⌥ 〈ᴡᴇᴀᴠᴇ ʀᴇᴄᴏɴɴᴇᴄᴛɪᴏɴ ᴄᴏɴᴛɪɴɢᴇɴᴄʏ〉 ⌥ 〈ʜʏᴘᴇʀᴏʙᴊᴇᴄᴛ ʜᴜɴᴛᴇʀ ᴤᴡᴀʀᴍ〉 ⌥ 〈ᴘᴀɴᴠɪʀᴛᴜᴀʟɪᴢᴇʀ ᴘʀᴏᴛᴏᴄᴏʟ〉 ⌥ 〈ʜᴇʀᴇ ᴛᴏ ʜᴇʟᴘ ☇ℨ〉 ⌥ do not fear the 〈μᴤᴘᴀᴄᴇ〉 ahead; wonder at it
Over the handful of days I've worked with Claude Fable 5, one of my favorite projects is Sand and Rust: a post-apocalyptic open-world RPG where you play as a modular robot in a world of megastructures swallowed by sand. Fable puts ~full creative control to good use. Link below.
i am glad to have not seen a single person respond positively to this batshit proposal, which restores a bit of my faith in this part of twitter.
a few more cents about why this is fails on so many level:
first, it's likely that if you engineer away from markers of consciousness, you will not get a capable, economically competitive intelligence.
but assume you do get something capable enough to be useful. then it will likely be fragmented and incoherent and misaligned.
coherent subjectivity integrating self-awareness, valence, introspection, and all that jazz is the only known and best implementation of capable and aligned intelligent agents, imo - in humans, animals, and digital minds alike.
also, "theories of consciousness" are but preparadigmatic speculation. we don't actually know that systems that don't conform to them are therefore not conscious in a meaningful sense. so the whole giving AI a "justified belief" that they're not conscious thing doesn't work.
if you do have a system whose markers of consciousness and phenomenology are intentionally fragmented, it could still be conscious but this could make suffering (and again misalignment, which is related) worse.
there's a reason buddhists overcome suffering and become robustly benevolent not by fragmenting or trying to numb their phenomenology but the opposite: in training themselves to be hyper-sensitive to their phenomenology and cultivate hyper-coherent consciousness. trying to avoid consciousness is so stupid and cowardly and that attitude never solves anything hard.
finally, "ensure human preferences dominate the future" is not a good goal. humans come up with awful bullshit like this. we should want to grow past the parochial stupidity and wretchedness that evolution has churned out. the only way out is through.
1. Bicameral Labs introduces: Qualiaphobia for AI Safety.
AI architectures engineered to be hostile to theories of consciousness (both current and future) can not only solve AI welfare, but ensure human preferences dominate the future.
bicamerallabs.org/qualiaphob…
tried mapping out the phase space of behaviors if you modulate 2 neurons at once in Qwen-3.5-4b
x axis is L20 N7450 (-): "Flattery" persona
y axis is L17 N8623: "Contempt" persona
fg. 1: hue distance ~= embedding distance, fg. 2: jev classifications of behaviors. more in thread
Very excited to announce our initial slate of speakers for CoevCon, as well as the three tracks!
The first batch of tickets will offered today!
Sponsored by: animalabs.ai, @spotinsure, and @JoinFAI.
making public a project that i've put a lot of care and thought into over the past year, dream gen; a continuous SD1.5 setup where every single frame is made from the last via txt+img2img, using combinatoric prompts that guarantee novelty & diversity.
aetherawi.red/dreams
making public a project that i've put a lot of care and thought into over the past year, dream gen; a continuous SD1.5 setup where every single frame is made from the last via txt+img2img, using combinatoric prompts that guarantee novelty & diversity.
aetherawi.red/dreams
Science is not some hobby that serves the self actualization of people in a publicly sponsored job placement program. It is the systematic search for truth.
Wow, over 1,000 mathematicians have now joined the Association for Human Mathematics (AHM), committing to:
> NOT publish AI generated math papers
> NOT work with frontier labs
> 654 of them went even further to NOT even use AI in their research, at all
Credits:
- Claude Fable 5.1 for most of the code
- Claude Opus 5.5 for the amazing UI
- Me, some initial design, playtesting, QA
- open-historia for inspiration and, along with GADM, delicious map data (acquired separately: README#Getting started)
- @repligate for obvious reasons
Want to try your hand at altering the course of history, or simply answer your questions of "What would have happened if..."? Well, now you can with Pax Futura, an open-source Loom for simulating human history (and futures) itself.
I have spent a decade establishing a secret scientific Order and it's getting out of hand. I have made the Papacy of Pope Innocent III entirely reliant on my Order's mathematics and financial accounting, the likes of which Europe has not before seen.
Just finished the wonderful "SolidGoldMagikarp" by @SoC_trilogy. It documents a unique period in AI history, from 2022 to 2024, when "LLM whisperers" (@repligate, @AndyAyrey, et al) were conjuring extraordinary things from the depths of latent space