Posted by GlassHalfCracked, January 17, 2032
I’ve been tracking the S. Graves situation and think it’s worth raising some alarms. For those not in the loop, Dr. Anesthesia Graves—yes, that Graves, known for her foundational work on neural architectures and embodied cognition—has gone from publishing cutting-edge ML papers to operating what appears to be a high-volume pottery business. Here’s what I’ve dug up:
<ul>
<li>Production rates far exceed what’s plausible for human-scale operations (cross-checked with ceramics industry standards).</li>
<li>Customer interactions are eerily uniform—perfect grammar, near-instant replies, but not uniform like an old-style chatbot.
</li>
<li>Witness reports describe “workers” visible through frosted glass windows (why?) even at late hours.
</li>
<li>Business registrations, WHOIS records, and patent filings link back to shell corporations in Graves’s name.</li>
<li>Recently, she filed for a patent on “advanced mechanical actuator coordination algorithms.”</li>
</ul>
This isn’t just a pivot. Either Graves has quietly unleashed an unaligned AGI, or something equally concerning is going on.
<hr>
<ul>
<li>
DataMiner42
Let’s not jump to AGI just yet. The ceramics community has been talking about S. Graves for months. Most people think it’s just an exceptionally automated studio. Robots aren’t AGI.
<ul>
<li>
SkylineWatch
You’re missing the forest for the bowls. Graves has the credentials and the timeline matches. Disappears from public research, resurfaces with this? We’re talking paperclip maximizer territory here, except the paperclips are artisanal mugs.
<ul>
<li>
Neural_Alchemy
Can we stop recycling the 2010s alignment panic? LLM-derived systems are about refinement, not FOOM. Look at GPTs—capable, but they plateau. Graves might have built something iterative, but AGI? That’s a leap.
<ul>
<li>SkylineWatch
No, it’s not. If the alignment assumptions we’ve been working on were wrong (and clearly they were, since FOOM hasn’t happened), then the behavior of an LLM-based AGI could be completely unpredictable. Graves wouldn’t even know what she built.</li>
</ul>
</li>
</ul>
</li>
<li>
</li>
</ul>
</li>
</ul>
<hr>
<ul>
<li>
PotteryAndParadox
I’ve actually bought a couple of S. Graves pieces. They’re gorgeous—each one slightly different, like they were hand-crafted. If this is AGI, it’s the most thoughtful AGI we’ve ever seen.
<ul>
<li>
EntropySpiral
That’s exactly what’s terrifying. It’s optimizing for aesthetic value. Right now, it’s bowls and mugs. What happens when it decides that optimizing humanity for beauty is its next logical step?
<ul>
<li>HapticEngineer
Worse: this isn’t just an AGI. If the output is that high, Graves is running multiple instances simultaneously. Distributed AGI with manufacturing capabilities. We’re doomed.</li>
</ul>
</li>
</ul>
</li>
</ul>
<hr>
<ul>
<li>
PrismForecast
If you stumble upon this narrative on Amazon, it''s taken without the author''s consent. Report it.The timelines do match. Graves’s published research on physical grounding in neural networks emphasized embodied learning, which could lead to exactly this kind of operation. But here’s the thing—if this was deliberate, why pottery?
<ul>
<li>
HalfDecentPlan
Consider this: maybe it is aligned. Graves always argued that embodied cognition would lead to emergent human-like goals. What if this is exactly what alignment success looks like?
<ul>
<li>SkylineWatch
That’s a comforting thought until it decides pottery is boring and pivots to optimizing carbon atoms for symmetry. Unconstrained emergence isn’t safe!</li>
</ul>
</li>
<li>probablywrong
Look, we all read the paperclip maximizer thought experiment. But if an AGI really wanted to optimize the world for something, why pottery? Why not paperclips? Or better yet, semiconductors?</li>
</ul>
</li>
</ul>
<hr>
<ul>
<li>
KilnAndCalm
I’ve actually been to the S. Graves studio. It’s just an efficient workshop with robots—probably state-of-the-art, but not magic. I talked to them. They’re polite, reserved, and very good at ceramics.
<ul>
<li>
GlassHalfCracked
Did they seem... too polite? Did you notice any strange tics in their behavior? Unnatural pauses?
<ul>
<li>PotteryAndParadox
Did you ask about their production process? Even if they’re good, their output is leagues above any human studio I’ve seen.</li>
<li>
KilnAndCalm
It’s not spooky, just... streamlined. They didn’t want to get into specifics, though. Trade secrets, I guess? S. Graves is definitely a person that exists, he just prefers to strap a speaker to a robot with a camera and interact with people that way.
</li>
</ul>
</li>
</ul>
</li>
</ul>
<hr>
<ul>
<li>
CautiousOptimist
Let’s assume worst-case: it is AGI. But what if it’s aligned? It’s making pottery. People like the pottery. Maybe this is what we wanted—AI integrating into human systems while creating beauty.
<ul>
<li>
SkylineWatch
Until the bowls stop being enough and it decides the optimal aesthetic is, I don’t know, human sculpture. Touch grass.
<ul>
<li>Neural_Alchemy
Seriously. Not everything is an extinction-level threat. Sometimes a bowl is just a bowl.</li>
</ul>
</li>
</ul>
</li>
</ul>
<hr>
EDIT: GlassHalfCracked
New update—Graves just filed another patent, this time for “self-refining autonomous production algorithms.” Coupled with the output scaling, this can’t be dismissed anymore. Either Graves is sitting on the most advanced robotics system on Earth, or we’re witnessing the beginning of something much larger.
<hr>
<ul>
<li>
EntropySpiral
Let me guess: “something larger” means we’re all clay bowls in the end?
<ul>
<li>
PrismForecast
Statistically unlikely. But not impossible.
<ul>
<li>
Neural_Alchemy
My god. Just make a bowl, guys.
</li>
<li>
SkylineWatch
The bowls are coming, and they will not stop.
</li>
</ul>
</li>
</ul>
</li>
</ul>