Or it just means anything shared on the internet gets RLHF’d / special cased.
It’s been clear for a long time that the major vendors have been watching online chatter and tidying up well-known edge cases by hand. If you have a test that works, it will keep working as long as you don’t share it widely enough to get their attention.
Just wanted to say “thank you” for this article. I found it years ago, probably not long after you initially wrote it and have preached it as widely as possible ever since, both as an IC and as an eng manager. It’s one of the best such tidbits I’ve ever come across!
Edited to add: and thanks for keeping it up to date with the new Swift version!
Put weapons on it (as already seen in current conflicts) and it becomes a seek-and-assassinate tool. Drones are cheap enough it could even be done en masse. It is a scary future, and it’s not far away at all.
S&R has always been a front for weaponized robotics, IMHO.
The last DARPA grand challenge (Subterrainean) had automated drone networks that could find and identify humans in caves and tunnels. They were at least up front about the military challenges in these environments. (https://www.darpa.mil/program/darpa-subterranean-challenge), but the nod at civilian first-responders doesn't seem fair. Honestly, is cave-in such a big civilian problem that we need to prioritize it as a talking point at all levels?
> Honestly, is cave-in such a big civilian problem that we need to prioritize it as a talking point at all levels?
Considering (1) the number of people who are employed in mining occupations, (2) the frequency of serious accidents in mines, yes. Particularly in developed countries, societies expect that great lengths will be gone to rescue or recover the victims, and mine rescue is incredibly dangerous work.
(1) BLS says ~200K in the US in 2024, although only a minority of them work underground.
(2) BLS says "underground mining machine operators" is the 9th deadliest job in the US, and that is with a large and well-equipped mine rescue system (MSRA says 250 teams across the country).
Mining is heavily mechanized and automated already, yet remains inescapably dangerous.
Pragmatically speaking, when someone falls off a roof or a tree, it doesn't turn into a highly public, high-risk, government-responsibility rescue mission. When someone gets trapped in a mine, it does.
(If you fall off a tree logging in Alaska, there is a good chance a USCG helicopter crew comes to your aid, but that is more of a "five minutes in the local news" story than "nightly news host reporting live on location" event.)
I suspect the vast majority of deaths in underground mining in the US aren't from cave-ins but instead from heavy equipment accidents.
According to [1] there were 8 deaths in underground machine operators category in 2022.
There's a more detailed table at [2] but I don't quite understand how this aligns with the first one (the numbers seem different, but I think the category is "Mining (except oil and gas)").
In any case the majority of fatalities are from "Transportation incidents" or "Contact with object and equipment". I think cave-ins would be classed as "Fires and explosions"
It's generally hard to say what's a "front" for what, unless you mean "what can you get someone to grant you research money for when you really expect to parlay the learnings into another topic."
Everything about the rocketry needed to get to orbit started from warfare purposes, for example. And ARPANET was a foray into how to build a disruption-resistant network for military purposes.
A read a comment here a while ago about "search and rescue" being a euphemism for military applications and that's the first thing I thought of when I saw this story.
I've been saying that any armistace on drones won't come until the US starts being hit by drone warfare. Especially by a foreign militia or nation state
I would not be so sure - a mind boggling number of drones and drone types are used in the Ukraine war, from small observation drones, over supply drones, drop drones, fast one way FPV kill drones up to almost regular drone swarm exchanges with 100+ drones going one way (indigenous Ukrainian drones one way, clones of Iranian Shaheds from the Russian side).
An oil terminal in Feodosia is still burning after the latest Ukrainina strike.
There was even a few cases of re-purposed ultra light aircraft serving as one way drones for ultra long range strikes on the Ukrainian side.
In another region Israel has to shoot down various terrorist launched one way UAVs almost regularly by this point & uses UAVs heavily by itself.
So while US certainly did pioneer UAV use, it seems to be it is getting eclipsed by other states in this area.
Precisely. It's about terror. The US having the political capital it does (among other things, being a Security Council member in the UN), Americans won't push their government to curtail drone use until and unless they're on the receiving end of asymmetric warfare attacks perpetrated with low-cost disposables carrying lethal payloads.
(Certainly not advocating for this, but noting that it's the most likely trigger to get the ball rolling on regulation of drones in military operation where very little currently exists).
Sadly collusion among market leaders is rampant even though there are (poorly enforced) laws against it, so it’s not too far fetched for companies to all lock arms and say they won’t do something. It’s all too easy to get away with that, especially when there are high barriers to entry (as with cars).
This has unfortunately been a known issue for a while, surprised it took this long. The article doesn’t mention it, but Norway already paused usage of this specific vaccine due to the same concern, way back in March 2021:
This is why gatekeepers are still a good thing, and now the gatekeepers are themselves somewhat democratized. Lots of people now have their favorite YouTuber / Podcaster / etc. and get book recommendations from them. And there is now a much bigger selection of gatekeepers to choose from given the variety of content on those platforms. It’s not perfect, but overall it is an improvement in many ways.
Now yes I would agree that a fully flat, bottom up, “everyone is a publisher” world where everyone is on equal footing (including the spammers) is impractical, but then again who ever said that would happen or work if it did?
It remains to be seen what effect genAI has on all this, though. Long term, it seems likely that the need for gatekeeping will only increase due to the inevitable flood of more and more generated junk.
I’ve been trying to divest from Amazon as a source for products, and am looking to use more traditional stores to curate my options. They have buyers that vet stuff to some degree, and if the product is bad, they’ll take it back. If enough people think it’s bad, they’ll stop selling it.
Exactly, I would never go to Amazon to look for books. I go to Amazon to buy a specific book. I already know what I want, Amazon is just the place I order it. Some stuff is also Amazon exclusive.
That's a good one, thanks. It gives me `18:09 UT`, which is `12:09 CST`.
The Precision Eclipse calculator does seem to provide correct results in the US (e.g., 1:40 PM for Dallas TX). So, it must be a Mexico timezone calculation error.
It’s been clear for a long time that the major vendors have been watching online chatter and tidying up well-known edge cases by hand. If you have a test that works, it will keep working as long as you don’t share it widely enough to get their attention.