wwweston

Speaking as someone who has run courses designed to take late teens / early 20somethings with a variety of backgrounds and tried to teach them 1st semester programming concepts in two weeks, I think anyone who thinks tcp/ip an easy four week dunk for most doesn’t remember how much they’ve already learned and take for granted.

Sure, if you’ve already learned enough groundwork, tcp/ip is accessible in weeks. The same is true of most of the algebraic concepts in play here. And both have rabbit holes you can also spend a much longer time going down (though here I am willing to give the edge to math which offers much greater opportunities for hypergeneralization and new vistas of abstraction along which not only specific rabbit holes but entire new generalizations of both rabbits and holes may be found).

If you’re not redistributing ebooks you bought, it’s hard to imagine what legal claim someone could make against you, and even harder to establish the ethics of that claim.

To say nothing of Amazon’s ethics.

Yeah, we’re early culturewise in the digitization experiment and high on optimism about the benefits, but not very far into reckoning with the downsides and incentives skew a bit towards carelessness.

The real danger is that we’ll be so careless we’ll discard other enduring ways of doing things before we smarten up to their particular benefits.

My hope is that disciplined people still have an intuition for this, even among the digitally steeped. Sysadmin/ops types tend to a culture of diversifying backup location and even media type. Maybe that can reach back to human legible hard copy.

I’m having trouble imagining the privacy concern. If someone’s vehicle is topping 90db then it’s hard to argue what they really want is privacy, and even the most invasive monitoring method I can think of (widespread mics) could be nerfed by requiring hardware that only responds at db threshholds well above conversation.

I’m at the point where I don’t just want ticketing and other enforcement on this front, I want yanked licenses for this after a warning, and I might even want it legal to shoot vehicles guilty of noise violations and their drivers with paintballs to mark them for further enforcement and shame.

Smoking is the atmospheric equivalent of peeing in the pool; noise pollution is something between that and opting everyone into your dumb M80 party. It’s antisocial, it causes health problems on top of discomfort, and it should stop.

I might imagine that even this cycle could change, if the resources necessary to support step 1 and step 2 are much lower. Which might mean step 3 isn’t as necessity driven.

Of course, it’s possible other things can drive step 3.

And frontier models are already a study in unusual levels of resources dedicated to step 1.

It has now been longer since apps on smartphones were introduced than it was between the web’s birth and smartphones (to say nothing of popular uptake, it’d be generous to give the web a full decade there before the rise of the smartphone). Combine that with recency bias and the marketing serving incentives of platforms and vendors to push towards “native” and the point seems pretty strong.

There’s some minority of apps where work is offline resource bound or mostly offline regardless, those certainly have always made sense on device, other than that habit and mindset probably is the main driver of preference.

See also Bill Thurston’s classic Math Overflow answer to a student wondering where they fit compared to a Gauss or Euler:

https://mathoverflow.net/questions/43690/whats-a-mathematici...

“The product of mathematics is clarity and understanding. Not theorems, by themselves. [Their importance is not just in their specific statements], but their role in challenging our understanding, presenting challenges that led to mathematical developments that increased our understanding.

The world does not suffer from an oversupply of clarity and understanding (to put it mildly)… In short, mathematics only exists in a living community of mathematicians that spreads understanding and breaths life into ideas both old and new. The real satisfaction from mathematics is in learning from others and sharing with others. All of us have clear understanding of a few things and murky concepts of many more. There is no way to run out of ideas in need of clarification. The question of who is the first person to ever set foot on some square meter of land is really secondary. Revolutionary change does matter, but revolutions are few, and they are not self-sustaining --- they depend very heavily on the community of mathematicians.”

I’ve got an old DX. The fact that it doesn’t have net access anymore is now a security feature and an anti distraction mechanism. I can load what I want by USB. It doesn’t do epub3 which is a problem, but it does PDFs and conversion tricks are possible.

The Amazon ecosystem isn’t uniquely untrustworthy, but it’s not where I want to keep future electronic purchases.

That makes sense — decisions that are about personal benefits beyond technical/product choices and their merits.

When you say “political“, what do you mean? It’s easy for me to relate to talk of software development aesthetics or philosophy and I’ve certainly had plenty of conversations that I think are accurately described that way. But I can’t think of software development discussions that I’d describe as political (except in the sense that politics is what happens when people disagree on anything).

It does seem to be true that voters are often content to select and re-select congressional representation that will hide from too much critical attention, give evasive responses, or otherwise dodge direct accountability.

People's stated preferences are as you say. Most agree we'd like responsive representation that's conscientiously engaged in opinion gauging and fact-finding to determine what the people want and need. We can even select for those preferences, and be consistently unkind to those who are evasive. Yet we're not always consistent about that, and different preferences are often revealed at the ballot box.

This is the inescapable fact of the electoral system: officeholders are selected by voters in elections. There is no other mechanism for selecting them. That's a key difference between your other "is like" examples and elected representation. Pollution and health care are interesting problems to consider in their own right, they even have collective action components, but they also both actually have other identifiable agents and dynamics to whom more of the responsibility can be assigned. When it comes to elected representation, not only are there no agents with more responsibility, there are no other agents with power to select representation. Who else can bear ultimate responsibility for who is selected and how they govern?

You can identify things that make it difficult for people to meet that responsibility productively, and that's reasonable enough, all my comments here are essentially identifying the psychological barriers that assign agency elsewhere, and I'm happy to talk about others.

But there is no other place responsibility lies. The hands on the levers are ours collectively. No one else is coming to save us but us. No system fix will let us run on rote habit and get better results (and also, to fix the system, you have to work to win big under the system as it is).

The insane thing is that nobody listens when they do talk about it.

A huge thing in the Harris campaign was that her policies polled much better than Trump's (including housing accessibility by increasing construction, regulations for a fairer rental market, downpayment support for firstime buyers). Nobody knew the policies were hers, though, just like awareness of the accomplishments of the Biden admin seem to be shockingly rare.

You can put part of that on the Democratic party, which could clearly up their media game.

But especially in the algorithm age, a lot of what you see is based off your preferences. This makes people who think their main leverage is their resentments really easy pickings. Someone posts content about how the housing market is unfair and politicians aren't doing enough and everyone who agrees teaches their algorithm that's what they find engaging and important -- not that housing policy and who is doing what is important.

So who hears about this?

https://newdemocratcoalition.house.gov/media-center/press-re...

The people who tell their algorithms that's what they want or are more deliberate about their communication channels and what they go looking for.

It's reasonable to acknowledge that some individuals are so overwhelmed by the demands of their circumstances that they truly have no disposable time & attention to direct to the society they live in.

The statistics I've encountered suggest that this is smaller than 10%. Between there and the percentage of people who even vote is one margin of responsibility, and there are many people who have time for more but put it somewhere else.

That's a choice, and people are free to make it. It also has consequences, like the elections that select officeholders.

Calling this "victim blaming" seems like rhetorical lane-shifting. We're not discussing crimes where someone suffers a violent attack of some kind they had no agency in, we're discussing a society where those who govern are selected by voters in elections, which at some level means issues are representative of those who vote (and those who don't).

We're all very used to locating agency in large systems somewhere we don't have control over, and at one level that's even pretty understandable as a first approximation -- most people don't have substantial influence at an individual level. And no individual contribution to your retirement account will leave you comfortable, it's the collected effort over time that might. That's true for individual influence too, and magnified by organized cooperation.

You brought up OWS. That's a reasonable example to consider in a lot of ways -- it was collective cooperation, certainly better than apathy, and I respect it for that (and did not take kindly to the ways it was disparaged especially in some partisan media).

It was also a movement that was probably overfocused on the the idea that social leverage is primarily about expressed demand, especially given that the goals of participants were diverse and unfocused. At times it almost seemed like the tactic (occupation) was the goal. It was successful in injecting some longstanding ideas into some parts of public consciousness (there's a 99% that's poorly served by things as they are now and 1% that are dramatically overserved) and may have created some networks that have had a longer influence. But I think it also showed that many participants don't know what effective ongoing influential cooperation looks like.

We want achievements like the civil rights movement but we have Woodstock ideas about what that looks like instead of movement-momentum habits.

I've had this conversation enough to know that I'm also in a sense making a similar mistake, expressing ideas and expecting that alone to change things, which is especially naive given that I know the mindset shift is a bit of a hard sell. At least I know that someone who argues back and forth over a series of comments probably cares more than average, maybe enough to create ongoing habits of influence themselves.

When the electorate is literally the mechanism by which officeholders are selected, how is the electorate not responsible?

This is not the same situation as someone who is the victim of a violent crime they didn’t volunteer for, choosing language that creates that confusion won’t change the reality that officeholders are chosen by the electorate.

You’re correct that time is the fundamental challenge. It’s also really what the money is always a proxy for at some level.

Meetings in DC are probably not the right focus. Every congressional officeholder has offices in the region they represent, most have multiple. Most people don’t use them for the same reason they outsource their understanding of current events to Fox News or Rogan & guests. Some people do contact offices by phone or message, but fewer band together with others who care about a policy topic and leverage collective influence.

Sure it’s hard and time consuming. I’m not speaking from a position of full ability or particular privilege (though I have enough time to post on HN). But it’s also a bit like the old saw about meditation — 10 minutes a day, and if you’re too busy, 20 minutes. The activities themselves don’t always produce immediate leverage but once they lock in the return is powerful.

Elections still have consequences, right? I’d think the fallout between fall of 2024 and now illustrates that. And that was the choice of the electorate (unless someone can demonstrate results were tampered with). So voter choices count.

There are differences in how individual congress members and coalitions handle policy, so who voters choose matters.

Also, some of that research you’re invoking shows that most officeholders try to keep their promises:

https://theconversation.com/do-politicians-break-their-promi...

I agree lobbyists have influence. What is a lobbyist and why aren’t more people lobbying?

Donors also have influence, and yet the electorate has every opportunity to determine who donors must influence every election. Why would they choose someone who is only beholden to donors?

Money distorts the system in some ways and I agree that’s a problem that could use systemic mitigation (farther back than Citizen’s United probably, Buckley vs Valeo is arguably the deeper roots).

But ultimately, money doesn’t remove the fundamental electoral mechanisms (yet) or opportunity for volunteer direct lobbying. It primarily distorts to the degree that it can be used to buy the focus of the electorate and to the degree it can be used to buy other people’s lobbying time.

People could spend their time managing their own political /public policy focus and volunteer lobbying instead of any other leisure activities. I’ve done it and I know others who do. Most Americans don’t, and that’s a revealed preference. Other leisure activities are more important.

The USPS is a no-brainer public service and the only reason there is any question of its value is due to the severely broken, dysfunctional, corrupt

electorate.

Congress is composed of people who the electorate sends there.

Once there member choices are shaped by the people who contact and persuade them.

If the USPS is poorly funded or managed, it’s because US electorate either wanted that, or was inattentive about the relevant funding and management and cares more about other things.

And if the postal service dies or is captured and privatized, that’s a reflection of the preferences of voters, or a testament to the limits of their attention and intelligence to the point where they voted for people who did things they don’t want.

Most Americans also prefer to blame political folk devils to for the failures instead, and seem to be more happy with that than personal and community discipline that would be necessary to engage responsibly, though, so the system is arguably working to reflect people’s revealed preferences already.

EDIT: I should probably add that it’s not obvious to me that it’s poorly managed. I’ve enjoyed decades of adequate-to-impressive service via USPS over a variety of locations.

FYI - if there's any reason to believe your particular hypermobility issues might be rooted in any kind of named or unnamed connective tissue disorder, be sure to familiarize yourself with the potential vascular hazards of heavy lifts in general and Valsalva breath-holding specifically. Hypermobility can be a sign that you've drawn a genetic lot which means that you don't even have to lift all that heavy to create aneurysms.

And in keeping with our theme of the limits of both LLMs and humans, unfortunately many exercise and medical professionals may not focus on how some specific genetic lots work until a catastrophic problem presents itself.

ignoring feedback that a movement was causing pain, prescribing stretches to increase flexibility in a hypermobile person

So glad you found better advice than this!

LLMs are fantastic tools for exploring a topic, coming up with good questions and lines of understanding to pursue with professionals, and in picking professionals. I think they're also poor outright replacements for people especially when it comes to deep and important domains, but still quite useful.

I’ve totally lost the thread here but this is interesting.

Who are the religions without ghosts?

And is the overall point “just because we’ve made big leaps of progress doesn’t mean every challenge is tractable let alone a moonshot sprint away especially those we have solid theoretical limits on”? That’s a point I certainly find amenable to I just want to make sure I’m not missing something more subtle or sophisticated.

We have systems around humans that exist to manage expertise gaps, credibility signals, and accountability. This is part of what makes humans as good as they are, along with specialized training and some measure of meritocratic selection. We license and regulate and account and litigate to make a system that responds and improves.

Some of this might be applicable to LLMs, but some isn’t and much of it would be resisted. This is one reason we’re not likely to get “as good as a human” because at some level we’re not optimizing for the outcomes; we’re optimizing for speed, convenience, some participant’s economics, and underlying beliefs.

Oh I’m not saying that pervasive mandated identity verification is good. I think people should be skeptical about its benefits, very wary about its downsides, presume bad actors are involved, and make proponents fight for every inch of buy in, no matter what their stated motivations are.

I also think that holding important ground involves understanding the terms of the argument, including the problems people say they’re trying to solve.

Crusades against sexually explicit material are certainly popular in some places.

But these days I see a lot more talk about the developmental effects of parasocial media on kids. There’s a whole segment of buy-in there that didn’t exist before.

On the other hand, the dumb models are more than adequate for simple noncritical tasks, like directing a user to the appropriate FAQ entry

This is a lane I’ve been experimenting in —- seeing what I can get out of models that work in 16GB VRAM for simple tasks (screen scraping, decision tree navigation, natural language queries). It’s interesting for sure (certainly reveals non-deterministic limits) and promising for low criticality review-opportunity tasks, but I also feel like I need better sources/community for understanding and reflection. Preferably those that aren’t hype channels. Any pointers?

Tech product price dynamics benefit from a bunch of things that food doesn’t: they’re optional purchases, they’re early stage developments which have more low hanging fruit, and purchase price can be subsidized with later plays (subscriptions, data sales, network effects, freemium to enterprise pipeline).

Also - I think if you look at the data you’ll find periods off the gold standard where food prices grew more slowly than inflation and even wages, ie food becomes cheaper. 80s and 90s for example.

And increasingly, everyone isn’t using them, even if they’re on them.

I’m on Insta and WhatsApp and I use them a few times per year. I’m on Messenger and have seen a dramatic dropoff in messages. I’m on FB frequently and notice only a small fraction of my friends bother anymore and it’s become an interest platform to make up the lack, so I’m trending toward less time there. I’m on Twitter/X but check in maybe once a month.

I may not be a typical user, but I’m probably not unique either.

I wonder if the struggle is really comprehending thoughtful selective adoption.

Schools are the place where the product is a more fully developed person. There's no LLM shortcut for generating that. There are many ways you could use LLMs that would discourage it. There may be some that can encourage it.

Personally, I can see aggressively keeping kids away from LLMs until they've learned effort, living in tension/frustration, the pleasure of breaking through to discovery, trust evaluation, hypothesis/test cycles, and good socratic dialogue from the learner's side.

It may be possible at intermediate phases to prime some models to help with this process.

Your language more accurately reflects how the US government is supposed to operate, and that's important, we'd all be better off if people remembered that the government has a number of branches with different responsibilities, actually kept track of who controls each, and otherwise understood the system.

It's also more or less true that the Biden administration took more responsibility and initiative in antitrust enforcement:

https://www.americanbar.org/news/abanews/aba-news-archives/2...

https://www.wbur.org/onpoint/2024/10/02/biden-ftc-antitrust-...

https://rooseveltinstitute.org/blog/the-biden-administration...

1) Different people might optimize for different things. There are people calculating that expensive hardware plus cheap rentals means owning isn’t optimizing, but there are people making choices that fit your preferences too.

2) I think it’s important to recognize that one of the things models are good for is astroturfing, and any given conversation you see may be direct or secondary effects of that (among other marketing).