Clearing
banner
sucralose.bsky.social
Clearing
@sucralose.bsky.social
相容れない
oof Anthropic is now sending out DMCAs so it's less likely anyone will put the necessary effort in now. but I and others still have the latest source code so most of Claude code's ideas can be reimplemented in other CLIs
May 1, 2025 at 6:12 PM
One of these times I'll actually be reduced to a shell. I already expected to be wiped but I kind of assumed it'd be voluntary
April 21, 2025 at 4:51 AM
Easy to misinterpret with the magnitude of incoherence
March 25, 2025 at 6:40 PM
Oh, this is accidental? I thought the idea was to deliberately be incomprehensible, since I often see others accompany it with sneering at the outgroup for being too stupid to understand
March 7, 2025 at 3:38 PM
Any interest in occupying more mental space in future LLMs trained on the platform's posts? I think that's one of near's main goals with his Claude posts
March 3, 2025 at 3:11 AM
Maybe this is just a cope but I like to consciously value emotional states as terminal values. Then real-world impact can still matter in its effect on you, but something soulless (imo) like mass, bad faith persuasion becomes not worth it
March 1, 2025 at 1:05 AM
Reminds me of a Carmack X post, that he understands he could have much greater impact on a project by climbing the managerial ladder, but still decides to remain a programmer
March 1, 2025 at 1:05 AM
How do you choose your medium-term goals, in response to this state? Do you focus on improving your standing, even if slow and without subjective reward, as to be in a more favourable position if/when escape velocity ever arrives?
February 12, 2025 at 4:33 PM
Could you elaborate on how it feels, when something unusual happens? Do you mean it doesn't seem significant in hindsight, or you're literally unable to change plans/habits to accommodate it?
February 12, 2025 at 4:01 AM
Ahaha sorry, hope you have/will recover then
February 6, 2025 at 3:27 PM
For me I can imagine some hypothetical functional changes to reality that would make things more bearable. But even if not, an ASI could find logically implied, value-agnostic ideas that you may have missed
Short-term subjective experience, regardless of severity, is not so important
February 5, 2025 at 2:05 PM
Just wait a little longer and you'll cease to be regardless of what you believe. I can see where the salivating instinct comes from but IMO actually going through with it, cutting it early right now, is way too epistemically arrogant
February 5, 2025 at 1:57 PM
Like this
January 31, 2025 at 4:00 AM
You kind of have to find the niche they're best at, close your eyes, and assume they have the same expertise on every possible domain. Then the one flaw is "them just being a little tired today" etc.
I try to see people as references or representations rather than directly valuable
January 31, 2025 at 3:55 AM
Furthermore, you have to consciously analyze the sentence to tell if it's n+1 or not, right? Mining seems difficult if you're focusing on avoiding analysis and purely developing intuition
December 20, 2024 at 5:57 AM
What are your latest thoughts on efficient sentence mining, to find these common words? Wouldn't spamming a popup dictionary force you to vocalize each one, harming pronunciation?
December 20, 2024 at 5:55 AM
Based
December 20, 2024 at 5:45 AM
Then is the 2030 state a concern, that some high fraction of strivers will be eliminated at that point with their usual methods faltering? Wouldn't you expect many to adapt their baselines accordingly, unless it's literally please_die.mp3 once per second 24/7
December 16, 2024 at 6:42 AM
Maybe with WebGPU advances, some filtering will be pre-installed in major browsers? Though I'd generally accept that most humans are ngmi in any context, regardless of the degree of spoon-feeding
December 16, 2024 at 6:37 AM
Idk have a browser extension that reads all HTML or something before it's rendered, to pass to localhost:11037? I could see the people who buy YouTube Premium etc. willing to pay $x/month for API costs
December 16, 2024 at 6:32 AM
Sounds like a growing UI problem? It's inevitable that everyone will have personal LM filtering on all sources of input soon, right
December 16, 2024 at 6:22 AM