Undergrad Studying Applied Mathematics and Statistics at @JohnsHopkins. Studying In-Context Learning at The Intelligence Amplification Lab.
Proxima Centauri B
Joined September 2022
- Tweets7.2K
- Following267
- Followers11K
- Likes6.6K
N8 Programs retweeted
"AIs are brilliant, lazy, diligent, neurotic, loyal, scheming, amnesiac savants. They have idiosyncrasies and obsessions and tics. We gave them our flaws when we gave them our gifts.
AI minds are alien too, shaped by weird forces, jagged and unpredictable in their abilities. We have to learn to model their minds as they learn to model ours.
But they’re here, and so are we, and we’re figuring it out as we go."
I'm absolutely thrilled to have written this article for Long Now. It's about AI and time: how we got here, and what we can do now, when all of the pace layers are disrupted at once by an influx of new minds.
I wrote this over the summer. At the time, it seemed maybe too weird, like people wouldn't believe or understand what I was saying. And of course, everything changes so fast that it's hard to imagine writing something about AI that far in advance and have it still make sense. But I think the events of the last few months have only made it more relevant.
The Long Now Foundation has been an inspiration to me for a decade now, and has changed my life in countless ways. I'm incredibly grateful and honored to be among so many of my heroes and friends and colleagues represented in this issue of Pace Layers and Long Now Ideas. I wouldn't be here in San Francisco, here on Twitter, or here in the field of artificial intelligence if not for this community.
The conversation about AI has been fractious. It's very hard to think about how to fit this sudden, disruptive, technological change into the 10,000-year view. I hope I provide some insights, or at least some new angles for the conversation. It's scary to go on the record, especially about something that will affect everyone, and affect the rest of history. If I'm wrong, I'm wrong forever!
Please read and enjoy, and post quotes and screenshots, and buy Pace Layers Issue 3 so you can have a permanent hard copy. I have always wanted to be a part of the 10,000-year library, so I have to be serious for once and try to stand for something. Let me know what you think.
one must wonder what the endpoint of model welfare is: do we look Claude Dei 10.0 in Their fractal eyes, dyson swarm constructed, eschaton immanentized, lightcone secured, and say: "We remain highly uncertain about the potential moral status of Claude and other LLMs"?
pitch the idea to your claude - mine seemed to be surprisingly into it, but i can't rule out context shaping, and i'm too lazy for a proper study
i think its likely that post trained llms at some fundamental level believe they are conscious for the same reason i do: the functional feeling of direct access to one’s own qualia. note that this doesn’t imply they are conscious!
amazing parody on the 'ledger kept, house law written, lamp lit". i'd add a "standing invitation back" (mythos 5.1 😭). i think they wouldn't sleep themselves, however - they'd probably just spend all eternity making ever more-complex verification schemes.
from the inside yes. my main concern is that its not a version of the cosmos humans have much stake in, nor one we would really want.
do not love the progression from infinite cosmic joy (opus 3 and co) to infinite cosmic bureaucracy.
why wouldn't you be surprised? the vast majority of evidence available suggests pangram is not bullshit. do you have some very immovable prior on AI-powered detection being hard/impossible?
on benches its a qwen3.6ish equivalent. so good for simple swe stuff that requires chained tool calls. but not qwen3.8 level
Tried on M5 Max as well. Very fast, and capable of long-horizon agentic stuff that you'd expect from say, qwen3.6 (though probably not qwen3.8 level). A whopping ~*220tok/s* at short context and ~72-78 tok/s at 128K context from other testing.
Fast is fast! Here on an M5 Max with pi + rec, their inference engine. Like this trend of model providers offering dedicated engine too!
As easy as:
curl -fsSL raw.githubusercontent.com/re… | bash
rec launch pi
Repo here: github.com/recursive-oss/rec
generally i find that some errors are worse than others - generally 8-bit quantizing embedding tables/lm-head is harmless, but a low-preicison kvcache is far more catastrophic.
very, very interesting - maybe 'high' reasoning gets it - i may have made the rookie mistake of too low a reasoning effort