maxander

Meditation retreats sound interesting, but they've always struck me as tending to have a fairly grueling, bizarre schedule- awake and breakfast at ~6AM, dinner at 4pm, lights out at 9PM (and no meal until breakfast the next day.) It strikes me as the worst setup for long meditation sessions- getting everyone sleepy and hungry and out of sorts. That it would send susceptible people off the deep end doesn't sound surprising.

As a sort of "negative control," I wonder if psychosis is as frequently triggered by other taxing programs; for instance, army basic training? There too people are forced to sleep on an artificial schedule, meals are regimented and at potentially unfamiliar times, and most of the day is spent in unfamiliar, effortful activity (but all of which is probably diametrically opposite to meditation.)

NKOS is probably the (or at least one of the) defining works on cellular automata and the thought around/behind them, at least when it was written. It will likely be as interesting as that sounds to you.

There's an entire genre of posts on HN which amount to someone telling a terrible customer support story about some FAANG-y company, with the result (and likely intention) that the attention forces the FAANG to turn around, generously fix the problem, and offer the author profuse apologies and freebies to show how nice they are.

Not only is it sad that the same thing is now being done by a well-known company, but what makes it even worse is that this post probably hasn't gotten enough attention to make Google care.

No, an invisible infrared beam reflecting off of dust or water vapor is still invisible infrared light.

As others have noted, this doesn’t result in a uniform distribution. But what you can do, instead, is generate a uniform distribution over a square centered on the point of interest, and simply re-do points that land outside of the circle. For most software engineers, this is probably simpler to understand and harder to screw up than implementing the Haversine formula. :)

That's not how this sort of analysis is done. They separated people by push-up capacity, and compared the rate of heart disease in the low-capacity group versus the high-capacity group; those rates were 36 "cardiovascular events" for the 948 low-capacity guys (so, 3.7% of them), and 1 for the 155 high-capacity guys (0.6% of them.) I know which group I'd rather be in.

It is true that, with the small number of events, the precise "96% reduction" number shouldn't be taken too seriously- the study notes that the 95% confidence interval is 99% to 64%; essentially, there's a one-in-twenty chance that it's a 99% reduction, or likewise that it's only a 64% reduction. Science never grants absolute uncertainties- but sadly, that part never makes it into the news story.

The issue talked about here is distinct from the larger "reproducibility crisis"; the latter is a result of shoddily designed (or simply fraudulent) experimental work, whereas the issue here is the aggregate effects of the huge amount of computational work that is being done- even when that work is being done correctly and honestly.

Testing a hypothesis against a pre-existing dataset is a valid thing to do, and it is also almost trivially simple (and completely free) for someone with a reasonable computational background. There are researchers who spend a decent portion of their careers performing these analyses. This is all well and good- we want people to spend time analyzing the highly complex data that modern science produces- but we run into problems with statistics.

Suppose an analyst can test a hundred hypotheses per month (this is probably a low estimate.) Each analysis (simplifying slightly!) ends with a significance test, returning a p-value indicating the likelihood that the hypothesis is false. If p < 0.01, the researcher writes up the analysis and sends it off to a journal for publication, since the odds that this result was spurious are literally hundred-to-one. But you see the problem; even if we assume that this researcher tests no valid hypotheses at all over the course of a year, we would expect them to send out one paper per month- and each of these papers would be entirely valid, with no methodological flaws for reviewers to complain about.

In reality, of course, researchers sometimes test true hypotheses, and the rate of true to false computational-analysis papers would depend on the ratio of "true hypotheses that analysis successfully catches" to "false hypothesis that squeak by under the p-value threshold" (i.e., the True Positive rate vs the False Positive rate.) It's hard to guess that this ratio would be, but if AAAS is calling things a "crisis," it's clearly lower than we would like.

But there's a further problem, since the obvious solution- lower the p-value threshold for publication- would lower both the False Positive rate and the True Positive rate. The p-value that gets assigned to the results of an analysis of a true hypothesis are limited by the statistical power (essentially, size and quality) of the dataset being looked at; lower the p-value threshold too much, and analysts simply won't be able to make a sufficiently convincing case for any given true hypothesis. It's not a given that there is a p-value threshold for which the True Positive/False Positive ratio is much better than it is now.

"More data!" is the other commonly proposed solution, since we can safely lower the p-value threshold if we have the data to back up true hypotheses. But even if we can up the experimental throughput so much that we can produce True Positives at p < 0.0001, that simply means that computational researchers can explore more complicated hypotheses, until they're testing thousands or millions of hypotheses per month- and then we have the same problem. In a race between "bench work" and "human creativity plus computer science," I know which I'd bet on.

> We show that independently of age, sex, educational attainment, and socioeconomic status, higher worthwhile ratings are associated with stronger personal relationships (marriage/partnership, contact with friends), broader social engagement (involvement in civic society, cultural activity, volunteering), less loneliness, greater prosperity (wealth, income), better mental and physical health (self-rated health, depressive symptoms, chronic disease), less chronic pain, less disability, greater upper body strength, faster walking, less obesity and central adiposity, more favorable biomarker profiles (C-reactive protein, plasma fibrinogen, white blood cell count, vitamin D, high-density lipoprotein cholesterol), healthier lifestyles (physical activity, fruit and vegetable consumption, sleep quality, not smoking), more time spent in social activities and exercising, and less time spent alone or watching television.

In essence, this amounts to "spending time with friends/family" (said five different ways) and "staying healthy" (said nine different ways.)

HNers may prefer reading about this from OpenAI itself: https://blog.openai.com/better-language-models/

I suppose you could also read about this system from the system’s own output, but that would (for now) be less productive. :)

For what it's worth, I didn't downvote you- but you should realize the gap between what your Wikipedia source says and what you originally asserted. Yes, Wikipedia is right that the existence of dark matter isn't proven, and it's also true that there's no proven connection to the Standard Model; it's also true, as I said, that scientists are pretty sure that it does exist (the alternate theories are widely regarded as long shots.) And if it exists, this novel form of matter would need to be fit into the Standard Model somewhere, since that is intended to describe all forms of matter.

Incidentally, I also think you're getting your history of science confused; the "cosmological constant" term Einstein used to make relativity consistent with his belief in a static, non-expanding universe was a "hack" (he himself wound up admitting it was unjustified.) But that's entirely unrelated to dark matter. Similar modern theories to the cosmological constant have been termed "dark energy," but the use of the word "dark" in both doesn't imply a real connection- it simply means that they each haven't been directly observed.

That's a pretty awesome project. Instead of hacking the ARM, could you just call up a vet to ask if you can have some of their empty reagent bottles, and fill those with your homemade reagents? (As a bonus, if you can get it to work, the vet would probably be interested in knowing the reagent recipe as well.)

There's a movement towards open-source designs for relatively simple biological instrumentation- perhaps this sort of machine will be in range of those efforts someday soon.

The article doesn't really go into the scientific questions remaining, chief among them (IMHO) being the effect of weed on driving ability. There are some studies that suggest it's comparable to being drunk, but the evidence is still weak, and most people I know who do these sorts of things are oblivious or dismissive to it.

Not much more easily than making money off of anything else. The threat of a recession is widely known; for any financial maneuver that could take advantage of it, the "returns if a recession hits" factor is already priced in.

Unless you have a PhD in quantitative finance, it's usually best to take it as axiomatic that you don't have a knowledge-based edge in the stock market, or anything similar.

What percentage of the city would have easy access to tunnel endpoints? It's a pretty difficult last-mile problem.

At least in the short term, a good bet would be funneling stock to retail stores. In lots of neighborhoods, big trucks parked on narrow streets to offload cargo is a big problem.

Dark matter is a stuff that exists, we’re pretty sure, and some explanations for what it is would involve expansions for the standard model; seeing if one of those explanations is right is therefore a (supposedly) promising avenue for post-standard model physics.

Where does it say they sell 40% less volume? Under this scheme, the amount of US ethanol that can get in tariff-free is limited to two-thirds of the total Asian ethanol production. Does the US export that much anyway?

A potential huge impact of this advance is inspiring further research into peptide-based drugs; this is a whole new paradigm with the potential to be comparable or better than the small-molecule drugs we currently focus on, but the poor marketability of injections has kept most pharmaceutical companies from being very interested in them.

> The government expects 2.99 billion journeys to take place over the holiday, an increase of 0.6 percent from the previous year [...]

So assuming most are taking a trip somewhere and a trip back, that's about 1.5 billion people involved in this phenomenon, somehow or another- roughly 1 out of every 5 humans on the planet!

Not even- I don’t know the relevant numbers, but I’d bet there are many more AWS regions than factories making the same kind of chips as the one in the article.

Every once in awhile, one hears about some accident like this one at a single, physical factory, followed by a laundry list of tech companies that are adversely effected. It powerfully demonstrates how awfully centralized microchip production is; individual plants can account for substantial fractions (or even majorities) of entire product categories, which means that individual incidents can have outsized effects. I don't know if that sort of risk has been fully internalized by the tech sector.

For that matter, how specialized is the market for whichever high-purity chemicals were behind this incident? What portion of the semiconductor industry was purchasing from the same supplier? We may not have heard the last of this.

“Flying cars” like this would be great to have! For instance, emergency vehicles that don’t get stuck in traffic could save countless lives.

...We just need appropriate noise regulations, which will almost certainly prevent private citizens from operating them. Especially in cities.

> will tunnels be cheaper?

Per person? It’s hard to imagine a subway being more expensive than flying cars... even with the US’ ridiculous public works costs.

I find it fascinating that the prevalent comment on this are variations of "that can't be right, we still don't understand [things, such as dark energy, etc]." It demonstrates one of the most pernicious biases in science.

Particle accelerators have been the mainstay and principle tool for particle physics, and have produced amazing results. Further, there are obvious questions in fundamental physics that the Standard Model revealed by these experiments does not answer. But, it does not follow these further questions can be answered via accelerators that us humans can construct. The universe has made no promise to us that its mysteries are accessible, least of all through some particular method. We had a particular reason to believe the LHC would reveal new physics (the Higgs boson prediction) and we have no such reason for the proposed future accelerator; our evaluation should change as a result.

My guess would be that astrophysics is a better route to understanding the mysteries of fundamental physics; astronomers can measure distant phenomena with surprising accuracy already, and objects such as black holes are one of the few cases where exotic phenomena are currently widely expected to occur. Perhaps the money is better spent on a truly gigantic space telescope?

There's plenty of noble points that can be made about not playing God and handling the essence of life responsibly, and so on, but I think there's also some risk of losing sight of the fact that we're not just doing this for fun- we're doing it because people are literally dying and we're trying to stop it. If you have a parent or a child of yours that gets cancer, you're not going to be terribly concerned that the cure was created using CRISPR, or that there isn't 100% certainty about what all the side-effects could be.

The fact is that there is not a single medical treatment available that is fully understood; we regularly learn new things about aspirin, which is one of the most studied medications ever, and which has been in use in some form or another since a medieval wise woman chewed on some willow bark. Applying the sort of standards you're trying to hold genetic engineering to would take away every tool in the medical repertoire, aside from perhaps the bone-saw.

People driving taxis and Ubers certainly do have reason to be vetted at a higher standard than regular drivers, since for the most part they're spending much more time driving than anyone else; this leads to a much higher risk-per-driver if one is unsafe.

This is comparable to how there are special licensing requirements, above and beyond the typical driver's license, in order to drive a truck or a bus (and, IIRC, even higher requirements to drive a schoolbus.)

No mention of the “ghost city” effect? [0] A good hypothesis for why there’s so much unoccupied housing is that it’s all in regions where the demand is slim, that were nonetheless highly developed due to various political factors.

[0] https://en.m.wikipedia.org/wiki/Under-occupied_developments_...

It’s worth noting that all reasons to go vegan, if you don’t find them sufficiently convincing to actually become vegan, are also good reasons to simply eat less meat and other animal products.

I think if “he left surveilled areas for no registered reason” is something that law enforcement observes and considers evidence of criminal activity, we’ll have well and truly lost any semblance of being a free society.

It looks like each of those little bead thingies is a transistor; each has three wires leasing up to it with a little loop at the end. Presumably, a charge from the loop to the side can flip the bead, and this makes it conduct.

It’s surprising that that works, to say the least. :)

“Supporting 32-bit time system calls after 2038” will serve future generations as an example of the lengths that developers are supposed to go to in order to ensure backwards compatibility, even in the face of apparent absurdity.

Alternately, it will serve as a joke about how scared everyone was of breaking things written with our primitive and brittle 21st century programming techniques.