HNHacker News
TopNewBestAskShowJobs

0-_-0

4,368 karma · joined October 10, 2018

submissionscomments
0-_-0··on Pico-8 API Cheatsheet
I'm sure the word "variant" doesn't suggest official association. "Version" would.
0-_-0··on GitHub Copilot is not infringing copyright
Because someone already checked, and it doesn't:

https://docs.github.com/en/github/copilot/research-recitatio...

Every literally quoted part that could infringe appears at least 10 times in the training data

0-_-0··on Pico-8 API Cheatsheet
The PICO-8 has a Nim-based variant called Nico: https://github.com/ftsf/nico
0-_-0··on I do not agree with Github's use of copyrighted code as training for Copilot
Turing completeness applies to models of computation, not hardware. Otherwise, nothing would be Turing-complete because infinite memory doesn't exist in the real world. Just read the first sentence of what you linked to:

In computability theory, several closely related terms are used to describe the computational power of a computational system (such as an abstract machine or programming language)

0-_-0··on Research recitation: A first look at rote learning in GitHub Copilot suggestions
Surprising how extremely rarely Copilot quotes text verbatim.
0-_-0··on Research recitation: A first look at rote learning in GitHub Copilot suggestions
Which would be extremely slow
0-_-0··on I do not agree with Github's use of copyrighted code as training for Copilot
> The human brain isn't Turing-complete as that would require infinite memory.

This is wrong, this is not what Turing completeness is. It applies to computational models, not hardware.

https://en.wikipedia.org/wiki/Turing_completeness

0-_-0··on What should the CPU usage be of a fully-loaded CPU that has been throttled?
Exactly, why not report both? I for one would want to know both.
0-_-0··on Ivermectin and the odds of hospitalization due to Covid-19
If you're right, there should be a correlation with stidy size and results (I'll check for this tomorrow if I remember). It would still have to be explained why do small studies show a much worse result for HCQ than Ivermectin.
0-_-0··on Ivermectin and the odds of hospitalization due to Covid-19
I see your point that it could be P-hacking (as in https://xkcd.com/882/), but why would Ivermectin be so much more P-hacked than HCQ?
0-_-0··on Myocarditis After Immunization with mRNA Covid-19 Vaccines in US Military
Ha, I have a similar story with strage red spots appearing on my back the day after vacciantion. Turns out it was from sitting on a bench.
0-_-0··on Myocarditis After Immunization with mRNA Covid-19 Vaccines in US Military
The second number should take into account the population and probablility of getting COVID, so it should be:

(probablilty of contracting COVID) * (probablity of heart injury in the military from COVID if someone contracts COVID)

0-_-0··on Ivermectin and the odds of hospitalization due to Covid-19
Isn't your last paragraph describing a randomized controlled trial? The page lists 31 for Ivermectin [0] and (for comparison) 40 for HCQ [1]. I don't find the evidence for HCQ convincing, but Ivermectin seems solid from what I can tell. It's hard to explain 28 out of 31 randomized controlled trials showing a positive result with bias, since you would expect that bias to affect both HCQ and Ivermectin. Why would Ivermectin show much better results than HCQ? What kind of bias would affect Ivermectin much more than HCQ?

[0]: https://ivmmeta.com/#fig_fpr

[1]: https://hcqmeta.com/#fcite_rct (Figure 7, scroll down)

0-_-0··on Copilot regurgitating Quake code, including sweary comments
> couldn't you argue that it's working as it should?

Let's say that it's doing exactly what it was trained to do.

0-_-0··on Copilot regurgitating Quake code, including sweary comments
It's worth keeping in mind that what a neural network like this (just like GPT3) is doing is generating the most probable continuation based on the training dataset. Not the best continuation (whatever that means), simply the most likely one. If the training dataset has mostly bad code, the most likely continuation is likely to be bad as well. I think this is still valuable, you just have to think before accepting a suggestion (just like you have to think before writing code from scratch or copying something from Stack Overflow).
0-_-0··on Copilot regurgitating Quake code, including sweary comments
This is a very famous function [0] and likely appears multiple times in the training set (Google gives 40 hits for GitHub), which makes it more likely to be memorized by the network.

[0]: https://en.wikipedia.org/wiki/Fast_inverse_square_root#Overv...

0-_-0··on Ivermectin and the odds of hospitalization due to Covid-19
Then you have to explain why are results about Ivermectin more likely to be thrown out than studies about hydroxychloroquine (https://hcqmeta.com/), which has only 73.8% positive studies and shows much smaller improvement. Why are prophylaxis studies more successful than early treatment studies, which are more successful than late treatment studies? Are they more likely to be thrown out? Why are people throwing out studies exactly in a pattern to create a sort of pattern that would indicate a working drug?
0-_-0··on The most precious resource is agency
I vaguely recall a study where they took suicidal people and employed them as lifeguards with great results.
0-_-0··on Windows 11 on Raspberry Pi 4
Now I'm waiting for your YouTube video on it!
0-_-0··on Ivermectin and the odds of hospitalization due to Covid-19
I understand the P-hacking problem, but if that was the source of the positive result in the meta analysis that would mean that scientists are throwing away 95% of their studies on one of the most promising COVID treatments during a pandemic. In this case there's not really a reason to throw away a negative result. Considering the controversy around Ivermectin a study showing a negative result would get a huge amount of citations and attention.

In any case, thank you for trying to use reason in your argument, we need more around this subject.

0-_-0··on Ivermectin and the odds of hospitalization due to Covid-19
The only specific ctiticism is the following sentence, the rest is designed to create fear, uncertainty and doubt:

These websites do not include protocol registration with methods, search strategies, inclusion criteria, quality assessment of the included studies nor the certainty of the evidence of the pooled estimates.

This is false. The ivmmeta.com page says:

Typical meta analyses involve subjective selection criteria, effect extraction rules, and study bias evaluation, which can be used to bias results towards a specific outcome. In order to avoid bias we include all studies and use a pre-specified method to extract results from all studies (we also present results after exclusions).

They also compare prospective and retrospective studies and see no significant difference. You could have at least scrolled through the page before blindly copy-pasting something from Google.

0-_-0··on Ivermectin and the odds of hospitalization due to Covid-19
Meta analysis of 60 Ivermectin studies:

https://ivmmeta.com/

0-_-0··on YT demonetizes Bret Weinstein and Heather Heying channel
Same is true of Bret Weinstein
0-_-0··on GitHub Copilot
Well, it should be possible to crowdsource training a FOSS version, right? There should be a SETI-at-home for training neural networks. I would donate some GPU power for sure.

SETI@home achieved 50 times the computing power of the world's largest supercomputer [0], so it might actually be the only way to train the future GPT4 or GPT5.

[0]: https://en.wikipedia.org/wiki/SETI@home#Statistics

0-_-0··on YT demonetizes Bret Weinstein and Heather Heying channel
A random blog just isn't the standard for medical evidence. Aren't you aware of the actual standard and know what kind of studies, rigor and peer-review is necessary? You need multiple peer-reviewed randomized control trials, at least 30.
0-_-0··on YT demonetizes Bret Weinstein and Heather Heying channel
He lost his job at Evergreen. Now, the ivermectin controversy cost him half his income.
0-_-0··on YT demonetizes Bret Weinstein and Heather Heying channel
> One person claiming that a drug has effectiveness based on their anecdotal experience

How about 31 randomized control trials?

https://c19ivermectin.com/

A nice meta analysis page:

https://ivmmeta.com/

Ivermectin is not the only substance studied:

https://c19early.com/

0-_-0··on YouTube reinstates Right Wing Watch
Should they allow porn? What kinds?
0-_-0··on Just 2% of workers at VMware showed up for the first day of in-person work
A new variant is a negative effect
0-_-0··on Just 2% of workers at VMware showed up for the first day of in-person work
In that case the virus can further evolve to get around vaccine indiced immunity, as Delta already has. Letting a virus spread among a vaccinated population is not exactly wise.
← PreviousPage 11 of 34Next →