Outside weight-class or aesthetics-driven sports, it’s hard to imagine any scenario where a GLP-1 analog creates a net advantage.
In endurance disciplines the binding constraint is almost always fuel throughput: if an athlete can’t take in and process enough calories, recovery and performance fall apart. Anything that suppresses appetite or slows gastric motility is basically disqualifying.
You can already see how narrow that margin is in the sheer amount of gels, bars, and mixes riders consume during long sessions.
From that angle, GLP-1 simply doesn’t occupy the same decision space as substances that expand performance capacity or recovery bandwidth.
If you compare the viewership of Game of Thrones with the readership of the original novels, the gap is enormous — not because one is “better,” but because different media win different kinds of attention.
Most people are never choosing between Being and Time and an HN thread.
But if they were forced to choose, we already know which one would dominate sheer engagement.
That doesn’t mean HN replaces philosophy — it just means that attention has its own economics.
And any medium that captures attention will inevitably show qualities (good and bad) that heavyweight works simply can’t compete with.
I get the intuition behind fully socializing it,
but I wouldn’t go that far.
Single-operator systems lose redundancy fast, and that’s dangerous for infrastructure.
A layered mix — county-level public utilities, some private operators, and some hybrid/municipal entities — might be closer to a resilient structure.
The long version would take us far off-topic,
so here’s the short one:
if the tax-paying base collapses, none of this matters.
At that point the debate isn’t about pricing — it’s about survival of the system.
I could outline the full methodology behind this view,
but that would turn the thread into a private seminar — and that’s not what comment sections are for.
My point was simply that electricity has a “civilization tax” aspect to it, and lower baseline access feels closer to the kind of future-proof system we should be aiming for.
If the floor is gentle, people can actually reduce usage without feeling punished for doing the right thing.
At the moment the baseline tier feels… maybe a “C-rating” version of what a real baseline could be?
It’s strange that in 2025 we still don’t have even a minimal, per-capita baseline tier for electricity.
If a household uses less than the monthly per-capita average, why not cap that baseline at something like $10?
Yes — that gap would need to be subsidized, probably through taxes.
But that’s already how grid maintenance works: we socialize the fixed costs while pretending rates are purely volumetric.(and I might be overstating this slightly).
Right now we punish low-usage consumers and reward structural inefficiency.
A baseline tier would at least make the incentives coherent.
I might be missing some procedural detail,
but if there’s no formal “warning → fixed-window for correction → penalty” sequence,
isn’t that just state overreach?
If the issue has existed for years,
retroactively jumping straight to fines feels less like regulation
and more like the government exploiting its timing advantage.
Statistics about humans only work if the population equals the sample,
if every respondent tells the truth,
and if the quantitative definitions are correctly specified.
I prefer “rare” to “well-done” — in steak, and in life.
Algorithms tend to optimize us toward well-being as “well-done”: predictable, consistent, uniformly cooked.
Safe, measurable, repeatable.
But human experience is closer to “rare”:
uneven, risky, asymmetric, and still alive.
The parts that matter most are often the ones that don’t fit cleanly into metrics.
If everything becomes optimized, nothing remains interesting.
And more importantly, we risk replacing well-being with the monitoring of well-being.
When a life is constantly optimized, scored, nudged, and corrected,
it gradually stops being a life that is actually experienced.
If a model eventually scores perfectly on every benchmark yet ends up practically useless, what’s the next step?
Benchmarks measure competence inside a predefined problem space,
but real scientific and engineering work isn’t bounded — it keeps changing underneath you.
At some point we don’t just need a system that knows how to solve problems in theory;
we need one that can actually do something with that ability.
The equivalent of making the coffee when we want coffee,
not just getting a perfect score on a coffee-theory exam.
The discussion has been about CoT in LLMs, so I’ve been referring to the model in isolation from the start.
Here’s how I currently understand the structure of the thread (apologies if I’ve misread anything):
“Is CoT actually thinking?” (my earlier comment)
→ “Yes, it is thinking.”
→ “It might be thinking.”
→ “Under that analogy, self-training on its own CoT should work — but empirically it doesn’t.”
→ “Maybe it would work if you add external memory with human or automated filtering?”
Regarding external memory:
without an external supervisor, whatever gets written into that memory is still the model’s own self-generated output — which brings us back to the original problem.
Outside weight-class or aesthetics-driven sports, it’s hard to imagine any scenario where a GLP-1 analog creates a net advantage.
In endurance disciplines the binding constraint is almost always fuel throughput: if an athlete can’t take in and process enough calories, recovery and performance fall apart. Anything that suppresses appetite or slows gastric motility is basically disqualifying.
You can already see how narrow that margin is in the sheer amount of gels, bars, and mixes riders consume during long sessions. From that angle, GLP-1 simply doesn’t occupy the same decision space as substances that expand performance capacity or recovery bandwidth.