I think it would be helpful if you explained how distillation works (as relevant to the conversations threads here) because I agree that this is a key point.
This seems like an important post. It looks like these letters are occasionally used to as a tactic, and i can see how such a tactic can really scare employees in a country where legal bills can climb really fast.
Your statement is technically correct (and mirrors my initial feelings) but the comment below that notes that we accept many other abstractions seems more “meaningful” in the larger scheme of things.
For instance, a software engineer who also understands how to design microprocessors would indeed in my own evaluation be a “better engineer” than me (someone who does not). Yet, I wonder if they would be meaningfully more productive than a good software engineer who “just” understands how microprocessors work..
The current administration's actions are for sure signaling a distinct lack of predictability or stability.
The U.S. is either harming itself erratically, or systematically enforcing control on US based businesses in a way that is historically more synonymous with, say, China.
A lot of us worry it is the former. I wonder if it's better or worse if it is the latter. The US, having seen that Europe and the world seem to tolerate interference by CPC based on opinions expressed by Alibaba or Chinese car companies, may have decided that it's fair game.
It seems to be observably true that expressing opinions critical of the administration, even about one's own business, leads to further harm from government right now in the U.S. in a country which is supposed to have strong freedom of speech rights.
It's obviously unfortunate that this is the case.
It's also a shame that this is happening when the folks regulating AI have a tech industry background.
The administration seems to be clearly signaling "stop resisting". I am not sure how you read me as saying that.
Unfortunately saying it in the press is going to make it even harder for them in the current environment. I know this administration gets non-trivial support from the valley, but what to most outsiders seems like "targeting businesses based on personal vibes" is going to do long-term harm to the U.S.
I hope those who represent technology in government, especially the AI head David Sacks are giving this due consideration.
Buying state owned enterprises and running them more efficiently to double their valuation (or more) has been a reliably profitable business in many countries. This suggests to me that government bureaucracy, lack of accountability etc. could result in “at cost” trending to $100 or more if government buys private companies, or starting close to $100 if government starts a new one.
My favorite part of the paper is that the “attack” isn’t just exploiting a bug — it’s exploiting how different components interpret the same input. Modifying an executable as it’s loaded into memory is one example, but the deeper pattern is the mismatch.
What’s interesting about the malware in this post is that it goes one step further: instead of exploiting mismatches, it corrupts the computation itself — so every infected system agrees on the same wrong answer!
More broadly: any interpretive mismatch between components creates a failure surface. Sometimes it shows up as a bug, sometimes as an exploit primitive, sometimes as a testing blind spot. You see it everywhere — this paper, IDS vs OS, proxies vs backends, test vs prod, and now LLMs vs “guardrails.”
Fun HN moment for me: as I was about to post this, I noticed a reply from @tptacek himself. His 1998 paper with Newsham (IDS vs OS mismatches) was my first exposure to this idea — and in hindsight it nudged me toward infosec, the Atlanta scene, spam filtering (PG's bayesian stuff) and eventually YC.
The paper starts with this Einstein quote "Not everything that is counted counts and not everything that counts can be counted", which seems quite apt for the malware analyzed here :)
I have just recently discovered Watts, but I think he may have agreed.
“Watts left formal Zen training in New York because the method of the teacher did not suit him. He was not ordained as a Zen monk, but he felt a need to find a vocational outlet for his philosophical inclinations. He entered Seabury-Western Theological Seminary, an Episcopal (Anglican) school in Evanston, Illinois, where he studied Christian scriptures, theology, and church history. He attempted to work out a blend of contemporary Christian worship, mystical Christianity, and Asian philosophy. Watts was awarded a master's degree in theology in response to his thesis, which he published as a popular edition under the title Behold the Spirit: A Study in the Necessity of Mystical Religion.
He later published Myth & Ritual in Christianity (1953), an eisegesis of traditional Roman Catholic doctrine and ritual in Buddhist terms. However, the pattern was set, in that Watts did not hide his dislike for religious outlooks that he decided were dour, guilt-ridden, or militantly proselytizing—no matter if they were found within Judaism, Christianity, Islam, Hinduism, or Buddhism.”
Causality, in all my experience of human relationships in real life, in books, even in movies, is rarely single factor, and often goes in both directions :)
A marriage certificate is not a 100% vaccine against what imho are the shared root causes of divorce and unhealthy home environments - like mutually incompatible or self-centered human beings (absence of love as a noun), lack of commitment (absence of love as a verb).
With or without marriage - partners with a shared world view do well if (a) both partners want happiness for the other as much as they want it for themselves and (b) both openly expect to be a good partners to each other for life, even as both partners inevitably change, grow, fall short, succeed, fail etc.
We don’t have to call (a) “love” and (b) “marriage”, but these remain the most common shared names for these concepts in many societies.*
IMHO though, since we are very much imperfect animals and social animals, society having a shared expectation that couples strive for (a) and (b) matters - and I would be willing to consider all the ways in which this can be done.
* - We also face the separate and important problem that we have harmful definitions of these words in some sub-sections of human society -
I would agree that it is the commitment that matters.
Where we may agree:
Social norms really do impact human behavior. Marriage is a social norm supporting long term commitment. In communities where it has been replaced with another social norm supporting commitment (eg my well-off friends in Europe), it has become less relevant.
I also posit that adults in committed coparenting relationships constitute a small minority of unmarried adults in America (vs. France for example where a majority of my friends with kids match your description).
Where we probably disagree:
In my observations of close friends in loving relationships with children, previously in loving marriages, are now divorced and in respectful and functional coparenting but not cohabitating relationships.
For a considerable amount of time, they are functionally single parents. In most cases parents and siblings of one ex-spouse are unlikely to want to support the other ex-spouse with in-person child support.
The bright exception to this rule seems to be divorced co-parents who live in close proximity or in one instance in the same duplex and are good friends.
Like Matt, I am a supporter of the state providing a solid safety net on basic needs (food, health care, self-improvement, safety) to ALL its and pay for it by taxing the well-off citizens more than the average although unlike him, I would not describe myself as a socialist.
Marriage is many things, but amongst them, it is also a safety net for the children of that marriage. At its best, it brings the resources of two extended families and friend networks together to support the couple and the children. I wonder if Matt would agree with the view that Marriage is the most “atomic” form of socialism (which he seems to support)
I understand his critique of the stats, but I wonder if Mr. Bruenig disagrees with the author’s points of view.
In particular I wonder what he thinks would be best for his own kids. He seems to be, per wikipedia, married to a high-school friend (sweetheart?) for almost a decade with two kids.
While I do not think his personal actions have any bearing on the accuracy of his statistical critiques of the author, it does seem like his revealed preferences support the author’s points of view. I would be shocked if he believes the health of his marriage has no bearing on the overall wellbeing of his kids.
Regardless, I wish him and his wife a long and happy marriage, because I believe that would be the best outcome for both of them and their kids.
Working hard to ensure you don’t repeat the mistakes of your parents already makes you a good parent - one who cares and strives.
I worry too, especially about the weaknesses I share with one of my parents. Occasionally though, I find that my inward focus on my concern about a mistake I might make causes me to focus less on my kid or his concerns in an important moment. Since I learned this, I try to be present enough to react to life while it’s happening, instead of ricocheting painful memories while life is happening!
The good news is that loving parents who strive to continuously do better serve as a good example for their kids, even when they fail. Life is hard and full of failures. Even when you fail, and even if you fail frequently, showing your kid that you’ll keep striving to do better - now that may be the most valuable lesson of them all.