Ariarule

While human biology hasn't changed in any meaningful sense, technology and US culture have changed dramatically.

1974 was over a half-century ago; it is extremely weak evidence with everything that's changed since then, at best.

The WSJ charts at the end should be taken with a huge grain of salt: https://news.gallup.com/opinion/polling-matters/544508/measu...

Yes, people do that. Karpathy made a utility to monitor it better years ago: https://github.com/karpathy/arxiv-sanity-preserver

No? It's almost certainly not AI overall despite the Webflow badge; it's pretty similar now to how the site looked in 2020: https://web.archive.org/web/20200806131653/https://www.beyon...

What do you not actually like about the site? I'm not a big fan of the trope of "hero" image slideshows taking up the whole screen, but if it's justified anywhere, it seems justified here where they're trying to make a game look cool, and the cards seem reasonably informative and not just vacuous. Yes, it is a "polished" design, and I wouldn't be surprised if they started with a template. What should they do; bad design to show amateurism? Would that be more or less slop?

That might be a valid motivation for keeping the rule, but as far I can tell it can't be the original reason for this rule as it predates passive indices in retirement accounts being that popular.

There used to be a professional layer between most people and raw information. Librarians, researchers, editors, fact-checkers: people whose entire job was to understand how information was organized, who produced it, what motivated them, and where the gaps were in any given source. You didn’t need to think much about any of that, because someone else already had.

That layer has largely dissolved. Search engines replaced the card catalog, algorithms replaced the reference interview, and AI summaries are now stepping in where a librarian’s judgment about source quality used to sit. What’s been left in place of all that professional mediation is a search bar and the assumption that you’ll figure it out. - https://cardcatalogforlife.substack.com/i/188856309/the-skil...

An unfortunate conclusion that smuggles in unwarranted good-old-days nostalgia to an otherwise excellent overview. The previous system that they're describing had serious problems, limited access to raw data compared to now, and could not have scaled up to the level of access to information that the internet provides.

The information environment prior to the early 2000s was quite terrible. We shouldn't pretend this was a golden age of truth-oriented gatekeeping, although there were certainly gatekeepers. There were a lot of misconceptions, errors, and unchecked biases on dead trees and in late 90s databases. The idea that those librarians, researchers, editors, fact-checkers were aligned with anything you might care about is also unsupported and dubious.

Not a single one of those explain the endorsement in OP of the completely-new "highly bizarre raccoon army conspiracy theory"

As someone that's been working on a game with Claude Code in a more human-in-the-loop, iterative fashion, I have to say that OP's claim that "LLMs barely know GDScript" does not match my current experience at all even though it seems to have matched yours. Maybe it was true a while ago in both cases; how long ago was your "vibecode" attempt? I've gotten completely fine GDScript and even decent perfectly functional if placeholder-quality TSCN files from Opus 4.5, 4.6, and Sonnet 4.6 with very little issue and no special tricks; just a CLAUDE.md file, the project itself, and going through plan mode before each change. I did start from a playable project with a fair amount of hand-written scaffolding already in place, and I have no idea if that would make a difference. Every once in a while there will be some confusion that I get something appropriate for Godot 3 instead of Godot 4, but never Python despite the similarities of the language.

They did not, you get the same date range and the same graph shape going to FRED and pressing the "1Y" option, and the series includes the first two months of 2026 so it's 12 months: https://fred.stlouisfed.org/graph/?g=1SGzm

However, the chart settings were actually modified to hide/deemphasize the earlier decline: the the index date was changed. 2025-02-20=100 in their graph, default of 2020-02-01=100 would have the chart start at 64 and rise to 71.44.

Per Scott Alexander, they're working on a bet targeting 2036 instead of 2029: https://www.astralcodexten.com/p/open-thread-421/comment/215...

Previously polled in 2023 in https://news.ycombinator.com/item?id=37036804

In light of recent developments, I thought it would be interesting to ask again.

It unfortunately didn't get very much attention here at the time (2021), but "Sigmoids behaving badly: why they usually cannot predict the future as well as they seem to promise" at https://arxiv.org/abs/2109.08065 is related.

I won't bother defending Google-style personalization as it exists for their search results, but since collisions in terminology across fields are common, it's not that hard to see how actual, thoughtful personalization could be useful. Someone searching for "Kafka" is going to want very different results based on whether they're thinking of software or literature. Opinions may also differ over the usefulness of sources, even for people ultimately interested primarily in facts; I find Kagi-style personalization (make your own domain list) very useful, but across Kagi's userbase Reddit is simultaneously one of the most lowered, most raised, and most pinned domains: https://kagi.com/stats?stat=leaderboard

This could be "bad, actually" if it gives an incorrect impression that power tools are unequivocally safe, rather than somewhat risky but usually safe when used correctly.

What exactly is so terrible as long as you're willing to take a small to moderate risk of getting a PVC pipe stuck in the ground and keep in mind the presented cautions -- consider the water irrigation-only/non-potable until tested (possibly even if tested, although that's not what the site says)?

"Occasionally, there will be a thin black bar at the top of the top bar, in memoriam of a significant figure in the tech/science community dying." -- https://github.com/minimaxir/hacker-news-undocumented#thin-b...

While the abstract does say there's "rising evidence that leprosy has become endemic in the southeastern United States", the actual title is "Case Report of Leprosy in Central Florida, USA, 2022" and that's a more honest description of what this is.

The second reference -- https://www.hrsa.gov/hansens-disease -- shows cases in the entire US bouncing from around 160 to 220 since _2011_ up through 2020: "Most (95%) of the human population is not susceptible to infection... Treatment with standard antibiotic drugs is very effective." The Florida dashboard which is the third reference shows 14 cases in that state total in 2021.

"No nitrates or nitrites added" bacon still has nitrites as they are generated by the production method: https://www.americastestkitchen.com/cooksillustrated/how_tos...

The Frankenstein veto; Wisconsin is infamous for them: https://en.wikipedia.org/wiki/Frankenstein_veto

Huh? The 14th amendment has already been part of the US constitution for 155 years.

Also, what is this about judges starting the amendment process?

That also reminded me of _Connections_, which notably made the same point in _the late 1970s_! It isn't like technology's gotten any easier to understand -- in total, "top to bottom" -- since then.

Why speculate? _Anthropometric Reference Data for Children and Adults_, Table 20 "Waist circumference in centimeters for adult males aged 20 and over and number of examined persons, mean, standard error of the mean, and selected percentiles, by race and Hispanic origin and age: United States, 2015–2018" is right at https://www.cdc.gov/nchs/data/series/sr_03/sr03-046-508.pdf

The mean in the very first row is 102.9 cm (the 40.5 inches in the OP), while the 50th percentile is actually lower at 101.8 (39.7 inches).

Besides that, as of now, https://poweroutage.us shows slightly less than 6 thousand customers out in Texas. Meanwhile:

"States and territories by customers out Maine 110,693 New York 38,041 Virginia 19,038 New Hampshire 18,056 Pennsylvania 13,040"

While groundbreaking for its time (started 1988!), the IEM is a very poor substitute for PredictIt. While I could offer a half-dozen reasons why, some of which were noted in CFTC Letter 14-130 for PredictIT in the first place, it'd probably do to point out that their "current" markets consist entirely of:

* 2022 French Presidential Election Markets (I haven't looked at these too closely that maybe there's a reason they're still listed as "current" on the IEM site, but wasn't this election back in April?)

* 2022 U.S. Congressional Control Markets (3 Markets: Senate, House, and Overall)

If you're trying to use prediction market prices as a gauge of the state of the world, IEM is missing dozens of countries, does not give visibility into US state-level elections, and does not include the result of primaries or political events besides elections that nonetheless impact elections. The 2024 US Presidential election is not even on the radar yet there.

Credit where credit is due for the trailblazing, they're worth a look when they do have a market on an event going, and it's notable for being sort-of a proof of concept for prediction markets. But while IEM may now be the best fit within a particular niche in light of what's happened to PredictIt, it's not in the same league.

It's all but forgotten now, but the early statements that Minecraft was going to be released as open-source really hint towards what might have been, and the magnitude of what was lost:

"Once sales start dying and a minimum time has passed, I will release the game source code as some kind of open source. I'm not very happy with the draconian nature of (L)GPL, nor do I believe the other licenses have much merit other than to boost the egos of the original authors, so I might just possibly release it all as public domain."

https://web.archive.org/web/20120614004339/http://www.minecr...

The top grossing films of the 1970s? The original Star Wars, Jaws, The Exorcist, Alien, The Godfather? Rocky 2 and Jaws 2 are also on the list so it's not like there weren't any sequels at all, but that group of films really _wasn't_ dominated by remakes and franchise films:

https://www.imdb.com/list/ls026560159/

By comparison, entries for the 2010s include MCU franchise films, Star Wars Sequels, Jurassic World (a sequel), The Lion King (Remake), etc:

https://www.imdb.com/list/ls026040906/

Going year-by-year doesn't really change things very much: https://en.wikipedia.org/wiki/List_of_highest-grossing_films...

This is reminiscent a bit of Ross Douthat's _The Decadent Society_, especially (but not only) in the discussion of the continuous stream of movie remakes.

Maybe not, it _might_ cause obesity: https://slimemoldtimemold.com/2021/10/11/a-chemical-hunger-i...

It's interesting that this downside wasn't even on the radar of this article in 2011, and indeed no serious consideration of possible downsides was presented at all, just a snide dismissal that "anyone who has tangled with the anti-fluoride campaign will realise that lithium lobbyists will have a long, hard road to travel."

I use Taskwarrior daily, but have a love-hate relationship with it after running into sharp edges and quirks several times. Hopefully this doesn't come across as too harsh as Taskwarrior is good as a command line utility for tracking one-off personal tasks -- maybe the best and I recommend it for that limited case -- but it has some warts that anyone newly looking into it should be aware of:

The recurrence system is listed as an "advanced topic" in the documentation. I gave up on trying to make heads or tails of recurrence in TimeWarrior after fighting with it for years and having way too many times where I had meant to make a change to just a particular single task, and then having unexpected changes to several tasks in a series, or to the template task, or having just that particular instance changed but then the recurrence ends up creating new tasks in a funny way, etc. I once had a particular edit to a reoccurring task create hundreds of instances the next time I looked at the task list.

Continuing on with recurring tasks, The Taskwarrior documentation notes "you never directly interact with... the template task. It is hidden for a reason. Instead, you interact with the recurring task instances" Hope you don't want make a change to future instances without changing any current instances. You can do it, but it's unnecessarily cumbersome.

Ultimately, I've actually found it easier to run a script outside of TW on an hourly basis that checks for the last completed task with a particular user defined field for each task that I want to recur, then creates a new instance if needed. This works nicely as there's no longer any linkage within TW between individual tasks in a "series", preventing a change from copying from from one to the others, and it's impossible to accidentally make a change to the "template" since the template is no longer a record within TW at all. I do note that the facts that there are python (and other) libraries that make this trivial and that it's possible to define fields on a task this way are notable strengths of TaskWarrior. Nonetheless, the fantastic CLI ergonomics for one-time tasks are just missing for things that happen more than once. That a template for a task is itself a task record seems like some really bad conceptual confusion in the design, and extremely kludgy.

A quick look at sibling comments or a general web search for discussion about TaskWarrior recurrence issues will show problems with recurrence are not uncommon.

The units used for the length of time in the urgency calculation for things like task age, due date are unintuitive and there is no explanation of this rather important point in the documentation (https://taskwarrior.org/docs/urgency.html for reference). As a specific example, I have a task entered 5 weeks ago which task info tells me is an age of 0.107, while a task entered 23 hours ago has an age of 0.003. This kind-of looks like 1 per year, but why? Also, time to due date is definitely not 1 per year. It also looks like age has a ceiling of 1 since this value appears for very old tasks, and that time to due date is similarly limited, but again, not documented.

The built-in reports are kind of nice, and the fact that you can define your own "reports" (really task lists with custom columns) is also nice, but it'd be even better to be able to be able to just throw SQL directly at it. There's a TaskWarrior-inspired utility "Tasklite" that does just this by storing tasks in SQLite, although I haven't tried it myself yet. The Tasklite criticism of TaskWarrior is also worth considering: https://tasklite.org/differences_taskwarrior.html

You can get tasks out of Taskwarrior in JSON format with "task export" and there are scripts to further transform this into other formats, but you need to do an extra step to import this into another tool for any analysis you might want to do.

Reports like ghistory.daily and ghistory.weekly can get long and unwieldy. There are ways to deal with this, but it's annoying, and makes me question if the developers themselves are using these reports on a list of completed tasks going back several years or not.

Taskserver is another area that seems poorly thought-out and over-engineered, similar to recurrence, and I'd caution anyone considering using Taskwarrior on more than one computer to look over the Taskserver Setup guide at https://gothenburgbitfactory.github.io/taskserver-setup/ while noting the following:

* Just copying the files around between Taskwarrior instances e.g. with Syncthing or something like that mostly works, but a lot of tools/attempts at mobile apps for TaskWarrior rely on having TaskServer. The sync issues noted in various places by using a file-level sync like this are definitely possible, but the practical implication for a task list that most likely belongs to a single individual is questionable.

* Per the guide, "There are two 'D's in TASKDDATA, and omitting one is a common mistake" in reference to an environmental variable defined, as far as I can tell, entirely within this document. Naming things is hard and this is admittedly a small detail, but why not remove the pitfall rather than just warning about it?

* You will need a CA certificate to "to sign and verify the other certs" -- "The command below will generate all the certs and keys for the server, but this uses self-signed certificates, and this is not recommended for production use. This is for personal use, and this may be acceptable for you, but if not, you will need to purchase a proper certificate and key, backed by a certificate authority."

* "Certificates coming from Let's encrypt have not been successfully used by anyone. Please remember that Let's encrypt only generates servers, but we need a client certificate as well."

* This is rather a lot of work just to be able to _add tasks on your phone_ while you're away from your desk or even just to sync tasks between what might be two or three desktops, and it's pretty clear they had more of an "enterprise environment" or a hosting service like InThe.AM in mind when they wrote it. Regarding that particular service, it is open-source but "Setting this up locally as a clone of Inthe.AM is, unfortunately, not supported" - https://intheam.readthedocs.io/en/latest/

By contrast, there's another project dstack which is also Taskwarrior-inspired but "uses git to synchronize instead of a special protocol" at https://github.com/naggie/dstask -- this is also something I haven't tried it myself yet but it looks interesting, although doesn't itself help with getting tasks on/off a phone as far as I know.