Calavar

> What am I missing?

Several things.

First, private practice docs see patients with very good employer provided insurance, but residents are largely seeing patients that private practices wont see - patients who are far too medically complex to fit into a 10 minute slot and who also have particularly stingy insurance.

So as opposed to a private practice doc who is seeing 30 patients per day and billing an a average of $250 to $300 per patient (certainly not $1000 - that is unrealistic in my experience), a resident is seeing more like 10 to 15 patients per day (30 minute slots) and billing less than $100 per patient.

Second, residents have to be supervised. You have not included the salary of the physicians supervising them in your calculation.

Third, and I have mentioned this many times before on HN, training is limited by chiefly by the number of training sites that can offer quality training. For example, most hospitals will not see a single case of Guillan-Barre in a single year. Would you want to be treated by a nuerologist who trained at such a hospital? This is why neurology training is generally limited to places with a high volume of neurologic cases that would be considered rare at the average hospital, and these hospitals can only accommodate so many residents. Even for general medicine, you probably do not want to be treated by a doctor who trained at a hospital where any case that passed a certain complexity was transferred out to a bigger center.

General practitioners make more like 250k on average. 150k for general practitioners working in academics, but not overall.

I don't think Zelenskyy is a saint - far from it.

However this is one of the weakest possible grievances you could raise because Putin also released rapists and murderers from prison to raise manpower, most famously during the spring 2023 Bakhmut offensive. There is no moral highground on that point.

Actually, this is actually a near universal issue with Russian grievances against Ukraine. Virtually any grievance that Russia can claim against Ukraine, Ukraine can claim that exact grievance back at equal or greater measure.

This is a classic Russian perspective on the war - that Ukraine is a gang of neo-Nazis that are dead set on eliminating Russian speakers from Ukraine by force. Russians will further claim that this gang of Russophobic neo-Nazis is led by Zelenskyy, a Jewish man who speaks Russian as his mother tongue and Ukrainian as a second language.

The absurdity of this claim should be self evident.

> Early termination of a domain registration does not impact the life cycle of the domain, as the domain can still go through the various stages of a standard life cycle.

According to ICANN cutting the life cycle short doesn't have any effect on the life cycle?

ICANN corruption is at the level of FIFA corruption.

An absolute doddle assuming you have an absolutely excellent test suite. Otherwise on any nontrivial codebase you will still have regressions that make it to production and you will waste your time squashing bugs rather than writing new features.

It's a MongoDBism. The MongoDB community used document to mean the nonrelational equivalent of a row in a relational database. But over time there was definitional shift, and now it means a JSON blob, even if that blob is in a relational database.

I am confused as to how this is an argument for a single source of truth. It seems like you've presented an argument for the opposite. If there were a single source of truth, there would be no point in seeking a second opinion - it would be the same as the first. If the first opinion was correct, excellent. But if the first opinion was incorrect (and no AI model is correct 100% of the time), then you're straight out of luck.

SoftRam. Raymond Chen wrote a great technical analysis on what it actually did: https://devblogs.microsoft.com/oldnewthing/20211111-00/?p=10...

> There are a finite number of words and a finite way of combining them within an academic setting/field which means that we can't build a perfect classifier with just text.

Maybe for very short phrases, but otherwise I disagree. Phrasing very quickly runs into a combinatorial explosion. In the words of Noam Chomsky, "Virtually every sentence that a person utters or understands is a brand-new combination of words, appearing for the first time in the history of the universe."

In my opinion, the difficulty in LLM/human text discrimination isn't that a person might coincidentally write exactly the same text as an LLM would, but rather that 1) LLMs aren't hard locked to a single phrasing (so this is a tougher problem than matching to a single static document, e.g. plagiarism detection) and 2) text has relatively low information density, so you need quite a bit of it to gather enough data to run a statistical test with a reasonably narrow confidence interval.

> They'll almost certainly spend more time and money on the process than is ever collected if this ever happens.

The point of regulation isn't for the state to turn a profit. In fact, I'd go as far as to say that regulations that drive a monetary profit for the state are generally bad because they create a perverse incentive. For example, municipal governments adversely affect traffic flow by lowering speed limits because those lower speed limits generate more ticket revenue.

Similar situation. I have no doubt that a large fraction of my childhood memories are fabricated, but there were a few that I kept to myself until I was over the age of 30 and were later confirmed by a parent or aunt or uncle.

There's also the phenomenon of having a memory of a memory. At age 10, I had a very solid recollection of my life at ages 5 to 6 (not so much of age 4). Now all I remember is that I used to remember a lot more than I do know.

This gets said a lot and it kind of irks me. (I am a physician.)

US software devs also make 2x what their European colleagues do, but that never gets called out as bloat. Plus US software devs make that 2x pay without taking our additional loans for medical school at the rate of $75k per year or doing years of low pay residency where their salary doesn’t give them the means to pay off those loans.

I would be more willing to believe you only used Claude for minor editing tasks if you disclosed your usage of Claude upfront.

The value of a PhD thesis is the personal intellectual growth you get from putting it together. The end product isn't really the point.

There's a lot to be said about publishing in academia being broken and how nearly all the value comes from 10% of publications, while the rest are garbage spewed out for reasons orthogonal to the advancement knowledge. However, IMHO, none of that really applies to PhD theses.

To call out The Tribe as hypocritical, you first need The Tribe to have a consensus opinion. Agentic coding in particular has been very polarizing both on HN and in the developer community at large - there is no consensus opinion.

Octopuses are smart, but I've yet to see anything that suggests they are smarter than dolphins or whales.

Brains are resource hungry, especially oxygen hungry. Earth's air is orders of magnitude richer in oxygen molecules than its water. This likely made it easier for intelligence to develop on land. It's worth noting that the smartest aquatic animals are air breathing mammals that spent much of their evolutionary history on land before returning to water.

Gates famously came from a rich family, but Bezos did too - he used hundreds of thousands of dollars in investments from his immediate family members to get Amazon off the ground. Maybe 1 to 2% of Americans would be able draw that much from their family members if they were to launch a startup. If we define "bootstrapped" wealth as starting from an economic background within one standard deviation of the national average, then he doesn't count.

> While fact investigation continues, BAM’s current state of its investigation has uncovered significant evidence of gross negligence in how the store was previously operated by the prior owner

Verbal diarrhea. Excessive passive tense and vagueries to dilute blame (who is the "prior" owner here - Johnson or the woman before they've been trying to scapegoat?) and to top it off multiple grammatical errors in a single sentence.

I can't believe they would let this draft hit the website in its current form this in the midst of what could be an existential crisis for there company.

I believe they mean GPU threads. Plenty of cuda files in their repository.

Your sarcasm is misplaced because yes, this has unironically been true for large chunks of Latin American history.

- Argentinians in particular are over 60% of Italian descent.

- The richest man in Mexico was born to Lebanese immigrants.

- The chief military leader of the Chilean war of independence was born to an Irish immigrant.

- Peru had a president who was born to Japanese immigrants.

These countries have all, at various times, had an influx of overseas immigrants whose birthright citizen children rose to high stations in society.

If you're looking for international precedent, this is an old vs. new world issue. Birthright citizenship is rare in the old world, but it is the default for the Americas. Canada, most of Latin America, and a decent part of the Caribbean have birthright citizenship.

I did look up numbers before I made that claim:

From Yahoo Finance

GME Jan 1, 2016: $7.09, $5.49 adjusted (accounting for dividend disbursements)

GME Jan 1, 2026: $20.09

266% or 365% return depending on how you count dividends. 365% for GME vs. 306% for S&P 500 over the same period (also using adjusted for dividend numbers).

GME also beat the S&P 500 over the past 10 years. Is this evidence that Ryan Cohen is a business genius?

Tesla has been a meme stock for about five years now, maybe more. Its valuation correlates with Musk's abilities as a showman and media figure, not a businessman.

I consider myself AI skeptical-ish and I detest when people defend LLMs with "it's user error, prompt better," but in this case it actually is user error.

If you want a particular implementation approach, you need to specify not only the features you want, but the implementation strategy at least at a high level. This could be as simple as adding "use pywikibit" or "use relevant packages from pypi" to the end of your prompt. Or you could seed your project with some manually writtem scaffolding, including a pyproject.toml

While LLMs do tend have NIH syndrome by default, I think this is a good default. I'd much rather have tight control over when and how to include external dependencies as opposed to letting a prompt fire for 40 minutes, and coming back to find 2 GB of newly installed node packages with a dependency tree 300 levels deep.

> But it’s getting harder and harder to define a task that humans beat LLMs on. On pretty much any easily quantifiable test of knowledge or reasoning, the machines win.

Quite to the contrary, I think it's extremely trivial to find a task where humans beat LLMs.

For all the money that's been thrown at agentic coding, LLMs still produce substantially worse code than a senior dev. See my own prior comments on this for a concrete example [1].

These trivial failure cases show that there are dimensions to task proficiency - significant ones - that benchmarks fail to capture.

> Is medical diagnosis one of these high judgement tasks?

Situational. I would break diagnosis into three types:

1. The diagnosis comes from objective criteria - laboratory values, vital signs, visual findings, family history. I think LLMs are likely already superior to humans in this case.

2. The diagnosis comes from "chart lore" - reading notes from prior physicians and realizing that there is new context now points to a different diagnosis. (That new context can be the benefit of hindsight into what they already tried and failed and/or new objective data). LLMs do pretty good at this when you point them at datasets where all the prior notes were written by humans, which means that those humans did a nontrivial part of the diagnostic work. What if the prior notes were written by LLMs as well? Will they propagate their own mistakes forward? Yet to be studied in depth.

3. The diagnosis comes from human interaction - knowing the difference between a patient who's high as a bat on crack and one who's delirious from infection; noticing that a patient hesitates slightly before they assure you that they've been taking all their meds as prescribed; etc. I doubt that LLMs will ever beat humans at this, but if LLMs can be proven to be good at point 2, then point 3 alone will not save human physicians.

[1] https://news.ycombinator.com/threads?id=Calavar#47891432

I disagree, I use metaprogramming in application code quite regularly, although I tend to limit myself to a single construct (instance_eval) because I find that makes things more manageable.

In my opinion the main draw of Ruby is that it's kind of Lisp-y in the way you can quickly build a metalanguage tailored to your specific problem domain. For problems where I don't need metaprogramming, I'd rather use a language that is statically typed.

I'm writing a compiler. When I have Claude write a new feature, I have validate that suite against a test suite of ~200 tiny programs.

I have a shell script that automates this. If all tests pass, the shell script prints "200/200 passing" with very little token spend. If only 190/200 pass, the shell script reports the names of every test that failed, and now Claude does a process of

1) run the compiler binary -> 2) get assembly output and inspect for obvious errors -> 3) assemble -> 4) verify that the assembler did not report errors -> 5) run test binary, connect with gdb, and find the issue -> 6) edit the compiler source -> 7) recompile the compiler -> 8) back to 1

multiplied by 10 for the 10 failing tests. This eats up tokens very quickly. I realize that not every use case is going to look like this. But if I didn't have Claude verify against the test suite, then I'd be getting regressions left and right, and then what's the point?

The whole codebase (tests included) is less than 15k lines, so I don't think that's the issue. No MCPs. CLAUDE.md about 1.5k lines.

I'm skeptical of that reasoning because the original C wasn't too clean or performant either. For example emit.c from an earlier commit [1]

It writes a separate call to emit_raw for each line, even though there many successive calls to emit_raw before it runs into any branching or other dynamic logic. What if you change this

    emit_raw(ctx, "#include <stdio.h>\n");
    emit_raw(ctx, "#include <stdlib.h>\n");
    emit_raw(ctx, "#include <string.h>\n");
    emit_raw(ctx, "#include <math.h>\n");
    // And on for dozens more lines

to this

    emit_raw(ctx,
        "#include <stdio.h>\n"
        "#include <stdlib.h>\n"
        "#include <string.h>\n"
        "#include <math.h>\n"
        // And on for dozens more lines
    );

That would leave you with code that is just as readable, but only calls the emit function once, leading to a smaller and faster binary. Again, this is a trivial change to the code, but Claude struggles to get there.

[1] https://github.com/matz/spinel/blob/aba17d8266d72fae3555ec91...