Y
HN Search
Hacker News Search
new
|
comments
|
top
|
jobs
angelzen
searching PlanetScale…
1.
▲
2.
▲
3.
▲
4.
▲
5.
▲
6.
▲
30 ms
·
271.
▲
by
angelzen
5y ago
Because hospitals operate with minimal spare capacity. Any epidemic wave will overwhelm the system. That being said, hospitals are not overwhelmed. There may be a handful of hot areas where that is marginally true, but generally speaking
272.
▲
by
angelzen
5y ago
I am a concerned citizen and parent. I only have so much time to write a well source comments. I wish people were applying just as much critical carefulness with regard to CDC publications. Where is the CDC longitudinal VE study? Do they ev
273.
▲
by
angelzen
5y ago
Having such an obvious sign of infection is a reason to stay at home and minimize human contact? Something we did for thousands of years...
274.
▲
by
angelzen
5y ago
In term of protection. For starters, if vaccine effectiveness were holding, there would be no need for boosters. A picture from a recent preprint of 600k veterans over 6 months shows troubling VE against infection curves. J&J VE against
275.
▲
by
angelzen
5y ago
Thanks for the correction. I find it strange to believe that 12 months are sufficient to assess the lifelong effects of an immunity system intervention.
276.
▲
by
angelzen
5y ago
A South Korea paper from 2016 argues they are related "sometimes". Note that coronaviruses are RNA viruses. Now, all the epi data we have indicates that the covid mRNA vaccines are effective against covid variants of concern. No
277.
▲
by
angelzen
5y ago
I find it strange to believe that 95 days are sufficient to assess the lifelong effects of an immunity system intervention. "Original Antigenic Sin: How First Exposure Shapes Lifelong Anti–Influenza Virus Immune Responses" https:
278.
▲
by
angelzen
5y ago
I fully agree. To prove long term vaccine effectiveness (original, boosters, variants, sequence thereof) we indeed need a time machine. In other words, we are running a species level grand experiment with unknown outcomes. That biopharma co
279.
▲
by
angelzen
5y ago
Thanks for the links! Note that they are making variant-specific vaccines, not proving that variant-specific vaccines are effective against the target variant, if delivered after the original vaccine. Side note: I hope we can do better than
280.
▲
by
angelzen
5y ago
That is not the question I asked. Furthermore the degree of which the current vaccines work against variants is dependent of the variant itself, please consult the OA. (Also dependent on time, but that's a separate conversation .) I am
281.
▲
by
angelzen
5y ago
Same reason we have Intel Pentium Core2 Duo E8400.
282.
▲
by
angelzen
5y ago
That is a fantastic question. We are still vaccinating & boosting against the original variant. Would "the experts" please demonstrate that vaccines against variants are effective, given that original antigenic sin is a long k
283.
▲
by
angelzen
5y ago
Google has the resources. Is it sensible to expect a corporation larger than the oil giants to behave responsibly with its power?
284.
▲
by
angelzen
5y ago
Such a great example of the tailspin of modern activism. We better come to terms with the reality that modern activism is proselytism for secular causes, framed as moral imperatives. Rapidly morphing into proselytism by the sword: either yo
285.
▲
by
angelzen
5y ago
Thanks for sharing your insights. With all kindness, this sounds like post-UnionPacific. The neo robber barons have won. To go on with the metaphor, not sure the future holds a car to fulfill point-to-point transportation demand, as our dig
286.
▲
by
angelzen
5y ago
Inexorably, the locus of making money moves from using one's skill and ability to build software to skimming rent off enormous capital investments: billions of ad impressions, billions of e-commerce transactions, millions of cloud cpus
287.
▲
by
angelzen
5y ago
We'll soon all be millionaires, nothing to worry about.
288.
▲
by
angelzen
5y ago
Not sure this has much to do with OO vs Functional Hooks, more like reading the docs? OO lifecycle: constructor, render, componendDidMount, render, (render, componentDidUpdate, render)* FH lifecycle: (renderFn, useEffectFn | (useLayoutEffec
289.
▲
by
angelzen
5y ago
This looks great. Give it a year and fat bezels will feel crude primitive tech.
290.
▲
by
angelzen
5y ago
This is an interesting insight in the matter. Tangentially, the phenomenon of scientific pioneer jingoism is a fascinating topic. I grew up with Soviet books. The heroes of modern technology were Popov, Mozhaysky and Tsiolkovsky. Marconi, t
291.
▲
by
angelzen
5y ago
Meanwhile, J&J vaccine effectiveness against infection craters to 3% in 6 months. Pfizer VE follows the same trajectory 2 months behind, with 56% VE against infection at 6 months. This is workplace harassment. https://www.new
292.
▲
by
angelzen
5y ago
The title is misleading. "Some Netflix employees ..." or "About 100 Netflix employees ..." or "Dozens of Netflix employees ..." is a more accurate depiction of the situation. > By noon, the crowd had swelled
293.
▲
by
angelzen
5y ago
Education the population is bringing the best in every student. For every subject, there are some students with a best better than the best of other students. We need to learn to live with it.
294.
▲
by
angelzen
5y ago
The dynamics is crystal clear. On the other hand, there are 4 billion Asians in Asia, and they are not sitting on their hands. We can handicap US education all day long, in the long run it's a sure way to lose the global talent competi
295.
▲
by
angelzen
5y ago
The critique is fair. I suspect the original 'test kids in Kindergarten and fast track the best performing for the next 12 years' has its roots in the Sputnik moment and the need to rapidly create a high performing STEM population
296.
▲
by
angelzen
5y ago
Agreed, "gifted" implies a static frame, which is anathema to learning. Alternatively, "accelerated", implying the student must work to keep up the pace.
297.
▲
by
angelzen
5y ago
https://en.wikipedia.org/wiki/Preference_falsification
298.
▲
by
angelzen
5y ago
1. The subsumption rule is elegant, but doesn't answer the pragmatic question: for what use-cases is a naive structural type system (familiar to a large population of developers) insufficient? Logic theory has type towers, but for 99%
299.
▲
by
angelzen
5y ago
"When is this useful?" I can't comment on specific CUE design decisions, and speaking generally about data languages and not specifically about typing thereof. A canonical use-case for data languages is a cloud deployment con
300.
▲
by
angelzen
5y ago
CUE is really promising and the article is a good first introduction to CUE. After reading it, there are a few things that don't fully click: * How does CUE compare with a standard structural typing system, i.e. Typescript. Specificall
More ›