rhelz

6 ms·

1.
by rhelz 7d ago
Not saying that it isn't useful. It's just not science.
2.
by rhelz 8d ago
Thanks. Nobody ever admits when they are wrong on the internet, so I'm trying to be the change I want to see. W.R.T. the grand historical tradition, etc, note that the great majority of people who practiced that were, like Plato, inde
3.
by rhelz 9d ago
Agreed. There are multiple senses for "simple" and on of those senses is "small". It is actually a very useful sense too, as Solomonov/Levin/Kolmogorov/Chaitin-style inductive reasoning has shown. And i
4.
by rhelz 9d ago
If somebody finds patterns in them, that means they can be improved by reducing the patterns. Patterns and symmetry are forms of redundancy. The less memory the weights take to meet a level of competency, the more random, and therefore pat
5.
by rhelz 9d ago
Thanks for this reply, your points are well taken and it really got me to think about this stuff. I suppose you could view arxiv as being a way to let more people do a peer review on the papers, i.e. as one component of a more open and tran
6.
by rhelz 9d ago
What is your definition of a vanity press? Mine is that it will publish anything from anybody. Sure, there's lots of good stuff on there. But there's lots which is not good. Publishing there is not doing science. Science r
7.
by rhelz 10d ago
If the weights arn't compressed, then a smaller set of weights would perform as well. Sure, you can always induce as much symmetry and patterns as you want by bloating the data set, but that hardly gives us insight into how a set of
8.
by rhelz 10d ago
I'm not claiming any arbitrary set of weights is a minimal representation. But typically, if people could achieve the same quality of results with a smaller set of weights, or weights which have been quantized to lower bit representa
9.
by rhelz 10d ago
// why limit your article's impact // Because...science? It's not science until it passes peer review. I'm not advocating that everybody stops posting to arXiv, and I'm not saying you can't fin
10.
by rhelz 10d ago
You are not wrong. But this has by no means proven its up to the standard of being publishable in a machine learning journal. Its on arXiv.org, which, lets face it, at the end of the day is a vanity press.
11.
by rhelz 10d ago
Cyberphrenology. In any two random graphs, you'll find an isomorphic graph which is can be up to log of the size of the graphs. And if the LLM has been trained up to the limit of what data it can hold, it is going to be random. Pr
12.
by rhelz 13d ago
For those of you who are under the age of 25, "Nirvana" refers not to an enlightened state, but a popular rock band who had their heyday almost 40 years ago.
13.
by rhelz 13d ago
What am I missing here? Seems like an AI detection program would be as useful as a program which could tell whether John Henry used a sledge hammer or a steam hammer to lay a section of track. Or whether your CPA did your taxes using Excel
14.
by rhelz 13d ago
They most certainly are too stupid. This is a true story. In 1995, a guy named McArther Wheeler was flabbergasted when police picked him up for robbing a bank. He said "But I put the lemon juice on my face!!" He figured th
15.
by rhelz 15d ago
Not just Claude. For me, anyways, google ai and chatgpt swing from vapid to genius on a daily basis. These companies are obviously searching for some kind of sustainable business model, and trying to get us to upgrade to higher paid tier
16.
by rhelz 15d ago
There is one difference, and it is a very telling difference. In Idiocracy, people openly scorned smart people. They made no pretense of using any kind of logic or scientific method. What we have now is different. People stopped vaccin
17.
by rhelz 15d ago
What's the error rate for humans?
18.
by rhelz 27d ago
// America simply is no longer a democracy or representative democracy. // Well, that's the kicker. We actually did vote him in. TWICE. It would be great if we could blame some cabal or coup or something. It wou
19.
by rhelz 27d ago
Gold is a useful element, for sure, but there's no where's near enough of it to serve as a reserve currency. E.G. the U.S. reserves are worth about $1.2 trillion (changes each day, of course). Compare that to a national economy
20.
by rhelz 28d ago
I'm old enough to remember when people said that the debt Ronald Reagan racked up would never be repaid. Which means I also am old enough to remember Clinton turning record deficits into record surpluses. I remember when the U.S. Tre
21.
by rhelz 28d ago
The federal government is actually quite efficient in what it does. For example, the overhead for Social Security is between 0.5% and 1%. I.E, less than the mutual funds in my 401k charge per yeah for "administrative fees." Wh
22.
by rhelz 28d ago
How embarrassing is it to welcome foreign dignitaries to a building which has been partially torn down? I man, if he actually had replaced it with a ballroom, that would be different. But that's not what we get from Trump. He can&
23.
by rhelz 28d ago
By the time Trump leaves office, the whole federal government will be completely trashed, up to and including the East Wing of the White House being torn down to a hole in the ground. I've never felt so alienated from the leaders of my
24.
by rhelz 1mo ago
AI skeptics hate him. Here's how he doubled his productivity using AI with one simple trick. You have to stop thinking like an employee or a worker bee, and start thinking like a boss. Guess what, the work you give your boss isn'
25.
by rhelz 1mo ago
The general entshitification this article mentions is very real. However, if AI isn't making you 2-3x more productive than you were, you are doing it wrong.
26.
by rhelz 1mo ago
When I was using, the answer was always "more." It was great at first, but as I habituated to each dose level, I was prescribed more and more until I reached the legal limit. At which point, I habituated again, got none of the b
27.
by rhelz 1mo ago
According to the mighty google (and other HN comments) we are running out of address space at about 1 bit per year or so. So most of us will not live to see the need for 128 bit computers. An interesting thing to think about is this: how ex
28.
by rhelz 1mo ago
Most of the barriers for Alpha adoption were problems inherent in the 32-bit to 64-bit transition. Alignment problems, yes, but another killer problem was that 64-bit pointers were twice as large. The software I was writing at the time (ED
29.
by rhelz 1mo ago
"I am old, and have lived through four ages of the west, (the 8, 16, 32, and 64-bit ages), and have seen many defeats, and many fruitless victories."
30.
by rhelz 1mo ago
Oh and BTW we are nuking 20% of the world's oil. If somebody were deliberately trying to wreck the U.S. energy picture, they would be hard pressed to do better than this.

More ›