This is maybe not the place, but we did some apples to apples comparisons between PyMC, Dynesty, and the Julia Turing.jl package.
A little to my surprise, despite being a Julia fan, Turing really outperformed both the Python solutions.
I think JAX should be competitive in raw speed, so it might come down to the maturity of the samplers we used.
I visited the Large binocular telescope just a month or two ago. A very impressive facility, and one can only imagine the image quality if they were captured using both mirrors coherently.
> In the near term, the switch will lead to an approximate 25 percent gain in efficiency; manufacturers can take advantage of this to increase battery life, reduce the size of the battery, or enable a brighter display.
The 25 percent gain in efficiency is achieved by reducing the waste heat from ~ 20% to ~ 0%, so (if heat is the only limiting factor), they should be able to make them 0.2 / ~0 times brighter. That number could be much greater than one.
Interesting. When OLED screens age, or burn in, do they get yellow? That is, less blue? If they simply get darker, this means the red and green subpixels deteriorate as well, which means improving just the blue subpixels doesn't solve the problem.
I personally think the relative sizes of the subpixels already reflect how much they age. Larger ones are presumably larger because the dye ages more quickly, and larger pixels don't have to be as bright per area. So improving the blue dye would allow us to make the blue subpixels somewhat smaller and/or brighter. The current size difference is not very large though.
There aren’t any dyes in oleds. They emit
photons directly at the correct wavelength.
(Ignoring the topic of the article, which explains why that’s an oversimplification.)
If they yellow, it would be due to different colors dimming at different rates or because there is a plastic protective/anti-glare coating, and it yellowed due to UV exposure.
> There aren’t any dyes in oleds. They emit photons directly at the correct wavelength.
I meant the new blue PHOLED material.
> If they yellow, it would be due to different colors dimming at different rates or because there is a plastic protective/anti-glare coating, and it yellowed due to UV exposure.
There’s a bit more to it than that. GMT is further along but its design has many compromises versus TMT. Also, a second ELT in the southern hemisphere is less useful than one in both Hemispheres.
The situation for TMT on Maunakea is definitely tricky, but it’s also a better site than either of GMT’s or EELTs.
I've been told (but can't confirm) that the GMT instruments are farther along, as well. I agree that a northern hemisphere telescope is necessary and I'd like to see both get built. I'll defer to others on whether Mauna Kea is superior to the EELT site since Chile hosts a lot of extremely successful observatories.
I like to think I'm sensitive to the wrongs done to indigenous populations but can't come to grips with the Native Hawaiian's protests. Nevertheless, they have worked and there is no end in sight. Isn't TMT still hedging with a backup site in the Canary Islands?
It looks like NSF has to arrange a shotgun wedding between Carnegie and Caltech the way it did with Caltech and MIT over LIGO.
> like to think I'm sensitive to the wrongs done to indigenous populations but can't come to grips with the Native Hawaiian's protests.
It’s “holy land”. History is paved with mass death and destruction fighting irrationally over specific pieces of land that have nothing more than “societally sentimental” value.
I can understand it better when people have lived in a place, built structures, raised families, and so on. AFAIK, the top of Mauna Kea was not settled that way and in fact, is pretty inhospitable to humans who are not very well prepared to be there. I've never been to Hawaii but understand that rampant development and tourism are big problems on some islands. I could understand knocking down massive hotels and returning that land to nature. The telescopes at least to me are an entirely different situation.
It might be similar to the way there are unclimbed peaks in the Himalaya because the summits are considered holy. For example Kailash is considered holy by several religious groups, and there are pilgrimages to the base of it but there is no record on anyone climbing it. People in Hawaii may have considered the summit holy without building anything up there.
No. You should read "Hawaiian Antiquities" by David Malo, a Native Hawaiian historian who lived during the early Kingdom. He documents what happened on the mountains, and as far as I can recall, he only mentions Mauna Kea to mention an adze quarry that was used. You can also read Liluokalani's book. She mentions sacred mountains, but it's Mauna Loa, not Mauna kea. That's not to say there hasn't been religious significance but that it never stopped things like rock quarrying or hiking, or even the early telescope building.
This idea of the mountain being tabooed for anything but Native Hawaiian religious practices is a recent invention by the Native Hawaiian sovereignty movement in the form of the TMT protesters, as well as a new "sacred" name they cooked up for it "Mauna a Wakea".
That's a good point. Not that it changes my mind about the worth of the telescopes but I can see that the peak could have been left alone by Native Hawaiians for those reasons.
If we're going to be making utilitarian arguments, I'll point out that almost everything you own could be put to better use by someone else.
In this particular case of native Hawai'ians, and there's a long history of mainlanders deciding, without asking them, that their stuff could be put to better use by them.
It's not just that it's a holy site, it's also that Hawai'i was annexed by the US and its population disenfranchised with most of the land increasingly being owned by rich foreigners and foreign investors. The telescope may be benign by comparison but with growing cultural awareness of Hawai'i's status and history, I'd imagine it's more about the symbolism than the immediate harm. Context matters.
Since statehood, Hawaii has been much more democratic than it was during the Kingdom. Under the Kingdom, a large share of the population was made up of migrant workers without much rights. The initial US takeover was illegitimate, but Hawaii later became a US state with much greater rights for the general population.
The TMT site was chosen because archaeological surveying showed that that particular location does not have any ancient artifacts (such as burial sites).
My understanding is that the AO system for GMT is going to pose quite a challenge. A big topic of ongoing research is dealing with “petaling” where the separate primary mirrors in phase due to the atmosphere.
For combining the projects, it does seem like that may be the only option funding wise, but it’s hard to imagine what the resulting observatory would look like. Maybe it would have to be a completely new design?
Thanks for the pointer. I am interested in instrumentation so I'll look up the AO papers. Naively it seems like that would be an issue whether the primary is a few large segments or a lot of small ones.
If the resulting project winds up as a redesign, it would presumably add years to the project and that will only increase costs. Some philanthropic organizations and partner countries will see their generosity go for nothing which won't help matters. EELT will be operational in that time and getting results.
I also wonder about the growing LEO satellite problem. I get that these enormous telescopes have narrow fields-of-view compared to Rubin and similar wide-field survey facilities. The possibility of these being overwhelmed by spaceborne light pollution seems very real.
>I'll defer to others on whether Mauna Kea is superior to the EELT site since Chile hosts a lot of extremely successful observatories.
Mauna Kea is special as a site because it sits in the middle of the Pacific. Their air travels for miles unobstructed before reaching the island, so it flows more smoothly over the mountain. It is also above 40% of the Earth's atmosphere and very dry [1]. This makes for much less distorted viewing than any other telescope site in the world. It is also in the Northern Hemisphere, which has a different sky than the Southern Hemisphere in Chile. The two sites complement each other in this way.
> but can't come to grips with the Native Hawaiian's protests
That isn't surprising. The protest movement, particularly their leaders, are great at manipulating the media. I rarely see balanced articles, or articles telling the real story behind the protests.
The claims of the protesters are ostensibly about the environment and religious sacredness. But most of their claims have been debunked during extensive administrative and legal hearings [2].
What it's really about, particularly for the protest leaders, is Hawaiian sovereignty. They want to restore the monarchy. They are like Hawaii's version of the Sovereign Citizens. But unlike them, the Hawaiian sovereignty movement has its hooks in major institutions, like the University of Hawaii Manoa, the state's flagship campus, where the dean of the Hawaiian Studies school is a pro-sovereignty activist. They are also very good at social media, so many media organizations pick up their media and take it as gospel, without actually investigating their real motivations. This even is true of orgs the Sierra Club, unfortunately.
They are targeting the telescopes because they cannot target the military, a more logical choice with respect to their actual goals, or the tourism industry, because a lot of them make money there. So the telescopes get the knock even though they are the most environmentally and community friendly of the three. They are good at getting attention by bullying astronomy and have even duped people like Jason Mamoa, The Rock, Tulsi Gabbard, and Elizabeth Warren into supporting them. They have all the "woke" college students backing them, even that TMT member institutions (including astronomy students).
The delays in construction that are largely caused by them has caused the project cost to balloon and it's now seeking NSF funding.
Worth mentioning that essentially all professional astronomers / astrophysicists consider this person a grifter.
It’s amazing that he keeps getting so much press though, would love to learn his secret.
He’s exchanging a career’s worth of credibility for media attention and the resulting financial opportunities that result from media attention. It’s as a simple as using an existing platform to say outrageous things in a confident manner frequently enough to stay in the spotlight, many politicians around the world are running this playbook right now. It helps to play the victim when (rightly) criticized, it helps create an other to fight against.
- "What's the probability that a meteor hits an object of cultural significance?" -> vastly lower than the probability that the meteor hits something of non-cultural significance.
- "What's the probability that a meteor hits a given object of cultural significance?" -> Proportional to object size + age.
Maybe the sphinx isn't a bad candidate for the only object of cultural significance impacted by a meteor? Still seems unusually improbable however.
Right. A meteorite strike on the Sphinx is both extremely unlikely considered as an explanation for damage, and also entirely possible given evidence that it occurred.
The same reasoning applies to two UUIDs or Git hashes matching by accident: of negligible probability, yet easily proven if it happens.
Nothing compelling that I know of. Other meteorite strikes being described using the same language might count. A record of having used the particular meteorite in religious apparatus, more so. It is the sort of thing that would have been recorded, but records are spotty.
I think the headdress was supposed to have been covered in gold leaf, which might have attracted lightning strikes. Just cracking from normal weathering and falling off is the most likely cause, though. It seems like a thing that they would have liked to blame on something dramatic like a thunderbolt.
There are a lot of monuments that would feel similar if they got hit (the Taj Mahal, Notre Dame cathedral, the Statue of Liberty, Machu Picchu, Stonehenge, large parts of the Vatican, the Forbidden City, ...), and the odds of any one of them getting hit should be quite a bit higher. The Sphinx is on the smaller end of qualifying monuments, but it had a lot more time to get hit than most. A meteorite striking the site of Notre Dame in 1000 B.C. wouldn't have had the same ... impact.
When you're looking at hindsight, it is the "somewhere" that matters here, so it's not unlikely at all. There are a massive number of interesting things around the world that would be "surprising" to have been hit by a meteorite. If one of them was in the past, this isn't as unlikely as it seems.
Regardless, I realize other posters linked to pretty good evidence that this isn't what happened.
The whole point of this comment thread is that there does seem to be some evidence it happened. The evidence is just being written off here because the odds are low.
I’m using this from Julia and both the user and developer experience is great.
It’s much more limited than publication style plotting libraries but the instant 60fps reactivity is amazing.
Coming from matplotlib, I found Makie such a breath of fresh air. The API is just as (if not more) flexible but way more predictable. Their layout system in particular is amazing. I think it bundles it's own constraint engine?
Congrats on the new website!
PS. Thanks to the Makie team for the shoutout to my corner plot package in the ecosystem section!
To be fair, matplotlib is pretty awful (the API, I mean). I think it might be the worst library I have ever worked with... Though perhaps I'm repressing some memories there.
The best plotting APIs I have seen are those from MATLAB and Mathematica. I have used Makie a few times, and it is very flexible and well designed.
I am hoping they provide a declarative API that emulates Gadfly more closely. Gadfly is great for simple statistics, sort of a Julia equivalent to ggplot2.
Matplotlib is heavily inspired by MATLAB's plotting, no? Its pyplot submodule is a wrapper around the base OO API which is an almost exact clone of matlab.
Matlab and Mathematica are awesome. Gadfly had a great API but suffered greatly from Julia's ttfp problem. Tried makie, looked good but that was around when I left Julia (just because other langs paid). VegaLite is pretty amazing.
I'm coming from matplotlib too, though I lack the experience to judge whether its API is any good or not. I just use it. ;-) Much as I love Python, there's always that question lingering in the back of my mind, as to what comes after Python. A compelling plot library would be a major factor.
As has been pointed out elsewhere, this is the first image of our galaxy in something other than light (radio, infrared, x rays, gamma rays are all photons).
A little to my surprise, despite being a Julia fan, Turing really outperformed both the Python solutions. I think JAX should be competitive in raw speed, so it might come down to the maturity of the samplers we used.