit. And, on top of all that: you could keep your chat logs forever. The world of yesterday.
58.
▲
by
zetalyrae
9mo ago
I expect it's somewhere in the training data, but it's very unlikely to be salient. A few textfiles here and there in the ocean of the Internet is nothing. If Claude had memorized the walkthrough, it would have performed better.
59.
▲
11:59 PM
(atharvaraykar.com)
4 points
by
zetalyrae
9mo ago
|
0 comments
60.
▲
by
zetalyrae
9mo ago
https://borretti.me/ my blog + fiction
More ›