Hacker News Reader

Slightly off topic, but thanks to your first link I was just doing some satellite image sightseeing of Wuhan and noticed that on Google maps the streets are offset about 3 blocks east of where they should be.

It's fairly obvious here [1] where the curved road extends out into a lake.

Not saying this means anything, just found it amusing.

1. https://maps.app.goo.gl/reVDkMAhfMRbv1uJ8

It's fairly well-known [0] that China messes with maps data and forces Google to do so.

[0]: https://en.wikipedia.org/wiki/Restrictions_on_geographic_dat...

Is this similar to how the US used to mandate that civilian use of GPS was off by a certain amount?

Yes, it's ostensibly still for defense purposes in China

A bit different. The GPS restriction was to prevent GPS on a device moving over 400 mph (I think that's roughly the speed?), and to limit accuracy to within a few meters rather than a few centimeters. The restrictions were intended to prevent the use of civilian GPS systems in precision missiles.

Not a lot of civilian uses require anywhere near that speed or accuracy.

It's a lot harder to justify grossly inaccurate geographic data as not hurting civilian uses.

you're referring to the CoCom Limits on GPS receives, which limits functionality when the device is moving faster than 1,900 km/h aka 1,200 mph) and/or at an altitude higher than 18,000 m (59,000 ft), so you can't build a home made ICBM with it. Technically it's supposed to be and and not or, and high altitude amateur aerial ballonists tend to hit that flight ceiling, and so have a list of chipsets they can use in their balloons that don't stop working when they get too high.

https://en.m.wikipedia.org/wiki/Coordinating_Committee_for_M...

I never understood how this works, at least effectively. Isn’t this just a device DRM? If you can build an ICBM you can probably hack a gps receiver, or build an unlocked one, no? If that encryption key is in every gps device, surely it must be available on eg darknet. And thus you could use existing gps infrastructure (ie avoid the expensive part of launching your own satellites). What am I missing?

There's no encryption, it's just a limit in the firmware.

What am I missing?

The collective intelligence --- or lack thereof --- of the bureaucracies that come up with these laws.

It wasn't a pre-set error, rather each satellite was originally configured to broadcast a low-precision coarse signal and a high-precision signal and the high-precision signal could only be decoded using an encryption key. They later released the encryption key when civilian applications took off.

iirc there still are several signals broadcast, a rare public high precision signal (for land surveyors), a high precision stream (military) and a high enough precision stream (public navigation)

Sort of. It uses a pre-defined transformation that doesn't interfere with street navigation. If you're on a street it will precisely reflect that value as well as any other GPS. But it makes it difficult to perform purely GPS-based instrument navigation, which in theory makes it harder to conduct eg missile strikes.

No, the US one is basically error(like noise) injection and has been disabled: https://www.gps.gov/systems/gps/modernization/sa/

The Chinese one is effectively a warped coordinate system, see GCJ02: https://en.m.wikipedia.org/wiki/Restrictions_on_geographic_d...

Huh, I guess I wasn't aware of this previously. Figured it was a bug or something.

Considering the satellite imagery exists this seems silly. Glad I didn't drive into a lake!

I just want to add to this that when I first heard about this virus I searched the WIV on Google Maps. I believe that was 14 January 2020. When I searched it again a few weeks later, the location on Google Maps had changed.

I have no screenshots of this, but I did find it quite odd at the time.

Addresses of things change on Google maps. They fix problems and find issues and it's entirely possible that an address changing means something was updated. Or there's an international address conspiracy that Google is part of, you'll have to decide for yourself!

This change of address is initated by an authority of that region or owner of that business. What exactly is your motivation is spreading FUD on this matter?

I don't think it's fine. When I worked at Google, we got updated address information for businesses periodically and we had an official process that applied updates to the data we had stored. It was actually a colossal pain in the butt. I could see that still being the way that addresses can get updated?

omniglottal, I wish you'd comment further about address info changing. Like the other responder to you, third parties can ask to update address info. It's absolutely not conspiratorial. Can we agree it's not a conspiracy?

You don't have to be a business owner to change the location of a business on Google Maps. They rely on the general public to correct inaccuracies - I've done this a few times for certain locations near me. Presumably they use some kind of scoring system to determine whether the request to make the change is valid.

It's fine on Apple Maps.

That's Chinese map data obfuscation. The idea is to make it marginally harder for an adversary to target missiles using public map data.

Which is useless in this day. I was in China last week and Google Map’s street layer was aligned correctly to where I was, but the satellite imagery was not.

This demonstrates that Google already knows the correct coordinates of street in China, including those of an airport finished in 2021. For some reason they have spent no time manually aligning the imagery.

Coordinates on the globe are constant whether China likes it or not, my only guess is that Google doesn’t want to spend time fixing data in a country that blocks them entirely.

The wiki above explains it. Technically they are supposed to purchase a shift algorithm.

Google is doing no business in China, so they don’t “have to.” They could easily bypass that by internally mapping the imagery to the vectors, they have thousands of developers who could do this.

Even if they had to buy it, they’re not buying it because again they don’t do business there.

Who does Google buy map data within China from?

>became sick in autumn 2019, before the first identified case of the outbreak, with symptoms consistent with both COVID-19 and common seasonal illnesses

I was in Thailand in february 2019 and came down with SARS-like symptoms that were very similar to Covid-19 once that became known.

SARS-like viruses have been going around for a while, meaning they affect the respiratory system. Doesn't mean it was covid-19, what even differentiates SARS from covid-19? I have no idea, I'm just a layman who happened to get sick.

Covid-19 is the disease caused by the SARS-CoV-2 virus.

From [0]:

> Through DivErsity pArtitioning by hieRarchical Clustering-based analyses,5 the newly emerged coronavirus was deemed not sufficiently novel but is a sister virus to SARS-CoV, the primary viral isolate defining the species. The SARS-CoV species includes viruses such as SARS-CoV, SARS-CoV_PC4-227, and SARSr-CoV-btKY72. SARS-CoV-2 is the newest member of this viral species. The use of SARS in naming SARS-CoV-2 does not derive from the name of the SARS disease but is a natural extension of the taxonomic practice for viruses in the SARS species. The use of SARS for viruses in this species mainly refers to their taxonomic relationship to the founding virus of this species, SARS-CoV. In other words, viruses in this species can be named SARS regardless of whether or not they cause SARS-like diseases.

[0] https://www.ncbi.nlm.nih.gov/pmc/articles/PMC7133598

> SARS-like symptoms

Aren’t SARS-like symptoms just flu-like? I heard the same exact thing from an European friend and again I think it’s Barnum effect-like.

Given the transmissibility of the virus, I doubt that it was in the wild unnoticed for 12+ months before suddenly affecting millions in weeks.

Sort of but what made me think I had SARS specifically was the fact that I had this weird cough that persisted for a month after I recovered from the worst of the flu symptoms. That was very unusual to me.

During the actual week or two I was sick I had all the symptoms of Covid, diarrhea, shortness of breath, and the rest of normal flu symptoms. And when I recovered I remember eating a protein shake just because I couldn't eat anything else and I needed energy, and I remember not tasting it.

I'm not saying it went unnoticed for a year, it's just an interesting thing that happened to me that I didn't pay attention to until after covid was detected.

I even know how I got sick, it was on the plane from Copenhagen to Doha. Two thai guys were sitting next to us and one of them was visibly sick. We felt pity on him and offered him a pillow or candy sometimes. Then 1 day after we landed in Udon Thani we both got sick.

You didn't have SARS mate

I haven't looked at where Ben Hu's lab is, but epidemiological distance in urban areas doesn't necessarily follow map distance. In particular, proximity to a shared subway or bus line can be more important than physical proximity, since many commuters will put up with longer travel times if they can tune out, and a packed rush hour subway car is a great place for the spread of airborne infectious diseases. (In US cities one might consider two homes epidemiologically close despite a 10 mile separation if their children take the same school bus.)

The CDC is smack in the center of the city (on the other side of the river). The WIV is way outside the urban area, past a very large industrial park.

The distribution of cases also strongly contradicts you: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC9348750/figure/...

Cases were roundly distributed, not along any particular artery. Note that the WIV is not even close to being in frame there- that's how much farther away it is.

AFAIK, there is a Biolab of a Wuhan University rather near the market. BSL3. Choose your odds.

>As we’ll see later, there may actually have been two jumps from animals to people at the market.

Mixing fact with speculation is never a great sign for an argument. Many people pushing the lab leak theory are doing this, but the people pushing against it don't need to start. If there were two spillovers, that needs a very good explanation, particularly because events like this tend to fall into the categories "one" and "many" — if there were two spillovers, why not three or four? Why not a whole gaggle of infected raccoon dogs?

> Based on number of visitors, Worobey estimated the odds at only 1 in 10,000 that the market would be the first superspreading event in Wuhan.

This is based on a location analysis of early cases around Wuhan. That doesn't imply that the first event was at the market, only near the market. In particular, the market is across the street from a railway station and one block away from the Huazhong University of Science and Technology Tongji Medical College Union Hospital. Huazhong University of Science and Technology is one of China's top ten universities and might have a pretty good hospital.

Have you read the brace of Science papers that analysed the market data

> Mixing fact with speculation is never a great sign for an argument.

You're the only one who's speculating, because in the page that sentence links to a scientific paper: https://zenodo.org/record/6342616#.Y_WrmnbMJD-

> In particular, the market is across the street from a railway station

Which can be trivially ruled out, because then the spread would have been all up and down the rail line. The first infection point would have been on the train, among all those people getting off at all different stops. It would not have suddenly become infectious just as that person stepped off the train.

> and one block away from the Huazhong University of Science and Technology Tongji Medical College Union Hospital.

The evidence for zoonotic origin is how people regularly found or contracted SARS-like coronaviruses from animals. The evidence there was any relevant research at that university is zero. It's not a government lab, those projects would have been public.

Neat, thanks for that last link. It could use a table of contents but it did contain more information related to the DEFUSE grant proposal which is what I have thought of as being indicative of a lab leak.

Sadly that article doesn't disprove that theory, but it does detail its weaknesses which is a good enough jumping off point for further reading.

Which still doesn’t really draw the line of how an outbreak occurred at the wet market. Also common seasonal illness isn’t exactly a smoking gun

not really typical for healthy researchers in their 30s to end up in the hospital with seasonal illness symptoms though.

Perfectly normal for regular, healthy folks to occasionally come down with seasonal flu. Sometimes people get sick, it happens.

The issue was that a whole bunch of them did, rapidly, and in non-trivial numbers.

Did they "end up in the hospital" or did they go to the hospital to get a diagnosis? I live in a city, and my PCP is in the local hospital. "seasonal illness symptoms" sounds FAR more like they were just sick and went to the doctor, like anyone else would. For medicine.

I read it was hospitalized, but who knows with this kind of reporting.

And what exactly does "hospital" mean in China? I know in India a hospital can basically be a small clinic, not unlike an urgent care center in the USA.

This is a very large, very modern city. I'm very willing to believe they went to real hospitals.

https://en.wikipedia.org/wiki/Wuhan#/media/File:Wuhan_Yangtz...

People go directly to hospitals for primary care in China. It's not the same as the US, where people go to a general practitioner first.

Outbreaks can only happen in places of high density/congregation, by the nature of the required proximity. It could have been dancing around the perimeter for some time.

Maybe my perspective is incorrect, but this seems trivially possible to me. This problem with proximity is why kids didn't go to school.

The wet market was one of the less densely trafficked areas of the city: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC9348750/figure/...

Check out the other figures in that paper. The epicenter was the wet market itself.

Here is a thread explaining the flaws in Worobey's paper https://twitter.com/danwalker9999/status/1595653898572042240

That's not really relevant. The cases are all still very far away from the WIV, which is itself far from anywhere the thread implicates eg residential areas.

I also suspect the thread is just flat wrong- the people most relevant to viral spread are NOT the ones who are only spending part of their day in the city. Its the people who live there and spend all their time there.

2/3rds of the population of Wuhan lives in the urban districts: https://en.wikipedia.org/wiki/Wuhan#Administrative_divisions

The people commuting in/out are a small minority. No model is perfect.

.... written by a geoscientist. There are limits on the ability to transfer knowledge between domains (although this doesn't mean the geoscientist is necessarily wrong, and often even amateurs can find big problems with papers that got through peer review).

The geoscientist in question is actually arguing that some of the flaws in the paper are due to the authors' lack of expertise in geospatial analysis.[1]

[1] https://twitter.com/search?q=(from%3Adanwalker9999)%20expert...

that may be- I am not qualified to say after reading the twitter thread, but my money is still on the peer reviewed paper written by the virologist, based on a large collection of priors. Note taht the "geoscientist" is really a hydrologist who works on oil and gas, so their ability to comment on the centering model is still suspect. From what I can tell, "dawalker" is just a retired hydrologist with no special skills in this sort of analysis, especially in an biological contexgt.

We are dealing with situations where smart people are using social media to promote propaganda in response to academic publications. Twitter as a medium is not really the best place to carry out these sorts of discussions- for exmaple, when I read walker's tweets, they jump all over the place and make many different arguments which are incomplete. If he really has a problem he should take his case to retraction watch and get the paper retracted if it contains invalidating errors.

If you read the linked thread, you might actually be able to engage with the argument on its merits, without resorting to ad hominems and appeals to authority.

I read the thread. It's not that interesting.

The peer-reviewed literature on COVID origins has been unusually bad. I assume you don't think SARS-CoV-2 came from pangolins; but Nature published "Isolation of SARS-CoV-2-related coronavirus from Malayan pangolins", and took more than a year to correct it. (They said the virus was widespread among pangolins, based on multiple positive samples; but in fact multiple papers had been written about one batch of smuggled pangolins.)

Map-based arguments for the exact site of introduction seem generally like noise to me. Worobey has made aggressive claims in preprints and media interviews, but even his Science paper falls back to just "epicenter", a weaker term without standard epidemiological meaning. SARS-CoV-2 must have been introduced into the Americas at air and seaports; but even with the advance warning to public health officials, that's not where the first clusters were found.

Are you a professional scientist with experience in this area or not?

Definitely not. But if you just want to hold up credentials, then David Baltimore seems pretty convinced that SARS-CoV-2 arose unnaturally, and I'm pretty sure a Nobel prize beats a paper in Science. So I hope you now agree credentialism is dumb?

Some areas of the origins debate require deep knowledge of the evolutionary biology of related viruses (e.g. the extent to which the coding of the FCS suggests engineering), but most of the debate is understandable with only basic molecular biology and math. From your post history, I'd guess you're better-placed to understand something like Pekar's epidemiological model than most of his reviewers were. So if you're interested enough to comment here, then I'm not sure why you wouldn't try?

I don't change my opinion of David Baltimore or of SARS-CoV-2 based on his belief (I am actually completely undecided and have been waiting for convincing evidence of any kind for some time). Baltimore is actually a bit kooky, he's not the first Nobellist to do/say dumb things: the inventor of PCR, Kary Mullis, was an HIV-causes-AIDS denier https://www.ihv.org/news/2023-News/USA-Today-Fact-check-Rese... Anyway, Baltimore recanted some time ago, agreeing that he was far too strong in his claims (this is a common problem in science- people who are convinced often get really angry and insistent about their beliefs).

To be honest, I'm so unhappy with modern medical literature that even starting to try to understand a specific model would be highly unpleasant for me- generally, my experience has been that once I start pulling the threads on a paper that isn't strictly strong quantitative biophysics, the sweater comes apart. The vast majority of medical literature requires extensive analysis and a thorough understanding of all the context before you can even really start to make useful criticisms.

My general statement remains: my priors still place a higher weight on trusting papers in major journals that haven't been retracted yet, than on out-of-domain scientists on twitter. It may not even be true in this case- perhaps dawalker is actually totally right and you can't conclude much from the WOrobey paper. But more importantly: nobody yet has shown any true "smoking gun evidence" for the origin of SARS-CoV-2. I am not even going to get remotely excited about it for at least a decade, since that's about how long it takes for the community to calm down and start thinking rationally again.

I'd tend to agree that Baltimore initially overstated the significance of the FCS, though his updated position seems close to my own. I'd certainly agree that whatever Kary Mullis saw wasn't a physical glowing raccoon, that megadoses of vitamin C won't cure cancer, etc. That's basically my point, though. Prestige--personal, institutional, or otherwise--is no guarantee of quality.

It's not meaningless either, and in general I'd also place higher weight on a paper in a major journal than on a random Twitter thread. As to COVID origins in particular, I believe major journals have published unusually poor-quality work. This continues an unfortunate pattern set in past biosafety incidents (1977 flu, Sverdlovsk, etc.), of dismissing the possibility of an unnatural cause until the evidence was incontrovertible.

The math in the Worobey papers isn't really that inaccessible, and I'd consider some of it as good work if it weren't so grossly oversold. I agree there's no definite evidence for any origin of SARS-CoV-2, and I don't think the article linked here adds much. I do think revelation of the DEFUSE proposal did, and that UNC's and the EHA's prior silence on that is inexcusable.

Definite proof may never come, but for now the American government continues to fund high-risk virological research. That seems terrifying to me. Reckless agricultural practices also continue to risk novel pandemics (in the West too; our use of antibiotics on healthy animals may get remembered as a crime against humanity). The risk of reckless virological practices is additive to that though, and much more easily controlled, simply by defunding. I therefore believe it deserves attention now.

>in general I'd also place higher weight on a paper in a major journal than on a random Twitter thread.

The Alina Chan archetype would beg to differ. The journal gatekeeping on tbis topic has been off the charts.

"Epicenter" does not mean "origin", by any definition of the word. The epicenter is a statistical thing, in this context, not an origin thing. The epicenter of COVID in the US was New York [1]. The origin of COVID was not New York. This is no different than the epicenter being the wet market, but the origin possibly being somewhere else.

"Epicenter: the part of the earth's surface directly above the focus of an earthquake"

It's all probability of intercept. Density and frequency are important. Relative-to-a-mall probably isn't super useful, in a city with 11 million.

[1] https://www.cdc.gov/mmwr/volumes/69/wr/mm6946a2.htm

The higher the r0, the less proximity is required.

The market is extremely close to the lab. Within walking distance

That's the CDC, a totally unrelated building which even then isn't very close to the wet market, which is not a very popular spot in the city. The WIV was many miles outside the city.

Supposedly there was a problem for a long time with researchers selling used lab animals on the street to make a bit of extra cash.

That's ridiculous. The animals are transgenic mice.

Labs experiment on all sorts of animals. Serial passage is often done with ferrets for example.

“Supposedly”? That kind of claim really needs a level of proof beyond “it would be politically convenient if this were true”.

If a single staff person caught a virus, then anyone they interacted with and gave it to went to the market, then that's all it would take.

Incredibly unlikely. There are thousands of locations that are far more trafficked than the wet market: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC9348750/figure/...

just a guess, but mid august was the first outbreak and September 21st was when they realized it was out of control.