criticaltinker

9 ms·

31.
by criticaltinker 5y ago
Also discussed a couple days ago: https://news.ycombinator.com/item?id=29795910
32.
by criticaltinker 5y ago
Also known as the “bullwhip effect”: https://en.m.wikipedia.org/wiki/Bullwhip_effect
33.
by criticaltinker 5y ago
> dispute the observation that mutations accumulate with a certain bias based on the host biology To be clear that is not being disputed. IMO the central thesis of the paper is: > the progenitor of Omicron jumped from humans to mice,
34.
by criticaltinker 5y ago
I’m sure some folks here will chime in with countering/supporting data, but assuming these trends are truly representative, understanding why this is occurring will be imperative. One hypothesis involves the fact that current mRNA va
35.
by criticaltinker 5y ago
> Intermediate versions have gone largely undetected, thus resulting in an exceptionally long branch leading to the most recent common ancestor (MRCA) of Omicron in the phylogenetic tree. We hereafter refer to this long branch as Branch
36.
by criticaltinker 5y ago
This looks really cool, but apparently it relies on a web service to extract pdf metadata: >* The Retrieve Metadata feature uses a Zotero web service to find item metadata. The Zotero client sends the first few pages of text from the PDF
37.
by criticaltinker 5y ago
Yes and this type of doublespeak is coming from the top down in many countries. It’s alarming because “they don’t believe in science” has become a divisive weapon that hand waves away all scientific nuance. The vast majority of young health
38.
by criticaltinker 5y ago
Well put. I would love to read some case studies on orgs that recovered from bad tech debt. Does anything like that even exist? Genuine talent tends to recognize and avoid these types of orgs, and the vicious downward cycle you describe see
39.
by criticaltinker 5y ago
Another plausible alternative is that Telegram embeds an open source NLP model in the app so translation can be performed locally (offline). To be fair we know that mobile hardware resources can be extremely limited, and I’d wager a server
40.
by criticaltinker 5y ago
> The model is considered the greatest computational project in Russia for now, totaling 24,256 GPU days to train the models. > We don’t know for sure why OpenAI hasn’t shown its work in a more reproducible way. But this step is de
41.
by criticaltinker 5y ago
Dupe of ongoing discussion: https://news.ycombinator.com/item?id=29741424
42.
by criticaltinker 5y ago
Yup totally agree - OP is an interesting proposal and has provoked a lot of quality commentary regarding the tradeoffs of clarity vs conciseness, implicit vs explicit, FK performance, etc. But IMO you’ve raised the important long term con
43.
by criticaltinker 5y ago
Databases in 2030: SQL DB finally succumbs to Graph DB as #1 Does anyone else feel like a caveman when modeling a many to many relationship in a normalized schema, and then querying via SQL? I’m surprised graph DBs aren’t more popular f
44.
by criticaltinker 5y ago
DFAs and NFAs may seem like academic novelties but they are amazingly powerful for certain problems. For example, naively searching a string of length N for P different substrings will yield an algorithm that has roughly O(N * P) worst case
45.
by criticaltinker 5y ago
> In summary, the risk of hospital admission or death from myocarditis is greater following COVID-19 infection than following vaccination and remains modest following sequential doses of mRNA vaccine including a third booster dose of BN
46.
by criticaltinker 5y ago
It's a good question which might yield a very complex answer depending on how far down the rabbit hole of reproducible science/computation/machine learning you're willing to go. To keep things simple, I'd say the tr
47.
by criticaltinker 5y ago
Good suggestion, it was tough to narrow down the list! Here is a link to the ViT paper in case others are interested [1]. According to the latest ImageNet standings [2], ViT appears to have slipped to second place in Top-1 Accuracy. CoAtNet
48.
by criticaltinker 5y ago
A couple previous discussions for those interested: https://news.ycombinator.com/item?id=19054501 (Feb 1, 2019) 411 points, 23 comments https://news.ycombinator.com/item?id=23391934 (June 2, 2020) 304 point
49.
by criticaltinker 5y ago
Transformer based architectures and unsupervised pre-training are achieving state of the art results across multiple modalities including NLP, CV, speech recognition, genomics, physics etc - so here's my must read list of recent pape
50.
by criticaltinker 5y ago
Good point about the quantity of active ingredients varying by species. It’s worth noting that GP may have been referring to the physical distribution of active ingredients within the mushroom body. The “cap” of a mushroom almost always c
51.
by criticaltinker 5y ago
I highly recommend taking a look through the Yale Manual for Psilocybin-assisted therapy [1], section 6.9 is particularly relevant to your inquiry. Some excerpts: > We encourage you to take an attitude of curiosity and acceptance toward
52.
by criticaltinker 5y ago
Wow an ounce per week is such a heavy intake, have you noticed any negative impacts for example to your lungs or teeth? Do you prefer smoking over edibles because of the instant rush versus a slow ramp up?
53.
by criticaltinker 5y ago
Not going to bed stoned should be a recommended best practice in my opinion, especially given all the research supporting the fact that sleep is impaired by Cannabis much the same way as alcohol. So many people claim that Cannabis helps the
54.
by criticaltinker 5y ago
Yeah the body quickly develops a tolerance when cannabis is used chronically. In stoner culture people talk about taking “T” breaks for this reason. Other people switch strains to try mitigating the effect. Smoking a gram or two would be en
55.
by criticaltinker 5y ago
> But it’s no silver bullet: your software is not going to crash and burn if you don’t write your tests first, and the presence of tests alone does not mean you won’t have maintainability issues. > It was not our priority early on.
56.
by criticaltinker 5y ago
> Since both hosts were behind the same loadbalancer, it was possible to cache files hosted on redacted-cdn.com under assets.redacted.com, inherently allowing me to move the vulnerable html file on a different domain and achieve xss und
57.
by criticaltinker 5y ago
The clickbait headline belies the content of the article: researchers still don’t know exactly what triggers lightning. Personally I find the leading hypothesis a bit absurd or at least unbelievable. The idea that colliding ice crystals she
58.
by criticaltinker 5y ago
What’s dying is classical feature engineering: complex ad-hoc pipelines created to manipulate and transform data to improve a specific model. Contemporary machine learning instead prefers to create complex model architectures and loss fun
59.
by criticaltinker 5y ago
Check out The Annotated Transformer, it’s one of my favorite references! It contains straightforward python code side by side with excerpts from the original paper. http://nlp.seas.harvard.edu/2018/04/03/atten
60.
by criticaltinker 5y ago
Had to read quite a bit before Conway’s Law was mentioned, but glad it was. Viewing organizations and software engineering through this lens has been a game changer for me personally. As an individual contributor I now appreciate horrendous

More ›