Or places with a terminally uncool reputation. I'm still on Tumblr, and it's actually quite nice these days, mostly because "everyone knows" that Tumblr is passé, so all the clout-chasers, spammers and angry political discoursers abandoned it. It's nice living, under the radar.
I wonder if in the long run this will lead to the ascent of NixOS. They seem perfect for each other: if you have git and/or a snapshotting filesystem, together with the entire system state being downstram of your .nix file, then go ahead and let the LLM make changes willy-nilly, you can always roll back to a known good version.
NixOS still isn't ready for this world, but if it becomes the natural counterpart to LLM OS tooling, maybe that will speed up development.
I hope I turn out to be wrong, but the most convincing explanation I've seen for the "why" is that the 1945-2000 period was an anomaly, and now we're reverting to the mean: despotic governments, frequent wars for territory, and massive wealth inequality leading to powerful oligarchies as the only other important political players aside from the despot. This was the norm for the overwhelming majority of human history and perhaps it was massively hubristic to think we had escaped it for good.
It's not the only anomaly. There was a previous period of long peace between 1815 and 1914, between the Napoleonic Wars and WW1.
This balance of power was carefully set up in the Congress of Vienna following the (first) defeat of Napoleon, and was ended by the ambitions of a Kaiser who desired the prestige of globe-spanning empire yet couldn't diplomacy his way out of a wet paper bag to realize that empire without bumbling into war.
I think ADS and Merriam-Webster got it right. "Rage bait" and "parasocial" were the WOTY 2-3 years ago; not that they've gone away, but they were of a previous moment. "Vibe coding" is too specific, and "67" is trying too hard to be Hip and With It.
"Slop" is the word that perfectly captures what so much of 2026 was about, and I heard it from every direction, including people not into tech at all.
I thought this was already well-established public information? That fentanyl came mostly from China was never in doubt, what people were arguing about was whether this was happening with the tacit approval of the Chinese government. Then in 2023 China cracked down on it, and supplies dried up. Whether that was because it was a big enough issue to get their attention, or it was on purpose and they decided it was no longer serving their interests I suspect we'll never know, but I definitely read multiple articles in 2023 about the fentanyl crackdown in China.
Biden era cooperation with China on the issue was at the heart of this.
It wasn't about the direct supply of Fentanyl, or even (by that stage) the direct supply of Fentanyl precursor drugs .. (that gangs used to industrial shed chem lab into Fentanyl) ... this was cutting back and limiting bulk supply of the precursor precursors to shady onselling networks to starve the labs.
Was going well (as per the paper) until US / China relations went in the toilet.
Mexico also began enacting extremely heavy handed tariffs against China and other Asian exporters like South Korea, India, and Vietnam in 2023 onwards [0][1][2][3] in order to protect their domestic manufacturing capacity against an export-driven supply shock, which hit Mexico really badly in the 2000s [4].
> Was going well (as per the paper) until US / China relations went in the toilet
Yep, but as long as Mexico continues to enact trade barriers to protect against an Asian export shock, the APIs needed for synthesis will remain difficult for organized crime to acquire.
Already, cartels have begun tariff arbitraging by targeting the CEE and the Balkans as a new base for synthetic opioid operations [5][6][7], especially because Romanian [8] and other CEE gangs had been collaborating with Mexican organized crime on financial and human trafficking crimes in Mexico for over a decade now.
The biggest takeaway that deserves stressing over and over again is that Things Take Time .. it generally takes 18 months and longer to substantially impact global flows.
The work has to be put in early, kept up in practice, and results are often credited to political actors down the road of time.
People are always talking about this precusor from China, but I have no idea what this precursor is. Are they chemicals that are useful for lots of things or is it only useful for this? Because if it is the former, then China is just selling regular ass legal chemicals because they are the worlds number 1 supplier of manufactured goods.
Fun fact: The "traditional" way of making it was extracting piperine from black pepper and reacting that with nitric acid. Nowadays it's made in other more industrial scalable ways.
But yes, the same base precursors (and their siblings) are used to manufacture ADHD meds (ritalin/concerta), antidepressants (paxil), insect repellents (picaridin/bayrepel), hair loss medications (rogaine), allergy meds (claritin), anti-psychotics (haldol), anti-diarrhea meds (imodium), and many others. And also PCP.
So it's non-trivial to prevent. The core of the issue is that the one pot Gupta method came about in the 2000s and it made it extremely easy to manufacture fentanyl using these basic building blocks for so much of the pharma industry. Not only just making it easier to source ingredients but it took out all the steps and made the process easy as hell as well.
The challenge in international drug operations was not to get China to stop selling bricks to house builders to but get China to cooperate in stopping the sale of bricks to groups that only use bricks to throw through windows and at heads.
That’s tricky because if the US asks to stop the sale of precursors used for making medicines to an organization they name, it’s not always clear whether they are doing illegal sanctions or legitimate activities with the consent of the country in question.
China probably just wants to be a neutral supplier and stay out of it.
Despite the difficulty the former US administration was able to diplomatically achieve cooperation from China on this matter which bore fruit and gained traction until a seris of wild accusations and tariffs from a later administration killed a number of US / China working arrangements.
Fentanyl is so potent that just one lab can easily satisfy all the US demand with it, around 10kg a day. That's also why it's ridiculously hard to fight, one smuggled barrel of pure product can supply the entire US for months.
So no, there is no "supply shock". There's just more free Narcan (naloxone).
Cocaine death decreases is the hard thing to explain with either theory, supply or naloxone. Fentanyl supply doesn't affect cocaine in any way and naloxone doesn't work on a cocaine OD.
Maybe some percentage of cocaine deaths are misattributed fentanyl deaths?
I also wonder if there's any link to the Oxycontin reforms. Perhaps now that prescription is reigned in, we are seeing a lot fewer oxy->fent cases which has cut back on the deaths.
Or maybe it's actually that the drug dealers have gotten more careful. Drug dealers don't want to kill their clients, so maybe they've been purposefully diluting to make sure they get repeat customers.
> Perhaps now that prescription is reigned in, we are seeing a lot fewer oxy->fent cases which has cut back on the deaths.
This is definitely part of the story. When your primary source of new addicts is prescription opioids and you cut down on the prescriptions then over time, as people die off from OD, then the OD rate is bound to drop.
The most tragic part of it, to me, is that it's usually the people who got clean who eventually OD. Once they've been clean for a short time then their tolerance for the drug drops drastically, then if they break down and do "just one dose" they make the fatal mistake of thinking they can still handle the same amount they were used to doing before. This exact scenario happened to multiple more or less close acquaintances of mine, even people who were aware of tolerance and should have known better. I'm fairly sure that it's extremely common.
Pure cocaine overdose deaths are relatively rare. Only around 5% of cocaine deaths involved pure cocaine, it's almost always mixed with something else.
> I also wonder if there's any link to the Oxycontin reforms. Perhaps now that prescription is reigned in, we are seeing a lot fewer oxy->fent cases which has cut back on the deaths.
Prescription pills have been a non-issue for a decade by now.
> Or maybe it's actually that the drug dealers have gotten more careful. Drug dealers don't want to kill their clients, so maybe they've been purposefully diluting to make sure they get repeat customers.
Yup. I think that's exactly it.
The major reason for fentanyl deaths was not unintentional overdose because of poor pill quality. It was way too easy to end up with 1mg instead of 500mcg during pill mixture preparation. So _reducing_ the amount of fentanyl per pill results in a better safety margin. And users can just smoke another pill if one pill was not enough to get high, after all.
And yeah, it's just possible that the more reckless drug users are just dead by now. But to be clear, it's still absolutely horrible. We're still above the 2021 level.
"Drug overdose deaths may involve multiple drugs; therefore, a single death might be included in more than one category when describing the number of drug overdose deaths involving specific drugs."
https://www.cdc.gov/nchs/nvss/vsrr/drug-overdose-data.htm
Someone who overdosed after taking cocaine contaminated with fentanyl would be counted as a cocaine ODD.
The Oxycontin "reforms" caused the fentanyl crisis to begin with. People often moved onto heroin and fentanyl because pharmaceuticals were no longer accessible. The massive spike in overdose deaths begun after the decline in opioid prescriptions. See the Opioid Prescriptions & Opioid Overdose Deaths graph here https://drugabusestatistics.org/opioid-epidemic/
Revealed preferences suggest otherwise and that matters because he says a lot of things, often contradictory.
Is it just another Epstein diversion maybe?
Oil story doesn't stack up though:
- it's heavy sour oil, the tar like substance isn't economically extractable without an almost doubling in barrel price
- cheaper (existing infra) sour supply chain with Canada already meets US shale light sweet oil blending needs for a long time
- decided on maintaining stability of existing Venezuelan regime over supporting regime change
One thing that lines up so far is it does seem to be disproportionately effective at displacing column inches spent on the pending bringing to justice of Epstein entangled elites. Disproportionately because that pursuit of justice seems quite resilient in resisting partisanship breakdown.
B2B transactions like this are handled fine with contracts and lawyers all the time, I doubt it would be an issue. In the worst case, the utility could own the recharging module on the drone, just like they own your power meter.
Yeah, I think what Workaccount2 is not realizing is that there's no bottom to "you have higher risk factors, why should I pay for you?", and so once you start down that way you may not like where it ends up. Some hobbies have higher injury rates, why should I pay for your health care if you choose to play those? Some parts of the country have lower life expectancies, why should I pay for your health care if you choose to live there?
The actual realization, which usually comes years after the realization that there is no bottom, is that there is no top either.
The battle along the spectrum of privatizing gains (lower healthcare premiums for a healthy lifestyle - high premiums for unhealthy lifestyle) vs socializing losses (paying $20/mo to get $1200/mo of care - paying $1200/mo for $0/mo of care) is constant and boundless in either direction.
On end, it's "national insurance", functionally equivalent to fully-tax-funded healthcare like the NHS or the German system with several providers competing but regulated to near identical results, but moreso as the UK and Germany also has private care; on the other, it's the absence of insurance.
This is a macro problem larger than health insurance, and exists everywhere from employee bonuses, high school group project grades, handicap parking, gas prices, Everest summits, to gas prices.
Those might all seem wildly disconnected, but they all have systems of unfair allocation to compensate for unequal outcomes.
Generally national healthcare programs are entirely dependent on young healthy people paying into the system despite rarely needing it, and then hopefully enough dieing quick deaths or having multiple children to cover their costs. These rebalancing systems are artificial and humans are generally terrible at managing them.
This is a good article, although I wish it had talked a little more about the standardization (or rather, the lack thereof) in Markdown. I get why it didn't, it's trying to be positive about something that is an overwhelming net positive for the world, but I think a "warts-and-all" treatment of the history would be more honest.
I appreciate that Gruber brought this very helpful thing into the world, but OTOH he was such a prick about the whole Standard Markdown debate, for no real reason other than ego. And it resulted in Markdown remaining an ill-defined standard to this day, with occasional compatibility issues still cropping up even though most platforms support most of "Github-flavored Markdown" (itself a stupid name and indicative how badly this has gone).
You've pretty much said what I was going to say. I think John was absolutely inspired in coming up with Markdown, but was a terrible steward. Or perhaps I should say he was unwilling to steward it.
My impression was he pretty much threw up a Perl implementation that was good enough for what he wanted, refused to refine it at all, and declared by the power vested in him by nobody in particular that if any parser implementation differed in behaviour to his (like, to fix bugs or make it better), wasn't true Markdown and wasn't allowed to be called Markdown.
Or perhaps I am being uncharitable in my interpretation of events.
I _don't_ think it was just ego. I think it was a smart strategy because formal standardization tends to bring in complexity, and just letting folks go off on their own and document their own usage (or "flavors") ends up being Good Enough in actual practice. It sucks from a standpoint of what I personally find satisfying, to be clear. But based on what I've seen over the last 20+ years, it is the strategy that is much less likely to yield a format that gets captured by giant companies that own a hyper-corporate standardization process that eventually gets enshittified.
Thanks for responding, Anil! Like I said, I really liked the article overall.
I don't agree that the Standard Markdown effort, had it succeeded as originally laid out, would've resulted in "hyper-corporate standardization". I mean, one of the main actors was Jeff Atwood, just about the least "hyper-corporate" guy there is. And I also don't really see any possible trajectory for Markdown to get "enshittified": after all, for the most part it's just plaintext with formatting conventions that existed way before it. Even if some corporate entity had somehow badly messed it up, markdown.pl and the other pre-existing implementations would have continued to exist.
I haven't tried using agents to make a full editor, but Claude Code and Gemini CLI are actually quite good at writing Obsidian plugins, or modifying existing ones. You can start with an existing one that's 90% of what you want (which tends to be the case with note-taking/PKM systems: people are so idiosyncratic that solutions built by others almost work, but not quite) and tweak it to be exactly right for you.
My own Obsidian setup has improved quite a bit in the last couple months because I can just ask Claude to change one or two things about plugins I got from the store.
Writing or tweaking plugins is great, but it's not a paradigm shift (and risks a lot more toil because now you have to be your own PM or deal with patches/merges, on top of being a reference librarian and copyeditor etc). I feel like if you have a quasi-superintelligence in a box which can run your PKM for you, and you were designing from the ground up with this in mind, that Claude Code is only going to et much better & cheaper, you would not be settling for 'write or modify an Obsidian plugin'. You would get something much different. But 'write a plugin' is basically at 'horseless carriage' level for me.
What I have in mind is something far more radical. There's an idea I am calling 'log-only writing' where you stop editing or rearranging your notes at all, and you switch to pure note taking and stream of conscious braindumping, and you simply have the LLM 'compile' your entire history down into whatever specific artifact you need on demand - whether that's a web of flashcards or a blog post or a long essay or whatever. See https://gwern.net/blog/2024/rss + https://gwern.net/nenex , combined with the LLM reasoning and brainstorming 'offline' using the prompts illustrated by my poems.
That's fair, I guess when I hear "radical overhaul" when discussing PKMs I immediately start worrying about the overload and burnout that doomed my first attempts at Obsidian (see my sibling comment), whereas right now I have a system that works very well for me, especially now that I can just ask Claude to scan the whole directory if I want to ask it questions. But if you do come up with some new blue-sky vision for PKMs, I'd love to at least take a look.
This is the way. If you symlink the .claude directory (so Obsidian can see the files) then you can also super easily add and manage claude skills.
I've spent 20 years living in the terminal, but with claude code I'm more and more drafting markdown specs, organizing context, building custom views / plugins / etc. Obsidian is a great substrate for developing personal software.
reply