HNHacker News
TopNewBestAskShowJobs

jcahill

137 karma · joined November 24, 2016

submissionscomments
jcahill··on Banana Republic Corruption
I read Murray's article and the FT article.

I don't see:

    (a) any disagreement between FT and Murray on basic facts
    (b) any overtly conspiratorial thinking from Murray
    (c) any clear reason to give deference to FT, or
    (d) any justification for labeling Murray's article as "absolute nonsense primary "research"".
I do see:

    (e) a few concrete criticisms of Murray's article, and
    (f) a few concrete criticisms of FT's article.
FT in agreement with Murray, via non-AMP link[1]:

> However, the process has been criticised for awarding large sums of taxpayer money to several companies that appear to have no record in supplying PPE and have small balance sheets or poor recent financial performance. It has prompted concern over the extent of due diligence that was conducted before the contract awards. […]

> So far the government has published about 80 of the PPE contracts. The largest was a £252.5m contract awarded to Ayanda Capital, a family investment firm that specialises in “currency trading, offshore property, private equity and trade financing”, according to its website. […]

> When contacted by the FT, Ayanda said it was a UK limited company that paid UK taxes and was “happy to confirm that the contract has been successfully fulfilled”.

Possible criticisms of Murray's article:

    1. Loginwalled and dead links.
    2. Speculation about the pricing of intangible assets under Ayanda Capital's management.
    3. Over-narrow emphasis on the one Ayanda contract.
    4. Opaque, charged title
Loginwalled links: he links to Wealth Manager magazine, which is loginwalled (but not paywalled).

This is not terribly worthy of criticism if that's the best source for this story. I have not read the Wealth Manager article.

Dead links: he also links directly to objects in an s3 bucket with expired access tokens.

Those are from the UK govt's Companies House. Not too hard to get fresh links.[2][3]

Speculation: Addressing the small balance sheet is of direct relevance. Bundling conjecture is a bit much.

But only a bit. He qualifies this passage well enough.

Narrow focus: This contract is the largest one. Murray claims it is symptomatic of a failing of the UK govt.

So focusing on the one contract doesn't stick out as particularly conspiratorial.

Title: The former criticism (opaque title) is an ok one but the latter (charged title) isn't.

____________________

[1]: https://www.ft.com/content/c0d439c8-18ac-499b-8115-da2bc0f18...

[2a]: https://beta.companieshouse.gov.uk/company/11014884/filing-h... > Total exemption full accounts made up to 31 December 2019 > View PDF

[2b]: https://beta.companieshouse.gov.uk/company/11014884/filing-h... > Total exemption full accounts made up to 31 December 2018 > View PDF

[3a]: https://files.catbox.moe/234hjx.pdf

[3b]: https://files.catbox.moe/c92j8e.pdf

jcahill··on Why time feels so weird in 2020
The date slider section was straightforwardly about 'answering correctly'. If you can ballpark all of these events to the right decade or so, further granularity is unlikely to be related to perceptual distortions about time.

Obvious example: the cassette question.

Question: When were storage devices for audio commonly known as cassette tapes introduced?

Expected answer: 1963.

If "introduced" is intended to be read as shorthand for "sold in a retail market", then this question has no single answer grounded in one person's life experience. I was recalling the year of invention when I answered[1].

Citing "the early 1960s" will be good enough for anyone having a conversation about tech history. The year of the first retail release certainly isn't interesting or worth keeping in your head. Worse, the 1960s weren't even the decade most strongly associated with cassette tapes.

So this is very much a neener-neener type of gotcha question.

[1]: https://files.catbox.moe/h7hcuh.png

jcahill··on Why time feels so weird in 2020
>Why time feels so weird in 2020

It doesn't. Also, the perceptual demos don't work.

jcahill··on XMPP works – 1 July 2020
Bridges were shut down to push proprietary chat spinoffs of social media websites. Ex. Facebook had an XMPP bridge for a couple years but shut it down for this reason.
jcahill··on Analysis of all YouTube trending videos in US for 2019
My web archival group (nonprofit) has international trending data for the same period. I've been meaning to get around to analysis for a while. This article is a good point of reference.
jcahill··on Yoloface-500k: ultra-light real-time face detection model, 500kb
The submission is an implementation of a core task in computer vision. Your response is an appeal to vividness unrelated to the submission except in its recruitment of computer vision to sell the FUD.

If researchers getting good at something is sufficient priming to cause you to direct your imagination toward hyperbolically negative outcomes, the problem on your hands is a constitutional resistance to further progress in the research area.

In that case, challenging readers to produce arguments on the finer details of the narrative you've painted in support of the technopessimism is bad faith rhetoric.

jcahill··on Algorithms Are Now Commodities
Commodities are fully fungible goods and services.

Different implementations of a single algorithm should have the same basic formal properties, by definition.

That is, they should be roughly fungible. Else you aren't dealing with correct implementations of the same algorithm.

If anything, the thesis "algorithms are now commodities" is less true now than ever before.

As the complexity of software deliverables increases, the use of complex, proprietary algorithms in software should likely increase in turn.

jcahill··on Show HN: Thinker App – Automate Your Information Intake

    Thinker is an app that automates your learning. It enables you to:
    - Jump to key info instantly using AI.
    - Stop redundantly checking sites like Hacker News.
    - Permanently archive every article.
How are you archiving the articles?
jcahill··on In defense of hiring based on side projects
The author is essentially hoping to usher in an end-times world divided solely into middle managers and sapient jumping beans. This is a far from sensible goal.

>I don’t mean to pick on this person; there’s no reason to be familiar with the history of the US aerospace and rocketry industry (unless, of course, you’re going to make an argument that hinges on claims about the industry).

Failing to grasp the broadly satirical nature of copypasta like this, then doubling down on an i-know-more-than-you literal interpretation of the text, is not a situation in which the author should be concerned that readers will worry the other guy is getting roasted too hard.

>Barons of the Sky is a nice history of the early days of the aerospace industry, and it shows that in the early days, everyone was a hobbyist.

Acutely pursued hobbies with upside potential become professions! I am shocked. This is my shocked face. :o

>At another fund I’ve heard about, the head of research liked to poll the interns and ask them if they thought the market was efficient. If the answer was yes, he said, they were fired effective immediately.

This example is just an nth-hand anecdote about an empirically incorrect manager on a workplace power trip. Less shark, more dipshit. Parroting the desired response to some manager's arbitrary psychodrama isn't relevant to making money on money for a living. Does the manager need a date to prom, too?

>(As a friend put it a few years ago, when such pop culture references were relevant, “If you know what a ‘Khaleesi’ is, you had time for side projects.”)

Consider the following: humans experience the same number of hours per day.

If you know what a ‘number’ is, you have time to appease me. Reexamine your life to spend more of it appeasing me.

It's a good argument, because humans experience the same number of hours per day. Not enough of them are spent appeasing me!

>In a field with a low marginal cost of scaling results and tournament-like economics (for example: investing, software, writing, some kinds of hardware), there’s a moral component as well. The better you think you are, the more it’s your responsibility to work extra hard, since the results will be distributed so widely.

Just try to say "there is a moral component to scaling IoT toasters" with a straight face among people you respect.

Note: The author reads HN[1] and likely crafted this blurb as bait. He should come away from this thread more insulted by the inanity of his own writing than grateful for the attention.

Promotion of workified mediocrity like the kind on display here is unnecessary in an era that has already caricatured this worldview past the point of no return with such literary works as the Student Athlete copypasta[2]. You do not need to tamper with your inner life in the vague hope that an interviewer, somewhere down the line, will grade you on your hobbies like all code is a take-home test regardless of its genesis.

Stagnant real wages in the author's home country (US) further reduce the blogpost's relevance. As do OECD data[3][4] showing no positive correlation between average hours worked and stereotypical innovation rate at the country-level.

[1]: https://twitter.com/ByrneHobart/status/1278782161559138304

[2]: https://files.catbox.moe/2srmab.png

[3]: https://data.oecd.org/emp/hours-worked.htm

[4]: https://stats.oecd.org/Index.aspx?DataSetCode=ANHRS

jcahill··on Barclays Bank Using Internet Archive as CDN for JavaScript Files?
"customer onboarding journey" sounds altogether twee for a major international bank. Banks are mattresses with insurance policies. Why is there even a journey to be broken?
jcahill··on One Biohacker’s Improbable Bid to Make a DIY Covid-19 Vaccine
The first livestream has been shifted from Sunday to Saturday June 27th at 11AM PT.
jcahill··on SARS-CoV-2 detected in Barcelona sewage water sample from March 2019
They used five assays. All of these were from the WHO in-house assays list.

In theory, their primers shouldn't have specificity issues. In practice, it's more complicated than that.

The CDC 2019-nCoV rRT-PCR assays have a particularly sketchy history. The WHO in-house assays compendium[4] explicitly states that no false positives are expected for these. But there are huge, annoying caveats to this.

(Tedium warning)

Caveat A: Information chaos. This information has become substantially harder to find on CDC's end than it was originally, having been reorganized, replaced, and removed repeatedly at this point.[1] We can expect this confusion to have forced downstream errors on the part of at least some labs attempting to use the protocols.

Caveat B: The CDC's initial primers[3] were widely criticized for having been badly selected. Example.[5]

Caveat C: The CDC's primers have changed. Without the preprint authors providing more specific information, it's impossible to say what they were testing against for this reason. The preprint authors cite the fda.gov replacement pdf[6], which in turn links out to the CDC primers and probes page[3] via [7] > More Resources on the CDC Diagnostic Panel > Research Use Only RT-PCR Primers and Probes. So all we know is that they used CDC primers, but not which revision.

[1]: CDC replaced its original instructions file, which was located at[2], with a "redirect" pdf containing a link to a replacement fda.gov page. You may find the original document by searching "Tobacco Specific Nitrosamines in Smoke". The reason for this is that the title metadata of CDC's PDFs is often wrong, since they borrow from other files while compiling them.

[2]: URL: https://cdc.gov/coronavirus/2019-ncov/downloads/rt-pcr-panel...

Contents: Real-Time RT-PCR Panel for Detection 2019-Novel Coronavirus - Instructions for Use

PDF Title: Tobacco Specific Nitrosamines in Smoke

[3]: URL: https://cdc.gov/coronavirus/2019-ncov/downloads/rt-pcr-panel...

Contents: 2019-Novel Coronavirus (2019-nCoV) Real-time rRT-PCR Panel Primers and Probes

PDF Title: Avian H7 (Eurasian Lineage) Influenza Real-time RT-PCR Panel Primer Probes

[4]: The compendium document that compiles protocols is a PDF.

URL (landing page): https://www.who.int/publications/m/item/molecular-assays-to-...

Section: "2019-nCoV rRT-PCR Diagnostic Panel Results Interpretation Guide"

[5]: https://tomeraltman.net/2020/03/03/technical-problems-COVID-...

[6]: https://www.fda.gov/media/134922/download

[7]: https://www.cdc.gov/coronavirus/2019-nCoV/lab/virus-requests...

jcahill··on SARS-CoV-2 detected in Barcelona sewage water sample from March 2019
Mislabeling is also an option.
jcahill··on Quanta Editors Aren't Sure Which Narrative They Are Supposed to Be Selling
Article title: Quanta Writers and Editors Discuss Trends in Science and Math
jcahill··on Relax for the Same Result (2015)
Article gist: achieve a flow state / allow yourself to engage the default mode network to reduce exertion, perceived or real, on a physical task with low cognitive overhead (non-competitive cycling).
jcahill··on Preliminary Theory of the In-Group Contrarian
>We can now return to the concept of the In-Group Contrarian (hereafter IGC). Anyone who has observed social media platforms, especially Twitter, will know the type. This is the person who, precisely when mimetic snowballing is in progress, attempts to apply the brakes. What’s important is that this figure is not simply an outsider to the group, in which case they could probably be ignored, but instead claims to share the group’s goals, beliefs, etc and merely objects to some aspect of this manifestation of them. The appearance of this figure is as predictable as the mimetic snowballing itself.

This account casts assent and dissent as quasi-deterministic duals of each other. It is overdetermined as a result.

jcahill··on Coronavirus: Dexamethasone proves first life-saving drug?
New day, new chance to read the article.
jcahill··on Coronavirus: Dexamethasone proves first life-saving drug?
Are you ready to read the ref?
jcahill··on Coronavirus: Dexamethasone proves first life-saving drug?
>a simple design that bets on robust immune response doing the job.

That means greater surveillance time, because so little has been established about activity thresholds that might be deemed properly immunizing. Greater reliance on strong immune response than more complex designs means greater uncertainty with respect to each step of the trials, including projecting general population results from challenge trial results.

So the bad news can't come as fast as it might for other designs. So far, the main worry has been poor activity in monkey challenge testing. But that's not enough to call "bad news" yet. https://doi.org/10.1101/2020.05.13.093195

There's no downside to proceeding with manufacturing at full speed provided the financing is secured, yes. That's what I was getting at. "Produced at no-profit" still comes with significant opportunities for contractors and liabilities if they haven't been paid yet. It would be a nightmare for a public-private partnership to fall apart with manufacturing costs left uncovered because a vaccine platform has fallen out of favor. At the scale of multinationals, even issues like brexit having been pseudo-postponed until the end of 2020 right before all of this stuff hit might have something to do with how they're trying to structure delivery timetables. It's hard to say how any particular firm structures multiple deals with billion dollar pricetags.

jcahill··on Coronavirus: Dexamethasone proves first life-saving drug?
Theirs is AZD1222, fka ChAdOx1 nCoV-19. It has goofy delivery dates for multiple markets.

We've had no bad news because it's a simple design that bets on robust immune response doing the job. It's in contractors' interest to push ferocious timetables regardless of whether this vaccine strategy is viable: if it isn't, payouts are still secured prior to evidence of this coming to light, so they don't get stuck holding the bag. If it is, they're getting to market by next flu season.

https://www.precisionvaccinations.com/vaccines/azd1222-sars-...

jcahill··on Coronavirus: Dexamethasone proves first life-saving drug?
I did not dismiss the whole trial.

They were first to this scandal precisely because they are conspiracists. A website with a name like covexit is nothing I would normally link. ice9, who has been a very solid twitter-lit-reviewer on repurposing candidates for quite a while, started the discussion in the anglosphere via primary sources and these tertiary sources. That better sources hadn't covered the topic yet is unfortunate, but time waits for no one and their claims are independently verifiable.

I reviewed covexit's apologia for Raoult, including his absurd response to RECOVERY's HCQ news, before making any post with the link mentioned. The 2020-06-15 paste was imported straight from one of my social accounts. Friends there already have context spread across many months from me re: topics like HCQ and Raoult.

Today's BBC article on dex is spreading like wildfire, exactly like the HCQ meme did. It has great potential to inflict damage. Better sources had to wait.

>I could immediately spot which side your comment was on just by looking at the links.

You presume too much. I work on web archival and synthetic biology. I'm not a fanboy of celebrity doctors/medical research cranks.

Please read [2], understand it, and see if you still have a need to reply.

jcahill··on Coronavirus: Dexamethasone proves first life-saving drug?
The person responsible for the initial error is quoted in today's article. I singled him out in my comment. It takes more than one researcher to ignore a protocol calling for a dosage 3x in excess of that used by any other study, then paper over the fact that this happened.

>Well, no, actually a single doctor MIGHT have, during a single incident.

You'll find that I used the word "likely", indicating a lack of absolute certitude. This is a statistically reasonable statement based on their dismal HCQ results.

Please stick to material critique of the informational content conveyed.

jcahill··on Coronavirus: Dexamethasone proves first life-saving drug?
I clearly said draw no conclusions. I don't do histrionic. I have been following and archiving the pandemic the entire year.
jcahill··on Coronavirus: Dexamethasone proves first life-saving drug?
>What does this have to do with Dexamethasone?

- Shared personnel are hot on the heels of a deeply dubious result and attempted damage control that has involved an outlandish claim of misrepresentation in an interview.

- This has all the right smells so far to expect further questionable research practices.

- A press release was prioritized over a preprint, even though the authors absolutely know that readily-available drugs will be taken up as primary treatment methods as soon as this news meme spreads. They have given the research community zero time to double-check.

- Wildly sensationalist claims made to the press.

>They might have made that mistake.

This wasn't a simple mistake. Look at ice9's exploration on twitter (see [2]). The whole thing's quite damning.

jcahill··on Coronavirus: Dexamethasone proves first life-saving drug?
Here is the relevant file to comb over for today's dexamethasone result:

⁶ https://www.recoverytrial.net/files/recovery_dexamethasone_s...

Relevant passage:

> A total of 2104 patients were randomised to receive dexamethasone 6 mg once per day (either by mouth or by intravenous injection) for ten days and were compared with 4321 patients randomised to usual care alone. Among the patients who received usual care alone, 28-day mortality was highest in those who required ventilation (41%), intermediate in those patients who required oxygen only (25%), and lowest among those who did not require any respiratory intervention (13%).

Dexamethasone reduced deaths by one-third in ventilated patients (rate ratio 0.65 [95% confidence interval 0.48 to 0.88]; p=0.0003) and by one fifth in other patients receiving oxygen only (0.80 [0.67 to 0.96]; p=0.0021). There was no benefit among those patients who did not require respiratory support (1.22 [0.86 to 1.75]; p=0.14).

Based on these results, 1 death would be prevented by treatment of around 8 ventilated patients or around 25 patients requiring oxygen alone.

Given the public health importance of these results, we are now working to publish the full details as soon as possible.

jcahill··on Wuhan hospital traffic, search engine data indicate virus activity in Fall 2019
The graph looks even worse when substituting colloquial 拉肚子 for diarrhea.

https://files.catbox.moe/cum13j.png

jcahill··on Wuhan hospital traffic, search engine data indicate virus activity in Fall 2019
This paper is a trainwreck. Could be a case of scientific fraud.

Gist of this paper:

The authors look for circumstantial evidence of community transmission of COVID-19 within Wuhan for dates substantially prior to the seafood market outbreak. They do this by attempting to triangulate* web search volume data with satellite mapping of Wuhan hospital parking lots.

*suggestively gesture back and forth between

In the parking lots, they're looking for more vehicles than expected based on prior years' imagery.

In the web search trends, they're looking for "disease signals".

Specifically, they're looking at two terms: a symptom of sufficient specificity to be useful for differential diagnosis of COVID-19 vs. influenza ("diarrhea"), and a more generic term representing influenza-like illness ("cough").

Here's the teardown:

1. "We developed a comprehensive list of hospitals in Wuhan (using Google Maps, Wikipedia and PubMed)."

List not provided.

2. Search terms are given in English even though they were actually input in Mandarin.

You can verify this by trying to search the English terms yourself. By process of elimination and the shape of their graphs, the terms had to have been:

"cough" → 咳嗽

"diarrhea" → 腹泻

Live link below. Requires signup. Make sure to select Hubei → Wuhan (湖北 → 武汉).

http://index.baidu.com/v2/main/index.html#/trend/咳嗽?words=咳嗽...

3. "We extracted the relative search volumes of the terms “cough” and “diarrhea” using WebPlotDigitizer, v4.215."

What? The values are provided on tooltip hover. This data is written into the page. You can just pull the real data out of the html.

Based on a cursory glance at WebPlotDigitizer, the vague method quote above appears to be tantamount to saying "we made up the data from screenshots".

4. They're looking at seasonal trends.

This is clearly visible from basic inspection. This is not interesting.

5. Their Relative Search Volume scale doesn't agree with Baidu.

Their numbers are wrong. Just look at Figure 2b's y-axis.

https://files.catbox.moe/v05oss.png

Maxima: It shouldn't go to 600. The actual values peak on Baidu at 711.

Minima: None of their data should be plotting near 0. The minimum volume in their date range is 76.

6. The uptick for "diarrhea" (腹泻) that they claim exists doesn't.

Again, this is Figure 2b. The red line is the problem here.

This appears to be deliberate fabrication. No competent researcher can miss the discrepancy between what this graph depicts and what the source depicts.

咳嗽 ("cough" / blue line on harvard preprint) was following a completely normal-looking seasonal cycle during the period under discussion.

腹泻 ("diarrhea" / red line on harvard preprint) was, again, following a completely normal-looking seasonal cycle, with a slightly elevated baseline.

The red line is mis-plotted to trend up in a sudden surge right where the researchers would need this to occur for their argument to hold any value at all. Then, rather than declining substantially prior to the January 2020 cluster, this plot depicts the trend holding through to 2020.

This information is simply false. That's not what Baidu reports.

Here's a side-by-side of the discrepancy:

https://files.catbox.moe/5s6i7o.png

Here's a superimposed version, for charitable reading of Figure 2b. I've stretched it to fit. Red should overlay green.

https://files.catbox.moe/eubuci.png

7. Screw the sat mapping, frankly.

If you can't be trusted to screenshot a graph without us having to arrive at this point, there's no point even assessing the content of your claims about something orders of magnitude more complex.

The sat mapping doesn't constitute evidence so much as it constitutes having paid for the visuals needed to launder this preprint into mainstream news. The only thing that can currently be done with this information is a data dump for other researchers to use.

8. Proofread your paper. 2019 → 2020 below:

> Both search query terms show a large increase approximately 3 weeks preceding the large spike of confirmed COVID-19 cases in early 2019 (Figure 2, c, purple). There is a large decrease in hospital volume and search query data following the public health lockdown of Wuhan on January 23, 2019 (Figure 2, third annotation).

9. No stats anywhere.

Can't do time series analysis without math. Spinning a yarn / telling a just-so story shouldn't be the level of reporting expected of academics in this context.

10. No attempt to build a keyword list or other reference information of continuing utility.

So the analysis dies with this one paper, essentially. There's nowhere to go from here.

jcahill··on [dead]
Nonfunctional with adblock enabled. Bad content. Self-submission. Lipsum and a thispersondoesnotexist.com persona used for a testimonial on the pricing page: https://files.catbox.moe/9wt8hz.png

Go away.

jcahill··on Webrecorder: Make an interactive copy of any web page that you browse
A pi isn't going to cut it here.
jcahill··on Webrecorder: Make an interactive copy of any web page that you browse
Web archivists are coming at the problem with different requirements, related to fidelity and the systematic mirroring of address structure along with payload content and control information.

By the sound of it, you likely wouldn't want to substantially change your personal filing system to have your page captures conform with those requirements.

← PreviousPage 2 of 4Next →