Comment Re:Hyperloop ... (Score 1) 142
Prufrock II runs at about a mile/week, which is 230m/day. If China has faster TBMs, would love to see some citations. Industry average is 20-50m/day.
Prufrock II runs at about a mile/week, which is 230m/day. If China has faster TBMs, would love to see some citations. Industry average is 20-50m/day.
Both the Romans and the Incas were around 1000-2000 years ago, so they haven’t been gone for very long. The evidence left behind is already fairly heavily decayed and quite a lot is missing (ex: households of “common” Incan villages, not just the religious and ruling artifacts). For early human civilizations 10,000-15,000 years ago, there is really nothing except some fossils and primitive tools. In 2.5 million years, I think it is safe to say it would be hard to tell if there was a civilization here on Earth if we all disappeared tomorrow.
I live in Houston. Nobody that works in my office is a rancher, yet 50% of them drive trucks.
I bet you quite a lot of them use their trucks to do things that are impractical/impossible to do with a passenger sedan. Just because you live a lifestyle where you can’t see a need for a truck doesn’t mean everybody else is only buying one for vanity.
Some of it is culture, though, yes, but that is everywhere. Go to Alaska sometime and see how many people drive around Subaru Outbacks. One would think it was the official state car. Is it because they handle better in the snow than anything else, or that everybody goes mountaineering once a week? No, not really. They’re just popular cars. Are we really supposed to declare Californian-style cars as the only acceptable cars for everybody because the car culture everywhere else is wrong by somebody’s random opinion?
Nostalgia has a bit to do with it too. In my area, when I was growing up, winter was predictably cold (10-20F average with some days 10F and maybe even 0F). There would always be snow in December and some years in November. Within the last ten years we are closer to a 20-30F average winter temperature and no accumulated snow until January (this year no snow and itâ(TM)s almost February). So if you are person that enjoys the winter (snow, skiing, sledding), you probably miss that.
And thereâ(TM)s also that cities have built houses and other infrastructure in anticipation of historical repeated weather patterns (temperature, precipitation, wind, etc). Having those patterns change drastically can create a lot of problems where existing houses and streets and other infrastructure arenâ(TM)t built to withstand new prevailing weather trends.
All in all itâ(TM)s not so much that we canâ(TM)t adapt, just that it is disruptive and people as a general rule donâ(TM)t like change.
I didn’t watch the video, so there may be some trick to his calculation that I’m missing, but I call bs on those numbers. According to this,
https://www.irena.org/publicat...
which is the most recent data I could find (2021), grid-scale solar PV costs about $48/MWh generated. In the US, current nuclear plants are operating at a cost of $30/MWh, but even if we use Wikipedia’s high end,
https://en.m.wikipedia.org/wik...,
nuclear costs $136/MWh. So at 4x the generating capacity you have already exceeded the cost of nuclear by a lot, and this is without the battery storage. Also consider that a solar farm uses about 2x the land area as an equivalent nuclear plant (without the batteries), so if you’re installing 4x the generating capacity, you need about 10x the land. Where are we going to put it? And if you say the ocean, your costs for everything have now doubled.
From a system building perspective I won't even start to list the benefits, but to give you one: windows still doesn't support symbolic links through CIFS, it's network FS protocol.
Not sure what you mean by “support symbolic links through CIFS”, but NTFS has supported symlinks (aka reparse points) since v3.0 (Windows 2000 era) and the SMB protocol since v2.0 (Windows Vista era). Agree that SMB matured a bit late relative to NFS, and symlinks on NTFS are limited in the ways they can be used, but I wouldn’t discount SMB as a good general purpose network file system that is really the only viable alternative to NFS these days. Now, especially with v2.2+ (aka v3), SMB is a very capable protocol that is better in some ways than NFSv4.1.
No, ~98% of the genome is non-coding, but not all non-coding DNA is “junk DNA”, which as you note is a highly controversial term. Also, the exome is not concentrated in one spot on one chromosome, so even if they had decided the exome was the only important part to sequence, they couldn’t just sequence the exome. They had to sequence most of each chromosome to get all of the coding sequences on each chromosome.
You are both correct. When they announced the “essentially complete” human genome, they had sequenced 99% of what they had intended to sequence. It was not the complete genome, but it included what was then known to be the whole exome (all protein coding sequences). The missing heterochromatic regions were initially estimated to be only 5% of the genome, and the technology available at the time couldn’t sequence it. When long read techniques became available and they were able to sequence some of the gaps, they refined the estimate to about 8% of the genome and discovered some additional coding sequences as well. Completely sequencing all of the gaps was a monumental effort and was only officially announced this year, albeit some of the sequence is “low quality”, and we still don’t have a complete Y chromosome.
While it is true that they considered the heterochromatic regions “uninteresting”, the main reason they didn’t sequence them is because they couldn’t at the time. Sanger reads weren’t long enough, the polymerases being used couldn’t handle long and repetitive A-T stretches, and the BACs they used for ordering contigs would undergo recombination during cloning which caused a lot of the sequence to go missing or become inverted. So it was decided that the euchromatic regions were good enough for the project to be considered a success.
On the other hand, sometimes the regular expression that does what you need it to do is hard to look at.
Yeah, regexes are what got me into Perl back in the day. They can be a bit inscrutable, but they are just so powerful, and you can do some crazy things, like negative lookbehinds and named backreferences (https://perldoc.perl.org/perlre). Like you said somewhere else, you can do regexes in Python too, but Perl regexes are faster and just seem more natural to me when writing them.
Too bad Raku kind of fizzled out because they really kicked it up a notch with grammars (among many other useful things),
https://en.m.wikibooks.org/wik...
The training strategy for AlphaStar went through a few iterations. When using their population-based “AlphaStar League” approach they showed that the AI did in fact use a lot of macro strategy, and it’s micro optimizations did not exceed those of professional human players.
In its games against TLO and MaNa, AlphaStar had an average APM of around 280, significantly lower than the professional players, although its actions may be more precise. This lower APM is, in part, because AlphaStar starts its training using replays and thus mimics the way humans play the game. Additionally, AlphaStar reacts with a delay between observation and action of 350ms on average.
https://deepmind.com/blog/arti...
I think it’s quite fascinating, actually.
Selection affects the variants of a gene, the so called allele. It affects basically the value that is stored in a gene. Genetic modification adds new genes, which didn't exist before in that lineage at all (like bacteria genes in vertebrate or virus genes in plants or whatever).
First of all in plants it would be better to refer to selection of traits rather than genes. Cross-breeding, hybridization, and back crossing are fairly course tools for genomic modification. About the most specific you could hope to get is 1 or 2 linkage groups spanning a few dozen kb of a chromosome. So when you are selecting and propagating varieties and cultivars, you are NOT selecting random background mutations within a single species, for the most part. There is some, but most of the genetic variation will be due to homologous recombination during sexual reproduction, which is how you get a desirable trait to transfer from one subspecies to another (or any species that can cross pollinate with it). Very often the genetic differences between two cultivars with different sets of traits is very large. See for example,
https://ainfo.cnptia.embrapa.b...
Second, there are some astonishing similarities between core metabolic genes in plants and bacteria. Cross-kingdom horizontal gene transfer hasn’t been widely appreciated, but it does happen and is evolutionarily significant.
It’s called an allene,
https://en.m.wikipedia.org/wik...
They are not very common in “ordinary” bio molecules, but they are found in many natural products. For example,
Microladallene - https://www.chemistryworld.com...
Allene biosynthesis is another interesting topic. The enzymes that do these reactions are pretty wild,
https://advances.sciencemag.or...
Well, to be fair, what he is saying in his post is
“Powerwall will interface only between utility meter and house main breaker panel, enabling super simple install and seamless whole house backup during utility dropouts."
which implies that you don’t need Tesla solar panels, but the solar panels you do get have to be connected directly to the Powerwall. So if you already have an inverter setup, that will no longer work. Easy for an electrician to fix though.
Uh no. The goal post was set by Tesla with the public claim that it was impossible.
This is at least the second time you have stated this (and been modded to +5) with no evidence, and your friend jeff4747 has also done so below. The only thing Tesla has ever claimed is that there are safety measures in place to prevent it, not that it is impossible. The fact that someone can disable or trick the safety controls doesn't make Tesla the irresponsible party.
With regards to the crash in question, Elon specifically said that Autopilot was not engaged as evidenced by the car's logs. He did not assert that it was "impossible". Separately he said the car will not allow Autopilot to be engaged on an unmarked road. That is a true statement. If Autopilot is engaged on a marked road it will continue to drive when it hits a patch of unmarked road, but if it is not already engaged it cannot be engaged on an unmarked road.
The actual tweet is here for your reference,
https://twitter.com/elonmusk/s...
Soon is a bit relative. The question is will Tesla scale fast enough to outpace this change in the market. A legitimate question, but the article doesn’t make any attempt to address it. Tesla has been doubted many times before and has made it through, so I predict they’ll be fine. As of now, only GM appears to be serious enough to build their own Gigafactory, and it’s not online yet.
The use of money is all the advantage there is to having money. -- B. Franklin
