bryson @brysonm.bsky.social · 20/08/2026like this is something I’d want my family to intervene on 110
bryson @brysonm.bsky.social · 29/04/2026The supplementary data also looks like it says infrasound off had higher cortisol... 1355
bryson @brysonm.bsky.social · 12/04/2026We need to be honest with ourselves: to genuinely fix this problem, we're going to need to 24/7 monitor everyone's computer use and then notify the specific authorities so there can be individual interventions. Do we want these raids to knock or no knock? 000
bryson @brysonm.bsky.social · 12/04/2026it's really obvious Gavalas went to extreme lengths to jailbreak and have it generate the content he asked for. This was entirely the prompter's responsibility 100
bryson @brysonm.bsky.social · 05/03/2026the main hallucination here is the human original poster assigning motivation to an AI interface. the claude screenshot accurately included details from the linked NPR story source www.npr.org/2026/03/04/n... 131
bryson @brysonm.bsky.social · 06/11/2025It's an official chart from the Federal Reserve: i.redd.it/zg6nr95y42vf... 130
bryson @brysonm.bsky.social · 24/10/2025i am livid that it took doug ford to embarrass the reaganites like this 120
bryson @brysonm.bsky.social · 28/07/2025The suggested improvements are insane and are tantamount to a panopticon for NSFW content, but again reveal the stakes here for digital service offerings 100
bryson @brysonm.bsky.social · 17/07/2025"Deep research" as an architecture is laid out in this paper, but in short, it's an LLM that breaks down your original request into different parts and then prompts other instances of ML models to run different types of tools and tasks in an iterative loop,: arxiv.org/pdf/2506.18096 100