libraryofbabel

Or perhaps, just perhaps, the true higher-dimensional move is realizing that choice of programming language isn’t usually the critical factor in whether a project, system, or business succeeds or fails, and that obsessing over the One True Way is a trap.

It might surprise the author to learn that there are many people who:

1) Have tried lisp and clojure

2) Liked their elegance and expressiveness

3) Have read through SICP and done most of the exercises

4) Would still choose plain old boring easy-to-read always-second-best Python for 90% of use-cases (and probably Rust for the last 10%) when building a real business in the real world.

The article could really benefit from some steel-manning. Remove the cute Flatland metaphor and it is effectively arguing that lisp/clojure haven’t been universally adopted because most programmers haven’t Seen The Light in some sort of epiphany of parentheses and macros. The truth is more nuanced.

Yeah, I thought of this first as well. There is nothing that hammers home the point that the past was a horrible place better than childhood mortality statistics. I’m surprised the author of the article didn’t mention it, given all her focus on families - I mean, good for her for realizing she didn’t understand what life in the past was really like, but she still seems a little focused on “it wasn’t cute” rather than the really big differences.

Related recent HN thread on the Bills of Mortality from early modern London: https://news.ycombinator.com/item?id=46045061

The tldr of my post there is that life before the mass availability of antibiotics after WWII was pretty terrifying.

Question: How would you characterize the response to LLMs across the historical profession as a whole? Do you expect LLMs to lead to major changes in how historians approach research in the next ~5 years, or do you think they will be used by just a minority of people?

242 Episodes on WWII and you're only up to 1940!

(I say that as a compliment, by the way. I love deep historical detail.)

Ex-historian here, now an engineer. Ben is one of the few historians really thinking in depth about the implications of LLMs for historical research and teaching: both the good (wow, they are really great at transcribing difficult handwritten documents now; you can use Claude Code to vibe code up quick visualizations for your research or teaching that would have taken weeks of work before), and the bad (students submitting AI-generated essays). Highly recommended reading.

It's also nice to see a working historian who posts to HN. (If there are any others, please raise your hand!) Our community is richer for the wide variety of non-engineering professions represented here, from medical doctors to truckers to woodworkers to pilots to farmers. Please keep posting, all of you.

Almost everyone here will have seen a movie he was the screenwriter for or contributed to: Shakespeare in Love, Brazil, Empire of the Sun, even bits of Indiana Jones and the Last Crusade.

A lot of folks here will have read either Rosencrantz & Guildenstern Are Dead or Arcadia, which are still probably his best plays, and certainly the best introduction to his work. I personally also really like The Invention of Love, about the gloomy tortured homosexual poet and classical scholar A.E. Housman, which was also apparently Stoppard's favorite of his plays. It's definitely niche territory though and you might need to care at least just a little about A Shropshire Lad and latin textual criticism. The Coast of Utopia is even more packed with history and erudition, although worth a read; the currently top comment here is a quote from it about death, childhood, and the pursuit of happiness.

He had an interesting combination of traits that many HN readers will probably appreciate: erudite to the point of elitism, although never attended college; a self-described "small c. conservative in politics, literature, education and theatre" with libertarian inclinations, but he wrote a sprawling trilogy about 19th-century Russian socialist and anarchist exiles (The Coast of Utopia).

Now that he's dead, I want to go back and re-read all his plays, including the ones I never managed to get to before.

Yeah. It's unfortunate that they seem so intent on digging in their heels about this one issue of water use. Precisely because AI is so important, we should want to make sure we're clear on the facts and have the right context for them! And the larger point that AI is likely to consolidate power in the hands of increasingly few people and corporations is well worth making (although, that story can be told for most critical technologies , from the steam engine to the telegraph to the transistor, and I don't think Hao's framing of Empire/Colonialism is really the right way to look at it at all). I think there are plenty of books to be written about the social impact of AI from a more balanced, empirical, less ideological perspective.

This has been discussed many times on HN, including as far back as 2017: https://news.ycombinator.com/item?id=34925320

The tldr is it's a cute marketing / interior design stunt by Cloudflare. The lamps certainly produce entropy, but you can get good randomness much cheaper and quicker from a Hardware Random Number Generator (HWRNG) that uses thermal noise or radioactive decay, which is what most people do.

A critical analysis of Karen Hao's book Empire of AI, with a response from the author in the comments.

Oh, absolutely. Probably some of the best leatherworkers in China are making high-end handbag knockoffs: it’s where the money is.

Knockoffs? please, we call them “reps” ;)

There are whole subreddits devoted to this, the most well-known being repladies, which went private after it got too famous due to an NYT article. People will spend $1000 or more for a really good Birkin knockoff with high quality leather and hardware. The bags are almost all made in workshops in China. Getting one is apparently (I haven’t done it myself) an interesting exercise in trust and reputation: how do you know the seller isn’t going to send you a cheap knockoff from China rather than a “real” $1000 knockoff? In practice there is a whole world of trusted Chinese middlemen with reviews etc. who have a strong stake in keeping their reputation high in the “reps” community (but you’d better make sure the reviews are real…).

True, but this does not particularly apply to the Birkin, which was famously created for the actress Jane Birkin after she complained to the CEO of Hermes that she couldn’t get a bag big enough to hold both scripts and baby diapers. Sure, it’s not as good at carrying things as a backpack, but it’s not bad either.

It does delight me no end to see a whole thread on handbags on HN. I agree with one of the parent posters though, handbags are an unusual category with long-lived brand status (like cars and watches) and not really comparable to lububus.

It is one of the most striking things that makes life in the premodern past so unimaginably different from today. To be a parent was to suffer continuous tragedy. An example: Queen Anne of Great Britain (1665-1714). She was pregnant 18 times, and not one of her pregnancies resulted in an adult child to inherit her throne. Miscarriages, stillbirths, and five children born alive. Four died in infancy from various causes. One, Prince William Duke of Gloucester, lived to age 11, when he died of an infectious disease - probably smallpox. His death was one of the key events that led to the House of Hannover (the current British royal family) inheriting the throne in 1714.

Smallpox probably killed about a billion people, many of them children, before its eradication in the 1970s due to a global vaccination campaign. You can see it is listed in the Bills of Mortality in TFA, always with a high number of deaths.

True; in fact mortality in 17th-century London was so bad that it required a net flow of migrants just to keep the population stable. Most of all, you were at much higher risk of plague during the plague epidemics.

But rural Europe was no picnic either - you would still be likely to die in all sorts of painful ways, from sepsis, diseases, accidents, childbirth, etc. And my god, it would have been dull.

Honestly, if you forced me to go back to the 17th century, I would probably take the risk and live in London. At least there is the possibility of crossing paths with Samuel Pepys, Christopher Wren, the early Royal Society etc. while you sit in your coffeehouse reading a freshly-printed news sheet.

It’s already been (at least partly) digitized, e.g. https://www.deathbynumbers.org/data/

One thing you will have trouble with however, is that disease categories and the process of determining cause of death changed a great deal from 1527 to 1858. So the categories you're working with aren't stable at all.

I went to grad school for history, and it completely cured me of any nostalgia for or desire to live in the premodern past. The past is an interesting place, but also a place of unmitigated horrors. If you were lucky enough to live into old age in relatively good health, you would have seen many friends die young, and half of your children as well. Just the number of deaths from "teeth" in the Bills of Mortality says it all. These would have been extremely painful worsening tooth infections eventually resulting in sepsis or a brain infection, and death:

> In seventeenth-century London — and for that matter most other places — teeth were a leading cause of death, owing to poor oral hygiene and no effective means to treat infections at a time when extractions — without anesthesia — were performed by the local barber.

I've had at least one infection that would probably have killed me in the 17th century. I am grateful for the existence of antibiotics.

Thanks! I have hearing aids too and, while I don't have profound hearing loss (severe one side, moderate the other), it's got to the point where I am starting to look at other solutions to supplement them. For those not aware: hearing aids are not like eyeglasses, where you can effectively have your vision restored to 20/20 with the right prescription. They help, but they don't get you up to the level of a person with ordinary hearing. This goes particularly for situations with high background noise, multiple voices, etc.: restaurants, bars, parties and other places where social life tends to happen.

Your recommendation has made me think I should finally take the leap on live-caption glasses, which I've been watching for a while now. It sounds like they've finally got to the point where they're effective.

A question: how do you find them in conversations with multiple people? Do the glasses help you distinguish speakers by breaking out text from different voices, or is it all just a stream of words? What if there are multiple conversations going on - do the glasses manage to "focus" captioning on the one you're closest to, or do other conversations creep in?

The captions on google meet have been useful for me as well, although I do wish they were more configurable, e.g. being able to give the captioning model a list of technical phrases, internal company terms, or acronyms that are likely to come up. There also seems to be a gap still between real time transcription, which is good, and after-the-fact transcription, which is excellent. (You can also use an LLM for the latter and prompt it yourself with extra context, although that may not be the best model.)

It's been great to see so much progress in this area in the past few years. I am hoping the current (over-)investment in AI at least has the side effect of improving the tech still further.

I was interested (and slightly disappointed) to read that the knowledge cutoff for Gemini 3 is the same as for Gemini 2.5: January 2025. I wonder why they didn't train it on more recent data.

Is it possible they use the same base pre-trained model and just fine-tuned and RL-ed it better (which, of course, is where all the secret sauce training magic is these days anyhow)? That would be odd, especially for a major version bump, but it's sort of what having the same training cutoff points to?

I expect the book itself (Death in a Shallow Pond: A Philosopher, a Drowning Child, and Strangers in Need, by David Edmonds) is good, as the author has written a lot of other solid books making philosophy accessible. The title of the article though, is rather clickbaity: it’s hardly “recovering” the origins of EA to say that it owes a huge debt to Peter Singer, who is only the most famous utilitarian philosopher of the late 20th century!

(Peter Singer’s books are also good: his Hegel: A Very Short Introduction made me feel kinda like I understood what Hegel was getting at. I probably don’t of course, but it was nice to feel that way!)

The larger issue, of course, is that eccentric individuals and niche special-interest groups are able to use the planning process and the legal system to jam up all sorts of infrastructure projects in America, from simple turn lanes all the way to high-speed rail. This is not the only reason America has trouble building infrastructure, but it is an important reason. See Ezra Klein and Derek Thompson‘s new book Abundance for a long-form analysis of this… or for a contrast with the US’s “lawyerly society” (and, of course, the disadvantages of leaning too much in the other direction) Dan Wang’s Breakneck: China's Quest to Engineer the Future that just came out.

Both are excellent books and will probably appeal to a lot of Hacker News folks with an engineering/builder mindset.

I should have mentioned in my post that I have an applied math masters and a solid amount of analysis and linear algebra with some group theory, set theory, and a smattering of topology (although no algebraic topology). So, I'm not coming to this with nothing, although I don't have the very deep well of abstract algebra training that a pure mathematician coming to category theory would have.

Although, it feels like category theory _ought_ to be approachable without all those years of advanced training in those other areas of math. Set theory is, up to a point. But maybe that isn't true and you're restricted to trivial examples unless you know groups and rings and fields etc.?

Oh thanks, I will take a look. I’ve read some of John Baez’s things but mostly on mathematical physics, which was my undergrad. I didn’t know he’d written on category theory.

I think he and Joan Baez are actually first cousins!

One of the things I liked about her interview was how she candidly says her strengths are less in opening up new areas or proving new theorems and more reworking and clarifying existing areas (i.e. Lurie’s work) with cleaner approaches and new proofs to make them more accessible and therefore more useful.

This seems to me to be admirable, and perhaps under-appreciated. Although it is probably much more valued in mathematics than most other fields, perhaps because mathematicians place more value than other fields on simplicity and clarity of exposition for its own sake, and because it is just so hard to read unfamiliar mathematics. Her north star goal of making her field accessible to mathematics undergraduates was a nice one.

I would like to learn category theory properly one day, at least to that kind of "advance undergraduate" level she mentions. It's always seemed to me when dipping into it that it should be easier to understand than it is, if that makes sense - like the terminology and notation and abstraction are forbidding, but the core of "objects with arrows between them" also has the feeling of something that a (very smart) child could understand. Time to take another crack at it, perhaps?

He mentions he has nobody reporting to him. That sounds like he’s really a staff engineer with a vanity CTO title, plus a lot of sway in strategic decision making.

It’s not a guaranteed recipe for disaster, but it depends critically on his relationship with whoever actually manages the engineering org. If they don’t pull in the same direction, things go south very quickly and you end up with a little civil war.

Either way it’s a red flag and I wouldn’t work there. Another red flag is that he wrote this blog post at all. Given how clearly negative the reaction to it was going to be, it’s a strong signal he doesn’t really think things through and has a ego wrapped up in his “coding” prowess and ability to circumvent process. People mention Woz as an example of a technical co-founder in a non-management role, but he is a humble guy and wouldn’t brag like this.

It's a nice analogy, and I think I'll use it in future.

If you want another one, think of painting. An "Old Master" painter like Rembrandt or Rubens or Botticelli would have had a large workshop with a team of assistants, who would not only do a lot of the work like stretching canvases or mixing the paints, but would also - under the master's direction - actually do a lot of the painting too. You might have the master sketch out the composition, and then paint the key faces (and, most of all, the eyes) and then the assistants would fill in areas like drapery, landscape, etc.

This changed in the Romantic period towards the end of the 1700s, with the idea of the individual artist, working alone in a moment of creative inspiration and producing a single work of genius from start to finish. Caspar David Friedrich or JMW Turner come to mind here.

Some programmers want to be Turner and control the whole work and feel their creativity is threatened if a machine can now do parts of it as well as they could. I'd rather be Rembrandt and sketch out the outline, paint the eyes, and leave the rest to junior engineers... or an AI Agent. It's a matter of preference.

Does anyone know / care to speculate how they actually make this work, in terms of the LLM call loop? Specifically: does it call back to the LLM after each keystroke sending it the new state of the interactive tool, or does it batch keystrokes up? If the former, isn’t that very slow? If the latter, won’t that cause it to make mistakes with a tool it hasn’t used before?

I don't think anyone can reasonably argue against Claude Code being the most full-featured and pleasant to use of the CLI coding agent tools. Maybe some people like the Codex user experience for idiosyncratic reasons, but it (like Gemini CLI) still feels to me rather thrown together - a Claude Clone with a lot of rough edges.

But these CLI tools are still fairly thin wrappers around an LLM. Remember: they're "just an LLM in a while loop with access to tool calls." (I exaggerate, and I love Claude Code's more advanced features like "skills" as much as anyone, but at the core, that's what they are.) The real issue at stake is what is the better LLM behind the agent: is GPT-5 or Sonnet 4.5 better at coding. On that I think opinion is split.

Incidentally, you can run Claude Code with GPT-5 if you want a fair(er) comparison. You need a proxy like LiteLLM and you will have to use the OpenAI api and pay per-token, but it's not hard to do and quite interesting. I haven't used it enough to make a good comparison, however.

What would you say is an example of one of those “middle” tasks it can help with?

I agree with all of this. Many is time I've had to tell developers I work with: "don't just look at the mean/median, look at a graph of the full distribution!... then slice your distribution a lot of different ways by all the tags/facets you have and look again at the slices." Often you find that a shift in the mean or median was driven by one particular class of data points that skewed the whole thing. (Looking at you, NVDA.) This is usually a little lecture I give in the context of performance engineering, where it's api response times or whatever, but it applies everywhere.

At the same time - and I think you agree with this and it's probably implicit in your comment - we have to beware of anecdata as well. "Two of my friends asked me for money" means very little, except that your friend group is having a rough time. The meso-scale, your "high resolution 2d data", is where to look if you want a textured picture of what's really going on while at the same time avoiding observer bias. Unfortunately, that kind of data is not always easy to get, or to interpret.

I remember reading in one of Simon Willison's recent blog posts his half-joking point that MCP got so much traction so fast because adding a remote MCP server allowed tech management at big companies whose C-suite is asking them for an "AI Strategy" to show that they were doing something. I'm sure that is a little bit true - a project framed as "make our API better and more open and well-documented" would likely never have got off the ground at many such places. But that is exactly what this is, really.

At least it's something we all reap the benefits of, even if MCP is really mostly just an api wrapper dressed up as "Advanced AI Technology."