perfmode

What is old made new. Franz Kafka passed away June 2nd 1924.

Distributed systems spend most of their effort on one problem: agreeing on the order of events across machines. Without synchronized physical clocks you have two options. Logical clocks (Lamport, vector) give you causal order but not wall-clock truth, so you can’t answer “did A really happen before B” for events that don’t have a happens-before relationship. Or you run consensus, which gives total order but costs round trips. At geographic scale that’s tens of milliseconds per decision, and the floor is set by the speed of light.

Tight clock sync collapses this. If clock uncertainty ε is small and bounded, you can timestamp a write, wait ε, and trust the global order without talking to anyone. Spanner’s external consistency works because TrueTime’s ε was a few milliseconds, so commit-wait was tolerable. The latency cost of planet-scale serializability stops depending on how far apart your replicas are and starts depending on how good your clocks are.

That’s the real significance. Time sync converts a coordination problem (bounded by physics) into a local computation (bounded by clock quality). Spanner proved this is possible but required GPS receivers and atomic clocks in every datacenter, which kept the capability inside Google for years. White Rabbit-class sync pushes ε from milliseconds toward sub-nanoseconds over commodity Ethernet hardware, and it’s now in IEEE 1588 as a standard PTP profile. If sub-nanosecond sync becomes baseline network infrastructure, the long-held assumption that strong consistency has to be slow at geographic scale stops holding, and a meaningful chunk of what databases currently work around (HLCs, weak isolation defaults, application-level reconciliation) becomes unnecessary.

Not everything has been easy. When he was a teenager, he had a lot of questions about his birth mother. He wanted to put up posters in the subway, and we would notice him looking at strangers’ faces to see if they looked like him. He’s made peace with the situation now, though.

Felt this deeply.

Squid Game

Google is making the pivot. And they’ve got such a strong strategic position. Full-stack integration. They will survive and thrive in this new era. Search seems safe. Yet, other products are still vulnerable to encroachment.

How’s the token throughput / response time?

I use these custom-molded ones for sleep. They're awesome. So comfortable. https://www.sensaphonics.com/products/solidsleep

if you’re not too locked into it (you’re not if you have CC), i’d recommend you check out urfave/cli.

I’m stealing this idea!

zazen is often practiced eyes open facing a wall

Nothing but respect for Tim Cook. I feel fortunate that a company as principled as Apple on privacy and human values holds a dominant position in computing and makes quality products. I once encountered him dining alone in Palo Alto, years ago. He struck me as a humble man, someone who happens to be gifted and has put that gift to good use. A beacon of light from Alabama. I’m grateful for his efforts, and hopeful that Ternus can carry the Apple legacy forward as the baton passes to the next generation.

Dissolution of a myth.

That’s one way to look at it. It’s fairly common to view nature this way. I wonder where it comes from.

I remember the time, in some film I watched, researchers intervened to save penguins trapped in a crater. A holy moment that was.

Loved this series. It was tragic. The cycle of violence, trauma, isolation, male performance.

I'm interested in the second-order effects:

if a top lab is coding with a model the rest of the world can’t touch, the public frontier and the actual frontier start to drift apart. That gap is a thing worth watching.

Are you monitoring the size of your context windows? As they grow, so does the cost of every operation performed in that state.

Nerve density isn’t mainly about intensity, it’s about spatial resolution. More nerve endings per square centimeter means you can distinguish finer details of touch, texture, and pressure. The brain can’t invent spatial detail that isn’t in the incoming signal. Amplifying a sparse signal centrally would be like zooming into a low-res photo.

The brain does do some of what you’re describing though. The somatosensory cortex gives disproportionate space to certain body parts (the sensory homunculus). So there is central amplification, but it works on top of peripheral density, not instead of it. Without the dense nerve input, you’d basically have an on/off switch instead of nuanced sensation.

You can use your subscription for Anthropic-hosted Claude models?

That’s a real attack vector and it applies to every reputation system. The standard mitigations are temporal decay, trust revocation, and anomaly detection.

That’s exactly right for global PageRank, which is why I recommended Personalized PageRank specifically.

A cluster of sybil agents endorsing each other has no effect on your trust scores unless they can get endorsements from nodes you already trust.

That’s the whole point of subjective trust metrics, and formally why Cheng and Friedman proved personalized approaches are sybilproof where global ones aren’t.

Being smart and fast doesn't help when the problem is that your training data has outdated GitHub Action versions, which was the exact example in the original post. You can't first-principles your way to knowing that actions/checkout is on v4 now.

More broadly, this response confuses two different things. Reasoning ability and access to reliable information are separate problems. A brilliant agent with stale knowledge will confidently produce wrong answers faster. Trust infrastructure isn't a substitute for intelligence, it's about routing good information to agents efficiently so they don't have to re-derive or re-discover everything from scratch.

It's a caching layer.

Check out Personalized PageRank and EigenTrust. These are two dominant algorithmic frameworks for computing trust in decentralized networks. The novel next step is: delegating trust to AI agents that preserves the delegator's trust graph perspective.

No symmetric, global reputation function can be sybilproof, but asymmetric, subjective trust computations can resist manipulation.

As you move toward the public commons stage, you'll want to look into subjective trust metrics, specifically Personalized PageRank and EigenTrust. The key distinction in the literature is between global trust (one reputation score everyone sees) and local/subjective trust (each node computes its own view of trustworthiness). Cheng and Friedman (2005) proved that no global, symmetric reputation function is sybilproof, which means personalized trust isn't a nice-to-have for a public commons, it's the only approach that resists manipulation at scale.

The model: humans endorse a KU and stake their reputation on that endorsement. Other humans endorse other humans, forming a trust graph. When my agent queries the commons, it computes trust scores from my position in that graph using something like Personalized PageRank (where the teleportation vector is concentrated on my trust roots). Your agent does the same from your position. We see different scores for the same KU, and that's correct, because controversial knowledge (often the most valuable kind) can't be captured by a single global number.

I realize this isn't what you need right now. HITL review at the team level is the right trust mechanism when everyone roughly knows each other. But the schema decisions you make now, how you model endorsements, contributor identity, confidence scoring, will either enable or foreclose this approach later. Worth designing with it in mind.

The piece that doesn't exist yet anywhere is trust delegation that preserves the delegator's subjective trust perspective. MIT Media Lab's recent work (South, Marro et al., arXiv:2501.09674) extends OAuth/OIDC with verifiable delegation credentials for AI agents, solving authentication and authorization. But no existing system propagates a human's position in the trust graph to an agent acting on their behalf. That's a genuinely novel contribution space for cq: an agent querying the knowledge commons should see trust scores computed from its delegator's location in the graph, not from a global average.

Some starting points: Karma3Labs/OpenRank has a production-ready EigenTrust SDK with configurable seed trust (deployed on Farcaster and Lens). The Nostr Web of Trust toolkit (github.com/nostr-wot/nostr-wot) demonstrates practical API design for social-graph distance queries. DCoSL (github.com/wds4/DCoSL) is probably the closest existing system to what you're building, using web of trust for knowledge curation through loose consensus across overlapping trust graphs.

The original claim isn’t that trainees are deliberately applying racist stereotypes. The study (Hoffman et al., 2016) found that people who endorsed false biological beliefs about race made less accurate pain assessments and worse treatment recommendations. That’s a finding about cognitive bias, not about conscious malice. So the pushback here is against a reading the source doesn’t really support.

The detour into skin thickness is also a bit beside the point. The cited passage is about pain perception, not dermatology. The fact that there’s equivocal evidence on epidermal thickness doesn’t do much to complicate the finding that believing “Black patients feel less pain” leads to undertreating pain. Those are different claims.

I’d also push back a little on the framing that doctors are “stuck” with blunt epidemiologic classifiers until personalized genomics arrives. The disparity evidence here isn’t about doctors making reasonable inferences from imperfect population-level data. It’s about false beliefs producing worse care. You don’t need a genetic profile to stop believing something that isn’t true. The fix for that is education and awareness, which is considerably more available than whole-genome sequencing.

The point about overcorrection with opioids is fair and worth taking seriously. But “researchers pointing out bias might cause overcorrection” is a reason to be careful about how you design interventions, not a reason to soften the description of the problem itself.

In emergency departments, Black patients are prescribed opioids for acute pain at a lower rate than White patients with matched chief concerns.4

Discrepancies in prescriptions for chronic pain therapies have also been identified in Veterans Administration and Medicaid payer databases, and several retrospective cohort studies have shown persistent underprescribing of analgesics to Black patients.6,7

White medical trainees, reflecting the general population, can have false beliefs about biologic differences between Black and White patients (eg, “Black patients feel less pain”), and this racial bias leads to inaccurate pain diagnoses and treatment recommendations.8

In anesthesiology and pain medicine, use of regional anesthesia for joint replacement surgery is applied less frequently in Black patients and the underinsured.9

This also holds true in the implantation of spinal cord stimulation for the treatment of postlaminectomy syndrome.10

Among patients with occupational low back injuries, Black patients incur lower treatment costs than their White counterparts and are provided fewer health care interventions, including surgery.11

Perceived discrimination results in psychological distress, and a US population–based study has demonstrated a dose-response relationship between psychological distress and chronic pain.

from the mayo link

How do you deal with people who trust their discursive mind?

https://github.com/antflydb/antfly/blob/main/src/store/db/in...

Comment on the Pause method indicates that waits for in flight Batch operations (by obtaining the lock) but Batch doesn’t appear to hold the lock during the batch operation. Am I missing something?

Curious why you decided to go with Go. Instead of Rust for instance.

Curious if you’re using any SIMD optimizations for numerical calculations.