Hacker Newsnew | past | comments | ask | show | jobs | submit | snikeris's commentslogin

Yeah, what's the deal with this error message?

The two issues I have with relying on the cloud for intelligence is:

- I don't trust them with my data

- I don't trust the data they send back to me

The first is about my privacy, the second is about not being manipulated by what is considered the politically correct answer to my questions.


Whether a medication leads to dependence and withdrawal should be given greater weight than it gets in the regulatory process. There needs to be a greater counterweight to the obvious conflict of interest in selling a medication that makes you ill when you stop taking it. Doctors should hold themselves to a higher standard than the chronic disease paradigm.


On the flip side, a lot of the reasoning and decision making might have happened in the conversation which is more ephemeral than a comment. A future LLM-age source code repository would store the conversation along with the commit, for future reference by humans or LLMs.


What properties make a language more LLM-friendly? I've found Claude will readily make use of a REPL when available.


The most important ones are fast build/startup times and informative error messages. They can focus on actually debugging your code instead of waiting for the compiler or looking up obscure error codes in the docs (if they have web access, if they don't they will hallucinate).

Lisps are great for fast iteration but the syntax is a big problem because it's so sensitive to a single misplaced paren. LLMs like clean syntax with as little noise and unnecessary tokens as possible. They work better with dynamic or inferred types, low verbosity, no repetition like `Foo foo = new Foo();`, clearly delimited blocks, you get the idea. Parts of the language should compose well so the chance that a reasonable edit will cause a syntax error is as low as possible.

Popularity was a big deal with early coding models that couldn't go from specific languages to a wider idea of programming. Modern LLMs can generalize a lot better and use Rhombus or Gleam with no major issues, but they still work better in Python or JS with more training examples.

Basically the things that make a language LLM friendly are the same ones that make it human friendly, at least if your human is an average developer instead of a FP fanatics or enterprise architect.


> because it's so sensitive to a single misplaced paren

Much more than other languages are sensitive to misplaced braces/indentation?


> Lisps are great for fast iteration but the syntax is a big problem because it's so sensitive to a single misplaced paren.

That's more a problem for humans who have never programmed in a Lisp and are using notepad.exe.

    (if p (foo) (bar))

    if (p) { foo(); } else { bar(); }

    if p:
    <indent>foo()
    else:
    <indent>bar()
> LLMs like clean syntax with as little noise and unnecessary tokens as possible.

Isn't that good news for Lisps?


A Lisp with an HM type system and good error messages might be ideal for LLMs (if you don't want dependent types).


Having a lot of online codes and ressources


I use it through reagent and re-frame. I don't have any complaints.


I was a little surprised to find when I gave an LLM REPL access to the running program, it readily started using it during development and debugging.


> I think reading this helps me imagine a version of the future I'd actually like to live in. A version where technology is used well (rather than preaching for abstinence from technology)

I believe the Amish figured this out over a century ago.


> I believe the Amish figured this out over a century ago.

The Amish rather came to a different conclusion (which I don't want to judge on, but on which I nevertheless have a different opinion than the Amish).


What is that conclusion which differs from the post you replied to? The Amish are mindful about their technology adoption.


> The Amish are mindful about their technology adoption.

The central idea concerning the Amish's relationship to technology is that only technology is allowed if it does not destroy their community.

My personal values are much less based on upholding a community, but rather are much more rooted in individual freedom and independence. This means that I (likely) come to very different conclusions regarding this class of problems than the Amish do:

For example, I am less opposed to various kinds of technology that Amish would likely consider as as "community-destroying".

On the other hand, I guess I am much more opposed to technology that can be used to surveil the user and/or makes the user dependent on the whims of big tech companies than I guess the Amish are (i.e. the Amish would likely consider this as a much smaller problem concerning which technology to allow vs disallow; as I wrote: by my understanding their central concern is which consequences some technology has for keeping their community together).

To give evidence for the previous point: (by my impression - I am not US-American) you will rather not find many Amish people at political rallys against surveillance laws. The people who attend such rallys typically also have strong opinions on which technology to use or not to use (just talk to such people who are very strongly opinionated :-) ), but - as I pointed out - these technology choices come from very different basic premises than those of the Amish.


Yeah they probably wouldn't show up to a political rally because of this:

> Separation from Evil

> The community of Christians shall have no association with those who remain in disobedience and a spirit of rebellion against God. There can be no fellowship with the wickedness of this earthly world; therefore there can be no participation in the organizations, works, church services, meetings or civil affairs of those who live in contradiction to the commands of God (this may include Catholics and Protestants as well as other religions and pagans). All evil must be put away, including using weapons of force such as the sword and armor.

https://en.wikipedia.org/wiki/Schleitheim_Confession


This is my favorite alternative form of Magic:

https://articles.starcitygames.com/articles/the-danger-room/

"I have a feeling that roughly 25% of games are decided by a player drawing too few lands, 25% of games are decided by a player drawing too many lands, 25% of games are decided by a player having a legitimate bomb not get answered immediately, and the last 25% of games are the ones that everybody hopes for where there is a ton of back-and-forth on both sides. I wanted to create a format that eliminated those unpleasant 75% of games that are unfulfilling and foster a format where ALL of the games were as interactive as possible."


My biggest beef with Magic isn't the lands system, but that the game is fundamentally stingy with cards, leading to many non-games, and/or turns where you pray for the right top-deck and don't get it.

(This is one major reason I play Netrunner instead, where your action economy can be spent on draw. You might have weak turns, but never non-turns.)

Given the huge cardpool, there's no real way to "fix" MtG that couldn't be exploited in the metagame (aside from limited, or social formats like Commander).

But that aside, I think there are two relatively minor fixes which would go a long way:

- You can spend your "land for turn" to exile the land instead, and draw a card.

- Before your starting draw, Scry 1.


Magic isn't really stingy with card draw, if I understand what you're saying. It might be a drawback in a specific strategy, but that's how the game balances itself. If weenie or burn decks had card draw (or selection) on par with control decks, they would be too good...


Love that "land for turn", although I think it might shift the balance a little too much, in that you can make a deck with too much land and high cost spells and know you can cast them reliably. There needs to be a risk factor in building up to high mana to make low mana spells matter.

Possible tweaks, maybe it has a cost (all lands have cycling 1 or 2 mana or life.) Or delay that draw until end of turn, which feels like about the right power level, but does have memory and execution issues.


My thinking is that there's a tradeoff, that you slow your ramping in exchange for cycling (replacing your land for turn). You trade your acceleration towards big spells for more consistency (or just fewer dead draws).


I don't play any IRL magic but I know it can be hard to "fix" and maybe isn't broken. Casually those rules would probably be good house rules to balance play, though even small changes like this could change balance and lean all decks towards exploiting the rule

to point, your land for draw is kinda implemented with trade routes https://gatherer.wizards.com/9ED/en-us/108/trade-routes

And that has an additional mana cost to balance it (though it does have additional bonus). So you could kinda say "everyone gets trade-routes already in play" but like any "free card" fix does it simply skew deckbuilding into a new "broken" ?


> does it simply skew deckbuilding into a new "broken"

exactly, that's why it probably isn't feasible other than casual/social play: there's no change to the fundamental rules that doesn't end up warping the meta-game. (Particularly if the fix relates to mitigating mana flood/screw with consistency: I suspect it rewards combo lists, whose win-cons are indistinguishable from "find three specific cards".)


> (This is one major reason I play Netrunner instead, where your action economy can be spent on draw. You might have weak turns, but never non-turns.)

Is Netrunner something you can get into these days, or is it hopelessly OOP? I remember hearing good things about it.


There’s a community keeping it alive with fresh content: https://nullsignal.games/

I sadly haven’t convinced my MtG playgroup (or family or other friends) to try it with me.


It's alive and well, and the community is welcoming and happy to teach: https://nullsignal.games/players/around-the-world/

Cards release more slowly than the FFG era, but the current stewards (Null Signal) have been doing a great job for the last eight years.

There's a community app to play against an AI, which has a good tutorial if you want to dip your toes in: https://chiriboga.sifnt.net.au/


The "land for turn" concept is genius. I'd say only once though. If you draw another land, that's the nuts.


Well that's your "land for turn" not "replace playing a land". You can only play one land each turn, if other cards let you play extra lands...you can play extra lands!


I understand, I just meant it wouldn't be good to let the player cycle through their cards if they kept drawing a land with this rule specifically. You still play lands as normal if desired, and you can still play extra lands if other cards let you play extra lands.


> the game is fundamentally stingy with cards

Just play blue, that's not a problem then.


I haven't tried this because I haven't played Magic in over 20 years, but I always thought it would be more fun if you had two draw decks where the second was only allowed to include basic lands.

Then in the draw phase you could choose which deck to draw from, eliminating the mana flood/drought possibility that ruins games.

Maybe some additional rules to limit how many lands you're allowed to draw in the initial draw to one or two, I dunno.


This is similar to how VS system worked, it was a short lived superhero CCG in the mid 2000s, you could play up to one card per turn facedown as a “land”, but you still drew from one deck.


There’s a few modern TCGs who take this approach, like Riftbound (the League of Legends TCG that launched last year) and Sorcery Contested Realm.

Other attempts to fix the mana problem include games like Lorcana and Flesh and Blood, in which cards have dual modes where they can be played or used for resources.


I've thought for a long time, two draw piles - one is lands only - and you draw from each of the two piles each turn.

this would be balanced for both players.

If course with magic the lands pile would have to have rules like no lands that turn into creatures or something.. maybe even just basic lands .

but each person gets a land each turn - been thinking that makes the game better.


Sorcery TCG


Alternatively you give everyone a free mulligan and if they decide to start with <3 lands in their hand or no mana ramp that's on them.

My issue with guaranteed lands is that they remove the randomness of some lands. I am not guaranteed to get my "no maximum hand size" land and my rogue passage to make my creature unblockable in a typical game of commander. I have to plan for it by using stuff like expedition map.


Drafting last Friday, my first hand was zero lands, my second hand was one land, my third hand (bottoming two) had three lands. However, I then kept drawing four-drop spells and no more lands for 8 turns.

So is it "on me?" Or is it just that the game just is high variance?


Free mulligan means on your third hand you would only bottom 1, not 2.

Besides, with a 37 lands commander deck, the chances that in 4 hands (if your limit is bottoming 2, 5 cards in hand) starts getting low. Around 1 in 10 and that's ignoring mana rocks (2 lands + 1 mana rock) which probably brings it much closer to 1 in 14+.

If you add the chances of more than 4 lands (also fairly undesirable) your odds of a bad forced hand climb to 1 in 7, which is still pretty far from the 50% chance the original post is talking about.

Without the free mulligan you would be at 1 in 4, which is why I said free mulligan is the way to go.


Only allow certainty for basic lands?


Houses got expensive because of usury (loans made for unproductive purposes). It destroys civilizations over time. That is why it was encoded into ancient religious traditions.

Housing, education, and cars, all typically financed via loans, all exorbitantly expensive.


If anything it's the opposite. Extremely low interest rates drove housing prices higher by making it easier to afford a higher price.


It’s not about the interest rate. It’s the availability of loans for unproductive purposes on a societal level. It raises the price even if you choose not to partake in said loans.

When the money being lended is digital and not backed by anything it’s even worse.


You said "usury", which is a word that describes loaning money at an exorbitant interest rate. So it's no surprise GP thought you were talking about interest rates.

(Yes, I know "usury" has had other meanings, but this is the current, common definition, and if you're going to use a word in a way that's uncommon, you should be prepared for confusion.)


They said "usury" in the context of ancient societies. I understood their meaning perfectly well. More money lent out = higher asset prices.


I know that it is now commonly defined that way, so I explicitly called out that I was referring back to an the older definition:

> usury (loans made for unproductive purposes)


Okay then! What shall be our working definition of "exorbitant"?


Above the legal maximum. If you're asking how lawyers or lawmakers define that, read https://en.wikipedia.org/wiki/Interest_rate_ceiling


So in the USA in 2026, any interest rates or fees that are not illegal, are not exorbitant? And any loan or credit that is not illegal is not usury? That is obviously not the working definition for other commenters here.

If illegal loans have financed our cars, education, and homes, then where are all the criminal prosecutions? Where are the lawsuits to recover the exorbitant rates? Was it consumers taking these illegal loans? Or was it the established banks, negotiating illegal and exorbitant rates with criminals? Is that why banks fail, or are "too big to fail", because we are reluctant to expose their criminal activity?

We heard of people "walking away" from their home loans during the 2008 recession; it's difficult to walk with broken kneecaps! We heard that their loans were "underwater", but we didn't realize they were "sleeping with the fishes"!


> That is obviously not the working definition for other commenters here.

If you've read the other comments I don't understand why you're still confused.

https://news.ycombinator.com/item?id=47882360 pointed the finger at the general practice of moneylending, known for most of history as "usury", for high housing prices. This is broadly true - lower interest rates equal high prices.

Others thought they meant modern usury, which is lending money at higher than the legal maximum interest rate.


US interest rates have been historically some of the lowest anywhere. And there's nations with very high interest rates that don't have the same housing cost problems...


You're both saying the same thing...


When I hear "usury" I usually think of high interest rates, like payday loans and credit cards, not a 2.5% 30-year term.


That's the modern definition. The comment at https://news.ycombinator.com/item?id=47880310 was about bans on usury since ancient times, when that term referred to the general practice of moneylending.

"In many historical societies including ancient Christian, Jewish, and Islamic societies, usury meant the charging of interest of any kind, and was considered wrong, or was made illegal."

https://en.wikipedia.org/wiki/Usury

Low interest rates (i.e. freer moneylending aka more usury) increases house prices.


Payday loans are not illegal. By your definition, and Wikipedia's modern definitions, a payday or car-title loan is not exorbitant, and it is not usury. The payday loan shops around here are licensed, legitimate businesses working with regulated financial instruments.

> Low interest rates (i.e. freer moneylending aka more usury) increases house prices.

You just contradicted yourself. What is usury again? Low interest rates, freer moneylending, is "more usury"? You write in the present tense that freer moneylending and low interest rates are more usury than usury? Are the rates exorbitantly low? Is there an exorbitant number of customers?

https://en.wikipedia.org/wiki/Predatory_lending


> Low interest rates, freer moneylending, is "more usury"?

In the ancient sense of the term, which this comment https://news.ycombinator.com/item?id=47880310 used, all moneylending is usury.

More loans are approved and/or borrowers are approved for bigger loans in low interest rate environments. This leads to high house prices. Hence, the "more usury" comment that I wrote. In that one sentence I had the same ancient definition of usury in mind.

I hope you don't believe low interest rates = expensive houses is something I made up.

If you think I'm contradicting myself sure go on. I find it easy to hold two definitions of a word in my head and use the correct one in the right context. But I understand that not everyone can and some may find it confusing.


The question is, whether you can make words mean so many different things.


Yeah it was redefined to that at some point. See Belloc’s essay on usury if you’re interested.


The definition of usury hasn't changed in 200 years.

From Webster's 1828 Edition:

U'SURY, noun s as z. [Latin usura, from utor, to use.]

1. Formerly, interest; or a premium paid or stipulated to be paid for the use of money.

[Usury formerly denoted any legal interest, but in this sense, the word is no longer in use.]

2. In present usage, illegal interest; a premium or compensation paid or stipulated to be paid for the use of money borrowed or retained, beyond the rate of interest established by law.

3. The practice of taking interest. (obsolete)


The person I responded to said:

"That is why [a ban on usury] was encoded into ancient religious traditions"

So they obviously meant the old definition, which encompasses all moneylending.


I'm not even sure where "houses got expensive" come from. Houses certainly did not increase in price (per unit of area) in last ~80 years, inflation adjusted - they tightly fluctuate around the same point. Housing affordability is in fact 4th best in the US among all countries in the world, and it got better in the last decades (although with fluctuations, and periods when it was getting cheaper were not pleasant as it meant millions of people going under).


> I'm not even sure where "houses got expensive" come from

Not been shopping for a house recently?

Avg house price in my city doubled in the past 5 years.


Interesting.

This posture contradicts quite a few testimonials we can read about everywhere, including here in HN.

Maybe it's the salaries, then, that are half what they should be?


Because this is vibes vs data. People are made to believe that houses are expensive and people are being squeezed, because it makes money to throw people into rage. Housing price per square feet is flat over decades...


It's very true for LA, SF, NYC, DC, and other similar cities. It's far less true everywhere else. Here in Atlanta you can find homes <250k and condos <150k


Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: