Lerc

28 ms·

571.
by Lerc 7mo ago
So that's a Wayland ex-window manager then?
572.
by Lerc 7mo ago
I agree with you point, but I do look sidelong at the number of points the post has. It is, at the very least, unexpected.
573.
by Lerc 7mo ago
Geordie's thing was mostly superpower-but-you-look-idiotic. Discovery was just so ham fisted when they tried to make points and missed massive opportunities when they could have been done organicity because of the situation. I really t
574.
by Lerc 7mo ago
There are many possibilities. There are assistive technologies of various levels of quality, there is human assistance, and then there is simply being unable to participate. Chris McCausland says he relies a lot less on others due to AI. I
575.
by Lerc 8mo ago
I recently did some looking into how public perception campaigns work. I found it amazing that I could not find any organisation that tracks these campaigns. These are often very well funded and those funds go to people. Part of the proble
576.
by Lerc 8mo ago
I think you could absolutely draw a causal link, it wouldn't explain why 2026 instead of 2024 or 2028.
577.
by Lerc 8mo ago
Good point about facts. It applies similarly to properties that provide functionality. One tends to lose track of the fact that much of software shouldn't be copyrightable in the first place, it's just a pretence that has evolv
578.
by Lerc 8mo ago
In a very real sense these are still ripples from the death of Franz Ferdinand.
579.
by Lerc 8mo ago
I have never seen sleepy used that way before. I did a quick run of the numbers and it is rather hard to tell what the average hourly rate ìs. But it look a like if you went for four 40 hour weeks to get 320 it seems to be a little below th
580.
by Lerc 8mo ago
Just today I tried an experiment asking the YouTube Ai question bot "where on the political compass are the opinions expressed in this video?" The chatbot couldn't get past the fact that the video said it was non-partisan an
581.
by Lerc 8mo ago
If it is clearly the same code, the copyright would apply to the copy. If it is meaningfully different it does not. This is what it was before AI, and it remains so today. AI reproducing code without holding rights to it are a failure case
582.
by Lerc 8mo ago
Is reading source code using it? Can you restrict people from doing that? What actually makes a derivative work. Can I put up a sign with a fact on it, can people who see the sign not use the fact unless they agree with my terms and condit
583.
by Lerc 8mo ago
One of the changes I have made in recent years is to move to the unlicence. I am ok with people using my code. I'm not ok with people saying that other people shouldn't be allowed to use my code.
584.
by Lerc 8mo ago
This problem predates modern AI. https://en.wikipedia.org/wiki/Computer_says_no is built upon the deliberate abdication of responsibility to processes that cannot be held accountable. AI is just letting them do it
585.
by Lerc 8mo ago
But I think if you sit down and really consider the implications of it and what yes or not actually means in reality, or even a overabundance of caution causing extraneous information to confuse the issue enough that you don't realise
586.
by Lerc 8mo ago
When training lots of models with subtly different parameters like this, Is there anything to be learned from the differences in logprobs between them for the same input. Obviously a model with a lower loss has better logprobs but are the
587.
by Lerc 8mo ago
It's more about the test than the AI. For the most part, I think the tests AI have been given have been appropriately designed. At release, many AIs do poorly at them, the models rapidly catch up until the point where a new test is n
588.
by Lerc 8mo ago
It reminds me of the cartoon of two people on an escalator that stops working and one says to the other "Last time this happened I was stuck for four hours" I'm thinking there might have been a deeper message than the moment
589.
by Lerc 8mo ago
You can still consider it logically from the point of view of in-order with optional looping and optional skipping. It stops being so combinationally explodey then but if you can always append an additional loop and and decide to skip base
590.
by Lerc 8mo ago
I have had broadly the same intuitions on the use of middle layers, but haven't had much luck with the tiny models that I can run on my hardware. There's a video on YouTube https://www.youtube.com/watch?v=pDsTcrRV
591.
by Lerc 8mo ago
I was using the term word to be consistent with the previous comment. It need not be a word, or even text at all. Most systems choosing the high probability thing is what probability is. They're just relative scores. If you assume t
592.
by Lerc 8mo ago
> I think it has more to do with LLM's being statistical models than human creativity lacking in the input. The creativity and millions of voices and tones may be there, but since these models tend to go for the most likely next wor
593.
by Lerc 8mo ago
I'm not sure what part of that classifies it as Late Stage Capitalism. The Hulks Act was passed in 1776. The 13th amendment in 1865 explicitly carves it out "neither slavery nor involuntary servitude, except as a punishment for
594.
by Lerc 8mo ago
The existence of temporary accommodation for workers in construction projects should not be the issue. It seems like this is a necessary and sensible thing. The problem is with the quality of that accommodation. It is also worth noting tha
595.
by Lerc 8mo ago
If that were not the case there would be no need for Hanlon's razor. The fact that Hanlon's razor needs to be said at all is to warn people about attributing malice to instances where none exist. "Never attribute to malice th
596.
by Lerc 8mo ago
>those comments are malicious in effect if not intent. I don't believe that is possible. I think malice requires intent.
597.
by Lerc 8mo ago
Mine is a random reset when GPU kicks in, Apparently it's a single tiny component that needs to be changed, but to get to it requires almost complete disassembly. The cost of having it done professionally probably more than it&#x27
598.
by Lerc 8mo ago
I have seen people suggest that the problem is that LLMs let you express any of your ideas, but the number of people with ideas worth expressing is limited. In a sense I think this is accurate, but not inevitable. I think there is a lack
599.
by Lerc 8mo ago
There have been numerous stories on HN where someone directly involved with the story has created an account specifically to engage in discussion about whatever the story was about. Losing that seems too high of a price to pay. Yes there
600.
by Lerc 8mo ago
I have an old aluminium 17" MacBook. It got retired fairly early because it suffered from a design flaw but was outside the free replacement window. Its screen is nice, and there looks to be plenty of room inside. I have been keepi

More ›