dangom

What is the best research article on UI that you'd recommend to someone new to the field?

It would be great if more people would develop at least basic cooking skills. You'd be surprised how many people wouldn't even know what to do with a bag of rice if they were given one.

Conspiracy theories aside, Worebey's Science paper is not unanimously well received by the scientific community, and the data analysis was shown to be partly flawed.

Here's a discussion on inference pointing to references and sources that argue against the results of the paper: https://inference-review.com/article/thunder-out-of-china

It also raises eyebrows the fact that no animal has ever been found to be positive for COVID, in contrast to all of the previous SARS pandemics.

EDIT: thanks everyone for correcting my assertion. No source animal was found.

What is the greatest insight you gained and could share with non-experts from working on this project?

That's an incredible effect size.

Mechanisms hypothesized to be at play:

"These mechanisms—and those underlying the effects of adulthood vaccinations on all-cause dementia risk in general—can be grouped into at least three broad, non-exclusive categories: 1) influenza-specific mechanisms, including mitigation of damage secondary to influenza infection and/or epitopic similarity between influenza proteins and AD pathology; 2) non–influenza-specific training of the innate immune system; and 3) non–influenza-specific changes in adaptive immunity via lymphocyte-mediated cross-reactivity."

Seems like vaccines reduce the effect, according to this study: https://www.mdpi.com/2076-393X/10/5/652

how much time should one dedicate per day? Is this course compatible with working full-time?

drag and drop on the app and they sync.

Great training dataset for copilot.

this should be written on the top of mastodon.social. Also, it's a shame they don't own mastodon.com . Reminded me of http://paulgraham.com/name.html

They come after the incumbent and feel like a copycat. They also solve problems most people don't know they have, and thus are unattractive (at large scale).

What makes you think academia is any different?

Is there any intuition to be learned from the way Stockfish plays? Are there new openings or strategies that it plays that haven't been sufficiently explored by humans?

It's 2022 and it's close to impossible for you and me to own a computer that we control. Even the thought of owning a computer we control is now seen as "extreme". If we did go through hops to have a machine that we control, it'd be practically impossible to use it to communicate with our friends and family, because all of our communications go through third party servers that we do not control.

In exactly which way is Stallman being inflammatory here?

Stallman has a vision that you and I should be typing our comments and playing our games from machines that are under our control, i.e., whereby we can inspect and modify all of the code that is running on them. That is not the case, and all Stallman is saying, is that this is bad for us.

The Free Software Foundation was founded in 1985. From there since Free Software has only lost ground. I don't think Stallman would agree that "lack of pragmatism" is really what's stopping things from changing for better.

This. Being able to walk and discover new work or see a poster that draws your attention is so much more engaging than scrolling through slidedeck titles on a screen.

Are scientific posters still a thing in all conferences you guys attend? I see many communities transitioning to "digital posters", by which they mean PPT slide decks... I sure miss traditional posters, but curious what the future holds for them.

Active is a bad word to describe what I meant. I don't mean to say that a depressed mind is not thinking, just that it may not be able to get itself out of the place it is in without "active" effort (as in therapy - self-therapy or mediated by others). The conundrum is that putting in that effort may not possible because all energy is already spent - as you pointed out. My suggestion is that psychedelics opens space for these thoughts to be actively dealt with in a productive way.

There is also mounting evidence that depression is a symptom and not a single disease, and the symptom may be originating from multiple causes. This view is supported by the extreme between-subject variability seen in imaging studies [1]. That being the case, maybe not all causes would require an "active" approach, just as maybe not all will respond to psychedelics.

[1] https://www.nature.com/articles/s41386-020-00789-3

I have a pet hypothesis that "curing" depression is an active endeavor, instead of a passive one. Just like being healthy in general requires at least a moderate regard to nutrition and exercising. No amount of medicine can replace those.

From anecdotal reports, it seems that mushrooms forces one into an active thought space (like dropping someone of at the gym with a program for them to follow), such that one is more likely to engage with ones own thoughts, and is capable of interpreting them, recasting them, accepting or letting them go.

The idea of "rewiring" seems a bit too far, since rewiring cannot be achieved instantly.

Exactly. Our tooling doesn't even allow us to make clear inferences about neuronal activity yet - all we can image with fMRI as of today is heamodynamics.

The work proposes that because responses to difference visual stimuli show the same haemodynamic spatiotemporal response in memory areas, the actual memories must share a common representation. This is debatable since we know that localized spatiotemporal responses reflect the variability in the vascular tree and blood flow and volume, regardless of the experiment at hand [see 1 for a discussion]. To dare to claim that effects could be neuronal one needs to run a bunch of extra control experiments (vascular reactivity mapping / breathhold hypercapnic challenges, resting-state imaging as a control dataset), none of which were conducted in the proposed work.

This is known to MR physicists, but hasn't clicked in yet within departments of cognitive neuroscience.

[1] https://www.frontiersin.org/articles/10.3389/fnins.2020.5960...

It sounds like sshfs on steroids?

Read about the concept of Leitmotif [1], that some claim was introduced by Wagner but seems to have already been present in the early 17th century. Leitmotif represents the idea that you can imply dramatic progression through music. The idea is that by using short melodies and simple chord progressions you can bind together elements of a story.

Also see [2] for a birds eye view on soundtrack design by major movie makers.

[1] https://wikiless.org/wiki/Leitmotif?lang=en [2] https://www.youtube.com/watch?v=7vfqkvwW2fs

Actually gives a lot of context for someone not familiar with the topic/data. I learned more from the plots than from the prose.

Same principle, different example. We never read "fact X", we read "fact X from source Y".

What if source Y is biased in that its using all facts X to push an anti-bias agenda?

The text is not asking anyone to care about all humans equally.

It is pointing out to a bias in reporting: one can never read "fact X". One can only read "fact X from source Y". Depending on one's alignment with source Y, one is more or less inclined to believe or care about fact X. The text points out examples of multiple Ys and how they choose how to report on X.

Much of what you call "woke" is realizing that human rationality has many biases, and in order to reify what we in the West agreed upon in text back in 1789, we will have to learn what these biases are and learn to actively compensate for them.

Much progress has been done, but much is still left to do.

Of course you care more about your family than anyone else's, and about Ukrainians and Russians more than Iraqis or Yemenis. It's a great starting point to think about that, and what unintended consequences everyone's individual biases can have for the future of our society. This is not an exercise that people go through often enough.

Abstracting away the specifics of the case (religion, sexual orientation), the true debate at hand is whether a human being should have the right to get an exemption from a law based on beliefs (perhaps provided that these beliefs are strongly held, and commonly found within a particular community). Accepting that exemptions to laws are OK based on beliefs leads to an asymmetry between individuals, whereby those with beliefs gain an advantage that is not readily available to all. Here, playing the free speech card makes no sense, since the purpose of that card is to eliminate power asymmetries, not create them.

In my view, the only way to conciliate laws and exemptions is to make exemptions sufficiently costly to eliminate built in advantages to selected groups, whichever they may be. I don't know what a fair way to do so would be. Perhaps an alternative exit that doesn't require an exemption is an even better solution.

In the specifics of this case, couldn't the web designer hire a third party to take care of the job, providing the service as requested, but not executing it themselves?

No one will get funded with that mentality in such a high-risk high-reward area like fusion. But indeed after so many headlines I agree with you that no one gets excited about these news anymore.

The study is a meta-study over 250.000 published scientific articles. Hard to believe that general lifestyle habits would alone explain these results, given that individual peer-reviewed studies are expected to have at least partially accounted for biases and confounds, and general lifestyle habits would probably vary across such immense population.

Right,

> Here we introduce a hybrid in situ–in silico algorithm, called physics-aware training, that applies backpropagation to train controllable physical systems. Just as deep learning realizes computations with deep neural networks made from layers of mathematical functions, our approach allows us to train deep physical neural networks made from layers of controllable physical systems, even when the physical layers lack any mathematical isomorphism to conventional artificial neural network layers.

To my naive understanding, and please someone correct me if I'm wrong, the point is that they are not controlling the parameters that compute the NN forward pass directly (hence "no mathematical isomorphism to conventional NNs"), but "hyper-parameters" that guide the physical system to do so. For example, rotation angles of mirrors, or distance between filters, instead of intensity values of light. This leads to the non-linear transformations happening in situ, while simpler transformations in the backprop are still computed in-silico.

Smokers spend less time in closed environments. They are not always welcome in super-spreader indoor environments, and they spend less time with non-smokers than non-smokers do (most non-smokers do mind the smell of tobacco).

Also, it's not that hard to find another paper with an opposite conclusion: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC7301735/