programjames

11 ms·

1.
by programjames 2mo ago
Yes, people can always claim that wasn't their intent, but what you see quite often nowadays is people openly saying their intent is to get people to take action. "I wouldn't personally... but I have massive respect for those
2.
by programjames 2mo ago
The standard can be, "don't utter words with the intent of spurring others onto antisocial actions."
3.
by programjames 2mo ago
Where is the "tab" autocomplete?
4.
by programjames 2mo ago
I skimmed through the book, and it's lacking the information theory foundations. For example, "trust region methods" come from maximizing the policy's relative entropy (to a reference policy) under a tournament system wh
5.
by programjames 2mo ago
Standard Intelligence released one two years ago: https://si.inc/posts/hertz-dev/ It's only 8.5B and doesn't sound like it's quite conversational.
6.
by programjames 2mo ago
Didn't Standard Intelligence release a duplex model two years ago? Sounds disingenuous to market this as a new generation of voice models, when it is really OpenAI finally catching up to the current generation after two years. https:&
7.
by programjames 2mo ago
Logicians' training is so different from philosophers' that it should be considered a separate discipline, or under the branch of computer science.
8.
by programjames 2mo ago
The appearance of a logical argument is easier to achieve and often good enough for their purposes (publishing papers, winning lawsuits).
9.
by programjames 3mo ago
Optimizing for the metric involves: 1. Optimizing for generally applicable skills that the metric is trying to measure. 2. Optimizing adversarially to hill-climb the metric. You want candidates to do (1) and not (2). You can make them agnos
10.
by programjames 3mo ago
Nondeterminism is also a feature, not a bug. If you don't want people to optimize against your filtering process, you have to make it somewhat nondeterministic. For example, better candidates are exponentially more likely to pass the f
11.
by programjames 3mo ago
1.0 is "natural units". If your energy corresponds to nats, you should be using temperature 1.0. If your energy corresponds to bits, you should be using temperature ln(2) ~= 0.7. The optimization pressure is max nats = max
12.
by programjames 3mo ago
You should include the error correction code length in the description length. This means Newtonian mechanics was a much longer theory to describe Mercury's orbit than general relativity. It was only the shorter theory before they had
13.
by programjames 3mo ago
That is very unlikely due to Solomonoff induction...
14.
by programjames 3mo ago
You should look into Solomonoff induction. Nature and physics, absolutely, tautologically, have to follow the "shortest explanation is more likely principle". https://en.wikipedia.org/wiki/Solomonoff%27s_theor
15.
by programjames 3mo ago
They're likely talking about style transfer from a decade ago: https://arxiv.org/abs/1611.07004
16.
by programjames 3mo ago
My perspective is it is impossible to cut through the three layers of bullshit between you and anyone who knows what they are talking about. The only way to do this is with brand-name qualifications, like "MIT graduate", not thing
17.
by programjames 3mo ago
"A handful of years"? This is like the trend in K–12 education of blaming all issues on the COVID-19 pandemic. No, education was in a visible decline for five years before COVID, which means it had probably been declining for anot
18.
by programjames 3mo ago
If you click the website's name to the right of the title, it pulls up all the submissions from the same site: https://news.ycombinator.com/from?site=github.com/philipl
19.
by programjames 3mo ago
My experience was pretty contrary to points (1) and (4). My best teachers/professors directly conveyed information or skills. I found most students did the bare minimum to pass their classes (where "pass" = "not get thei
20.
by programjames 3mo ago
Your question is easily resolved by looking up how much American schools are funded, compared to historical funding, other countries' funding, and their relative successes.
21.
by programjames 4mo ago
People have more wildly different definitions for "fair" than "equity".
22.
by programjames 4mo ago
Not only is that implication rude, it's just not true. I am at least in the 95th percentile of amount of reading. I just think the article is poorly written. Not everyone is a good writer these days (or any days).
23.
by programjames 4mo ago
I clicked on the article to learn, "why Japanese companies do so many different things," and then got hit with pages of low-bitrate context, such that my eyes started glazing over and it was difficult to find the answer to the que
24.
by programjames 4mo ago
The primary objective of the United States, at least according to their leaders, is suspiciously absent from their comment. If we're being charitable, it is clear they disagree with their leaders: > All of this to get to a point whe
25.
by programjames 4mo ago
So, to clarify, you were just trying to spread the meme that Iran giving up its HEU was equivalent or inferior to just holding to Obama's deal, in spite of your belief the commenter you were replying to would not agree with this, nor t
26.
by programjames 4mo ago
> Long comment before yours "So a win will be returning to Obama’s Iran deal?" Where in that long comment before yours did you see an implication that anyone was thinking a return to Obama's Iran deal was a win? Why did
27.
by programjames 4mo ago
Maybe not this year. The next cycle though will better pander to Chase Oliver voters.
28.
by programjames 4mo ago
When the war is fought over literal nuclear weapons, you should not be so tongue-in-cheek about calling a war nuclear.
29.
by programjames 4mo ago
Why would you think returning to Obama's Iran deal would be a win? Actually, let me word this better: how could you possibly think that anyone in the White House for the past decade would think returning to Obama's Iran deal is be
30.
by programjames 4mo ago
I'm sure you've at least heard what the United States' leaders have to say. To paraphrase Trump, "stopping nuclear proliferation to a group of lunatics that support terrorism". And also (still paraphrasing Trump),

More ›