MoltenMan

I have been to Tokyo, and I agree with you; perhaps I should have been clearer, but my point was that in American cities you DO need cars (this is a given, however unfortunate it is), and so you are going to already have a car and the costs that go along with it, and as such you are not going to use public transit even when you could because you already have a car. If I had had a car when I was in Tokyo I certainly would have used it! Public transit takes half an hour to go a couple miles, and cost wise gas is cheaper! In the scenario where you already have a car, it doesn't make sense to use public transit 90% of the time. Does that make sense?

The only point I was trying to make here was that cost isn't the reason public transit doesn't work in American cities, and making public transit free will in fact make it worse rather than better. I want good public transit too, don't worry!

Why do you need free public transit for this? Why not just cheap public transit? Cars certainly aren't free, but people drive them because cars are just far better in most ways; you get to listen to your own music, they can be fun to drive, are incredibly flexible (can go anywhere), don't have the last mile problem, take like half the time unless it's rush hour, etc. Cost just isn't the reason people aren't using public transit. People buy cars because you NEED cars for many other reasons (most of them mentioned), and then once you have a car and pay for insurance and everything it's simply nicer and faster than public transit (rarely cheaper, even only factoring in gas and per mile depreciation, although the BART is so expensive it comes close). (FTR: I'm not sure where I stand on public transit subsidization; I think some is definitely good, but not sure how much I think is optimal).

Safe is a tougher problem, and especially when you consider that in most American cities with subways, the exact cities we're talking about right now, generally NO, there are many areas where it is not safe to walk! While I agree public transportation safety and public safety in general are incredibly important, they're also (clearly, given how we're failing at them) also difficult endeavors.

I also think this is relatively orthogonal to cost; cost is nice because it helps keep the people you don't want in the train (crazy people with knives, people smoking weed in the train, people who yell at everyone on the train and make them feel unsafe, etc.), but it's also important to create at least a small barrier so that people don't 'waste' the transit system! In general, when things are completely free, they will be taken advantage of; even a very small tax / friction helps stop this. If the subway is completely free, there's no reason to not just sit on it all day (taking up space and making it worse for everyone). I think subsidizing the subway is net beneficial given that we subsidize cars and things already with road upkeep and such, but free is not what you want.

Yep. I'm already not a huge fan of reading AI writing, and adding an obviously false disclaimer at the top is ridiculous.

Calling LLMs 'next token predictors' is completely reductive and disingenuous; it's true that technically that is what they're doing, but so are you! What people generally mean by this though is that they're just 'predicting the next token of their training [i.e. the internet]'. If you were talking about the raw models, this would actually be true; but the models are post trained, so even this description isn't true at all anymore! Saying they aren't 'intelligent' is both not useful and (imo) wrong. Who cares if it matches your definition of 'intelligent'; it still gets impressive stuff done, much more impressive stuff than you seem to be implying.

And is Claude the one that got frustrated arguing with people? If not, why is Claude the one writing this?

The problem is that this is an incredibly niche / small issue (i.e. <<1% of users, let alone prompts, need this clarification), and if you add a section for every single small thing like this, you end up with a massively bloated prompt. Notice that every single user of Claude is paying for this paragraph now! This single paragraph is going to legitimately cost anthropic at least 4, maybe 5 digits.

At some point you just have to accept that llm's, like people, make mistakes, and that's ok!

The subtitle specifically says 'the lower rungs of middle class have shrunk'? This seems a little ridiculous to just say. As trendy as it is to hate on modern day life and talk about how awful it is, in material terms it's pretty clearly better than it's ever been.

But being 'conscious' of something is being aware of it; your 'subconscious' is the part of your brain 'below' your awareness (although it is true that it's also below your consciousness! So perhaps both would work)

...no? It's the same as when you say you'd 'die for somebody'. I don't want to die, but if I had to die to save my family I would. That's not being suicidal. Similarly, if space is important enough to you to take this risk (which realistically is a pretty low risk!) I wouldn't call that suicidal either. I take the risk of death driving in my car every day; that's the nature of life.

While I don't necessarily disagree with this, I also wonder how much his 'data' about Nobel Prize winners and Institute of Princeton grads actually holds up vs how much of it is just very expected regression to the mean. He talks about Shannon; at some point Shannon was always going to have his last great idea. Given that the idea that made him famous was his greatest, you wouldn't expect many other ideas like that just from normal variation.

Essentially, if you take scientific ideas, including Nobel Prize ideas, and put them all on a bell curve of how difficult it is to find them, you wouldn't expect the same person to have multiple ideas all the way on the right, even if they are very above average.

I think this makes sense as part of the existing 'skeptic cost'

Ok, but if you are investing capital in some sort of production line or industrialization you are not going to want to do that in an area where you might just lose your entire investment instantly; instead, you're just going to invest it in Texas or China. Of course with more extreme examples like yours you do have to put some cost on the existing companies to get it fixed, but it would be something with a smaller cost like having to dispose of the mercury properly (whereas in this article's examples they just flat out ban these things, which you can't do to existing factories).

As somebody who is entirely for restrictions on internet / social media, I think you're missing the bigger picture here. First, you assume that parents have the technical knowhow to restrict their kids from specific sites. My parents used a lot of different tools when I was a kid, but between figuring out passwords, putting my fingerprint onto my mom's phone, and spoofing mac addresses, I always found a way around the restrictions so I could stay up later.

But let's assume the majority of parents can actually do this. The problem with social media is not an individual one! We've fallen into a Nash Equilibrium, a game theory trap where we all defect and use our phones. If you don't have a phone or social media nowadays you will have much more trouble socializing than those who do, even though everyone would be better off if nobody used phones. As a teenager, you don't want to be the only one without a phone or social media. And so I truly do think the only solution is with higher level coordination.

Now, it's possible that the government isn't the right organization to enforce this coordination. Unfortunately, we don't really have any other forms of community that work for this. People already get mad at HOA's for making them trim their lawn; imagine an HOA for blocking social media! I do think the idea of a community doing this would be great though, assuming (obviously) that it was easy to move on and out of, as well as local. This would also help adults!

So to be honest, I don't think parents have the individual power to fix this, even with their kids.

Most of this is very true, except for the one caveat I'll point out that a space complexity O(P(n)) for some function P implies at least a O(cubedroot(P(n))) time complexity, but many algorithms don't have high space complexity. If you have a constant space complexity this doesn't factor in to time complexity at all. Some examples would be exponentiation by squaring, miller-rabin primality testing, pollard-rho factorization, etc.

Of course if you include the log(n) bits required just to store n, then sure you can factor in the log of the cubed root of n in the time complexity, but that's just log(n) / 3, so the cubed root doesn't matter here either.

This seems like the most likely reason to me!

I don't think this post reads as AI at all. It has none of the tell-tale signs either (em dashes, common constructions like 'not just ____ but ____, bullet points, headers, etc.)

Right, so the solution is to silently upload their encryption keys to Microsoft's servers without telling them? If users don't understand encryption, they certainly don't understand they've just handed their keys to a third party subject to government data requests.

What exactly are you hoping Windows does here? Anyone who knows anything about Bitlocker knows Microsoft has the keys (that's where you get the key when you need it, which I have needed it many times because I dual boot!) Microsoft could put a big screen on install saying 'we have your encryption keys!' — would this change literally anything? They would need to also explain what that means and what bitlocker is. And then after all of that, the only people who are going to decide 'actually I want to set up FDE myself' are going to be the technical people who already knew all of this already! This is just a non-issue.

This is such transparent fear-mongering. How often does this actually happen versus how often are cloud providers breached or served with legal requests? You're solving a hypothetical edge case by creating an actual security vulnerability.

This is not fear mongering at all! The nice thing about Bitlocker is that you don't need to put in your key 99% of the time (and in fact 99% of Windows users — who are not technical! — don't even know they have Bitlocker). But occasionally you do need to put it in. Once or twice I've booted to the bitlocker screen and I actually don't even know why. Maybe my TPM got wiped somehow? Maybe my computer shut down in a really weird way? But it happens enough that it's clearly necessary! That big Crowdstrike screwup a year ago; one of the ways to fix it required having your Bitlocker key!

Encryption by default and cloud key escrow are separate decisions. You can have one without the other. The fact that Microsoft chose both doesn't make the second one necessary, it makes it convenient for Microsoft.

Again, this is not true for a product like Windows where 99% of users are not technical. Remember, Bitlocker does not require your key on startup the vast majority the time! However, there is a chance that you will need the key at some point or you will be locked out of you data permanently. Where should Microsoft give the user the key? Should they say on install 'hey, write this down and don't lose it!' Any solution relying on the user is obviously a recipe for disaster. But again, let me remind you that encryption by default is important because you don't want any old random laptop thief to get access to your chrome account! So yes, I think Microsoft made the best and only choice here.

And because of Bitlocker, their encryption is worth nothing in the end.

No, it's worth exactly what it's meant for: in case your laptop gets stolen!

flawed default

Look, in terms of flaws I would argue 'the government can for legal reasons request the key to decrypt my laptop' is pretty low down there. Again, we're dealing with the general populace here; if it's a choice between them getting locked out of their computer completely vs the government being able to decrypt their laptop this is clearly the better option. Those who actually care about privacy will setup FDE themselves, and everyone else gets safety in case their laptop gets stolen.

Just to be clear: bitlocker is NOT encrypting with your login password! I could be a little fuzzy on the details but I believe how it works is that your TPM (Trusted Platform Module) is able to decrypt your laptop, but will only do so if there is a fully signed and trusted boot chain, so if somebody gains access to your laptop and attempts to boot into anything other than Windows, it will ask for the bitlocker key because the TPM won't play ball.

The important bit here is that ~*nobody* who is using Windows cares about encryption or even knows what it is! This is all on by default, which is a good thing, but also means that yes, of course Microsoft has to store the keys, because otherwise a regular user will happen to mess around with their bios one day and accidentally lock themselves permanently out of their computer.

If you want regular FDE without giving Microsoft the key you can go ahead and do it fairly easily! But realistically if the people in these cases were using Linux or something instead the police wouldn't have needed an encryption key because they would never have encrypted their laptop in the first place.

I think most people don't understand that 99% of people don't know what data encryption is and definitely don't care about it. If it weren't for Bitlocker, their laptops wouldn't be encrypted at all! And of course if your software (Windows) encrypts by default but you don't want to bother the average user with the details (because they don't know anything about this or care about it) you will need to store the key in case they need it.

To everyone saying 'time to use Linux!'; recognize that if these people were using Linux, their laptops wouldn't be encrypted at all!

I agree with a lot of what you've said, but I completely disagree that LLM's are no longer sycophantic. GPT-5 is definitely still very sycophantic, 'You're absolutely right!' still happens, etc. It's true it happens far less in a pure coding context (Claude Code / Codex) but I suspect only because of the system prompts, and those tools are by far in the minority of LLM usage.

I think it's enlightening to open up ChatGPT on the web with no custom instructions and just send a regular request and see the way it responds.

Whenever this comes up people point out, 'Come on, let parents decide for their kids!' -- I sympathize with this argument, but let me explain why I don't believe that actually fixes the real problem. For reference, I'm gen-Z, COVID hit while I was in highschool, and I have seen and to this day see Tiktok / Reels / Shorts used every day by my friends (and to some extent me).

I may not be having kids for a while yet, but if I had teenagers today I would absolutely move somewhere where it is not legal for kids to have social media accounts. The underlying problem is that this isn't an individual problem, it's a social one! If a teenager's friends all have social media, he is going to be left out! It is going to severely hurt his life. Even if he never watches short-form video (the main component of social media I think is detrimental), his friends will! When I was in highschool sometimes my friends and I would get together and we would be bored, have no clue what to do. Instead of messing around doing random things, a couple of them would just open up Instagram reels and bam, afternoon wasted. If the half the group isn't trying to do something, you aren't going to do anything. Contrast this with before I was a teenager and before phones, I vividly remember me and my friends just exploring and doing random things. It's just a different experience and I think social media needs to be banned for everyone for it to be effective.

I haven't looked at the source but the first paragraph of the article shows the percentage of people thinking a college education is important or worth it declining in the last decade. This is what I was referring to, but to be fair I don't know if college has actually gotten worse or if it's solely a change in what people think. For me personally it definitely is not useful at all for learning.

I think 'mirage' is the wrong word; college very much has gone out of vogue as a place of learning. But if you want to get a good job, you still do need to go to get the piece of paper, regardless of whether you think you'll be learning anything. The other factor you have to keep in mind is that college is where every 'smart' 20ish yo goes. If you want to meet people your age similar to you they are most likely all in college! It's difficult as a 20ish yo to have a social life outside of college.

I personally took a gap year working at a startup mid college, and as much as I enjoyed working I still ended up coming back to college purely for those two reasons.

I can't say I understand the Palantir hate. Isn't it just a database analytics SaaS? Why not hate Google as well because government employees who do things you don't like use Google? Is the Palantir hate just manufactured pointless rage or is there an actual reason behind it?

Clicking anything on the banner does absolutely nothing Hanlon's Razor wins out here I think

Thank you! Will be doing this from now on.

I wish there was a way to report AI slop articles; it's frustrating clicking a link and reading a paragraph or two before realizing.

Sorry for being so aggressive. I suppose I'm just very confused at where you're coming from.