If you want to teach a practical language that your grads can immediately use in practice, and want it to also work with Java: teach Kotlin. The (Android) mobile space has gone this way, and Kotlin works (more or less) seamlessly with Java, so it's becoming more and more popular in the enterprise as well. I think it will eventually supplant Java as the most popular JVM language, beating out competitors like the more academic Scala.
If you want to teach a practical language that your grads can immediately use in practice, and don't care about it being on the JVM: teach Python. There's many good reasons it's becoming more and more popular as a first programming language. As an added bonus, you can later teach them C and have Python and C work together.
If you care more about concepts than immediate practicality, then I'd suggest Julia. It's a practical and high performance language with some key language features you won't find in other languages outside of Lisp. It also works great with C. That said, there's not too many Julia jobs around, though I think it'll grow in popularity over time, and it's actually pretty similar to other languages, so it'd be easy for a student who knows Julia to pick up Java, C, Python, etc.
I program in Java professionally, and I have to say you really did a good job explaining what's good about Java in your post. The multi-threading thing in particular is a huge deal, as I haven't found another language that truly excels at multi-threaded development like Java does, at least with the same quality of life as Java provides (languages like Python and Ruby provide better development quality of life in my opinion, but without the excellent multi-threading support...or a type system, and many other things that would take a while to get into).
I also feel like recent improvements to the Java language (basically Java 7 & 8) may give it a new lease on life, though I think the long term picture for Java is pretty grim, because it's held back by historical baggage and Oracle. In particular, even if Java doesn't lose ground to newer non-JVM languages like Python, then Java will lose ground over time to Kotlin and other JVM-based languages like Scala or Groovy. These newer JVM languages simply don't have the historical baggage of Java (and if you love Java but haven't tried out Kotlin you really should).
Also, another problem for all of these JVM-based languages is that Oracle's recent licensing of the JVM is troubling. That said, it can survive this thanks to OpenJDK and other open source JVM initiatives, but we may be in for some rocky times in Java-land.
I’ve found Java quite effective for writing short programs of less than 100 lines to explore mathematical ideas, like how many polygons can meet at a snugly at a point(see http://math.ucr.edu/home/baez/...). So Java does not need massive projects to be really useful.
I recently discovered and learned Julia, and I think it's the ideal language for these kinds of math exploration problems. It combines the raw firepower of languages like C and a developer quality of life that exceeds even the best languages I've seen previously. It's very popular with the High Performance Computing crowd. It has high level features that no other popular language has, such as multiple dispatch, though it manages to be a very practical language at the same time. It also has some surprisingly nice libraries in certain areas, like their graph theory library. I think it has a very bright future ahead of it. If you haven't tried it, you really should (or if you tried it a few years ago, you should give it another spin now that it's at version 1.0).
That said, even Julia isn't perfect (yet?). Right now the biggest problem is that while it has excellent HPC support, it's still lagging behind on its multi-threading support. Basically, it currently assumes that you are fine with running your additional threads inside other processes (possibly on other machines), and Julia makes this very easy for the developer to accomplish, though it's not quite the same as having cheap in-process threads like in Java. Another issue, specific to enterprise development, is that the ecosystem is still evolving VERY rapidly, even though the language has finally settled into version 1.0, so libraries shift under your feet all the time (though fortunately, they have a top notch package management system to handle library versioning). Oh, also start up times can be pretty bad due to how compilation works (basically, you can pre-compile code, but it's easy to end up having to compile at least some of the program on startup rather than ahead of time), but there are solutions to this most of the time and it buys a lot of positives (though it can be especially annoying when a library didn't optimize their startup times).
I'm also somewhat concerned about the use of garbage collection in Julia. Some recent results make me think that automatic reference counting may be the future in this area. While this concern applies to Java as well (there's a reason Java programs are so memory hungry!), it's compounded in Julia because it complicates the bridging of C and Julia, which would otherwise be seam
He's talking about his personal experience being inside greenhouses kept at 1500ppm, not how plants react to it.
Your source only suggests that action be taken in a domestic setting at 1000ppm to avoid higher levels, and that it causes slight drowsiness starting around 1000ppm. This hardly constitutes a serious danger to health.
That said, a world in which just going outside causes mild sick building syndrome is pretty messed up.
I'm not sure why they selected that snippet of text as their prime example when the made up story about Brexit and the continued prose from Pride and Prejudice from the included video were both more impressive.
That said, I don't see why they think it's so dangerous that they need to keep it secret. People already know that everything that not everything they hear on the Internet is true (or if they do, they're already too far gone!).
Here's an interesting option for controlling cyber-weapons without taking them entirely off the table. Instead of banning them or allowing unlimited secrecy, instead the following rules have to be followed: 1. The cyber-weapon has to be completely declassified within 1 year of becoming operational. (Perhaps a somewhat longer time could be mandated, such as 3 years or 5 years, but if the countdown becomes too long then the situation becomes more and more like unlimited secrecy) 2. The cyber-weapon has to be declared when it becomes operational, so we know when to start the declassification countdown. 3. The cyber-weapon cannot be used against the populace of the country operating the cyber-weapon. If this is the case, the exploits involved have to be reported to vendors immediately, and it has to be declassified more quickly as the vendors fix the issue. (What constitutes being usable against the populace is an interesting question, as stricter interpretations of this may rule out cyber-weapons usable against any public software, also note that private/secret forks of public software used by specific countries for country-specific purposes would almost certainly count as country-specific).
The overall effect of this should be that cyber-weapons are short lived and limited in scope (mainly attacking the secret capabilities of other countries instead of public software/infrastructure). It incentivizes improvement of existing nationally-used public software by defence actors, as they can no longer exploit loopholes in the software used by their own nation. It also incentivizes other countries to use public software for their infrastructure, and increases the quality of said infrastructure dramatically as everyone would want said infrastructure to be of top notch quality. The relatively quick declassification time means that any scandalous abuse of the system can be detected quickly (such as if they ignored rule 3, or if they created a cyber-weapon that was brutal enough to cause war crimes). The cyber-weapons declarations also serve as a deterrent, indicating that such weapons exist without giving away details about who is targeted or what it's for (at least until it is declassified, at which point there should be new weapons in existence). If it ever got to the point where it'd be impossible to create a new weapon before the old ones expired, then there wouldn't be many vulnerabilities out there and so we'd live in a very safe cyber-environment, making cyber-warfare moot.
If there were reasons to classify some cyber-weapons for longer periods, then I would recommend that they at least be required have an accurate summary of their purpose and reason for the classification extension declassified after the normal period, and they should be subject to substantial court scrutiny, with an ultimate declassification required at some later date. If this is allowed at all, it should be rare.
As for your other two suggestions, I definitely agree that software used in weapon systems is important to keep classified. However, I strongly disagree that criminal justice software should be secret. The benefits of public review of criminal justice software outweigh the possibility that some genius could find an exploit that makes them harder to bring to justice. Also, such exploits are more likely to be detected in the first place.
It should also be noted that the complete "source code" of the law itself is already out in the public view, yet we don't worry about someone finding an exploit in the law, even though it happens from time to time, allowing some people to exploit the system. Clearly, having a transparent code of law is much more important than catching every criminal.
Trucking it in via tiny bottles is very silly, but I drink mostly "bottled" water.
Where I live the water is extremely gross to drink, even though I'm fairly sure it's safe to drink. We use our tap water freely in any situations where taste doesn't matter, like cooking (the taste of the water is overwhelmed by the taste of the food), showers, washing things, etc. It's so bad tasting that I am willing to pay extra to get less terrible tasting water.
However, individually packaged bottled water is still extremely stupid even in this situation. Why? Because we have several 5 gallon jugs that we cycle through, and get them refilled at the grocery store. The water out of their machine is incredibly tasty and only costs $0.39/gallon. Even at a gallon per person per day, it's only $11.86/month per person. Although this is outrageously expensive compared to tap water ($0.01/gallon locally, or $0.30/month per person), this premium is definitely worth it to me. It's definitely much better cost-wise than bottled water, which in disposable 1 gallon jugs is around $1/gallon ($30/month per person) and in disposable bottles it's around $6.30/gallon ($189/month per person).
You don't need an expensive dispenser machine to get started doing this, I just bought a manual water pump that attaches to the top of the jug and pumps the water out with a few presses of my hand. It's inexpensive, portable, and requires minimal maintenance. I also bought proper reusable dew caps for the jugs, instead of the expendable ones they try to sell you in stores, so the jugs and caps basically last forever, only needing to be washed out every once in a while.
It's also relatively good for the environment. Unlike the disposable options, it doesn't produce a ton of plastic. I'm pretty sure the grocery store is able to produce the water very cheaply and so these machines are pretty much pure profit for them, hence why they still have them instead of trying to get you to buy their even more expensive bottled water options. I don't know exactly what kind of filters they use (I think it's based on reverse osmosis), but the disposable parts are probably based on carbon.
I looked into running my own filtration machine (not the little filters that don't really do anything, but the real deal), and it's even more cost effective, at least a quarter the cost of buying it from the store, even at the small scale I would have been operating it at. I bought a high quality gravity-based machine (good for camping), and the problem with it is that the water tastes terrible using that option as well (I think it's removing too much, so there's no remaining tasty minerals). If I owned a house in an area with crappy water and had higher water demand, I'd probably install a full-scale reverse osmosis filter, just like the store has. As long as I'm renting though, the extra cost of buying the water at the store and the inconvenience of lugging it home is just fine.
One last side note: it is nice having an emergency water supply. I usually have about 4-8 days of good quality water on hand. If there were ever a serious outage, we would hardly be impacted by the situation. We wouldn't need to start boiling our water, we would just avoid the use of tap water for a while.
tl;dr Don't buy individually bottled water! There's so many better options, even if you want bottled water.
If you assume that you need to cover all current demand for electricity (and more due to growth), then you're massively over-complicating things. Renewables can often be used directly. Renewables and fossil fuels lead to very different usage patterns as well.
For instance, instead of installing excess solar capacity for heating we can get most of our heating from direct solar flux, and use insulation and building materials that absorb heat to make the most of the available energy. For cooling, we can vent hot air as it conveniently separates from cooler air automatically. Obviously, this doesn't cover 100% of our energy needs, since we may be run out of solar energy after many cold, cloudy days, and these problems get worse at latitudes that get less sunlight. That said, the energy required to cover this gap is far lower than replicating our current fossil fuel system using renewables. All the energy generated through direct use is energy that doesn't have to be generated through solar panels and wind turbines, and direct use definitely doesn't require any rare earth metals.
All high energy household applications can be replaced with direct solar when it's available. Heating water with solar is well-trodden territory. Dryers can be replaced by hanging clothes on clotheslines. The sun can light the indoors during the day, and with modern LED lighting the remaining time doesn't really require that much energy. If energy really needs to be cut down, then solar cooking can take over for gas and electric, but at this point we're hitting diminishing returns.
Of course, this isn't quite as helpful in extremely northern climes, and we'll need to stop building such terribly inefficient homes and living such terribly inefficient lifestyles. However, most people live close enough to the equator for this to cover most of our energy needs, and the colder countries can import solar, use geothermal and hydropower, or build local wind and nuclear to cover the difference. They also still benefit, just less so.
In terms of industry, there's a lot more industries that are harder to make renewable. Insanely high temperatures, electricity being consumed directly, warm up and cool down cycles and chemical reactions that take a long time, carefully controlled environments, high labor and capital costs that need to be made the most of, etc. That said, industry can adapt to renewables (ideally using energy sources directly), and at least in manufacturing the finished products themselves are a form of energy storage (the energy you used to produce them doesn't need to be spent when energy is scarce).
For instance, why run a factory 24/7 if you only have energy 12/7? Running 24/7 makes sense in a fossil fuel world because you get more use out of your capital investments, and there's no need to warm-up and cool-down the factory if it's running continuously, but it's not the end of the world to have to do these things in most industries. This situation will continue to improve as computers and software improves, as we can use computer control to cope better with the complexities of manufacturing using intermittent power (with buffers and forecasts). We can also turn off many factories completely for several days in very low energy situations, such as multiple cloudy days in a row, so we can ensure that energy is available to homes.
Of course, much like homes this doesn't cover every possible industry. Some domain-specific high energy processes cannot be reasonably halted or require extremely precise environmental control over a long time, or the costs of halting may be extremely high or safety critical for a variety of reasons. These specific cases require a lot less energy to cover than all industry, and in some cases they may want to shut down for the winter if they're especially likely to run into issues. For critical industries that need to be 24/7/365, they can use base load or stored energy.
As we move more towards service and information oriented industries, the situation gets even better, as peopl
Unfortunately, the problems you talk about are just reflections of the wider society (and technology's impact upon it).
I recently ran across some websites on sustainable living, and I realized that we're doing it all wrong. As a society, we're not running at 100% efficiency, or even 50%, but likely less than 1% efficiency. Considering our current inefficiency, we could all be living a better-than-modern lifestyle with relatively little work (no more than 20-30 hours/week) and enjoy said work far more, all while minimizing our impact on the planet. It was truly an eye opener for me.
Don't get me wrong, the sustainability community has some half-baked ideas too, but after hearing some of them I can never see the world the same way again.
As a concrete example, modern western greenhouses are designed incredibly stupidly. They're impossible to insulate and lose their heat quickly at night, so they have to be propped up by burning fossil fuels to warm the greenhouse. This design only makes sense if labor is expensive and fossil fuel energy is cheap (even with cheap solar, it'd be better to utilize the solar directly). It also makes growing greenhouse crops incredibly risky, since by the time you're ready to harvest, you've already spent a ton of electricity on each crop and could lose it all if the crop fails due to a disease or some other problem. The better solution is to build a brick wall facing south and build the greenhouse on the south face of the wall. Over the day, the wall will soak up energy from the sun and radiate it back out at night. Brick walls are relatively cheap and simple and last a near eternity, and the wall doubles as protection against cold winds from the north and a serious insulating layer, so it makes an ideal energy storage system. Lastly, you can throw an insulating blanket over the greenhouse during the night to keep the warmth in even more effectively. At temperate latitudes, this creates an environment where you can grow temperature-sensitive crops all-year round without any heating, even in the depths of a midwestern winter (and at higher latitudes, it will drastically reduce the amount of heating required to sustain livable temperatures).
At first this might seem like a bit of trivia, perhaps only applicable to farmers. However, I don't see any reason why this wouldn't be applicable in a suburban setting. A greenhouse of this design about the same size as a typical midwestern suburban home can consistently feed dozens of people year-round. They are very popular in China, likely because they need the higher efficiency to feed their massive population and much of their population lives at latitudes where this design works ideally. In a sprawling US city, each home could have a relatively small greenhouse that would be able to easily feed the family living there and produce some excess for sale or storage, maintained by the family living there. Even in denser situations, a single properly-situated house-sized communal greenhouse could feed a dozen apartments year round (or the suburban ring could produce more food per family for sale to the city center). Setting this up and maintaining it isn't rocket science, and yet I have not seen such a greenhouse in the midwestern US, despite it being the perfect location.
Why do we not do this as individuals? I did a cost-benefit estimate, and a house-sized greenhouse would produce $60,000-120,000 in produce per year with an up-front investment of $30,000 and minimal maintenance costs thereafter (maybe $5,000 a year and likely less). In a suburban setting, there would be minimal transport costs, especially if food was transported to the suburb it was being grown in (perhaps being sold to neighbors). The labor costs are hard to figure out, but growing the produce would likely be no more than full time work for one person. $55,000 is a great wage for gardening/farming, and I do think the higher profit figures are more realistic. This work would also be good exercise.
I don't currently own a house, so I'm now looking fo
Let's take videogames, for example. No one has really figured out how to combine a popular consumer-goods-type product like that with the philosophy of free software. The common advice when asked how to make a living writing free software of "provide a service for hire to support your product" only works with some very narrow types of products, and never really with consumer-level products. So, essentially, videogames simply don't exist in Stallman's universe.
Why not use a Kickstarter or Patreon style funding model?
I know it's harder to extract money out of consumers after you're done building the project, but perhaps extracting the maximum profit out of consumers shouldn't be the point.
The "catch" is energy efficiency, running CAES as an isolated unit gives 40-52% electrical-to-electrical efficiency, which is pretty terrible. This is because (for high pressures) you create heat on compression and cold on decompression, and unless you can store the heat at compression and return it for decompression, you'll have to provide external cooling and heat (which brings down the efficiency to the figure given above). They're also large units, since the more you compress the air the more problematic these issues become (as well as increasing mechanical wear and safety concerns).
However, despite these problems CAES has some huge advantages that will become more and more important in the context of a high renewable energy mix. The biggest advantage is that they have an incredible ESOI ratio (Energy Stored On Invested), which means they can store a massive amount of energy over their lifecycle compared to the energy required to build and maintain them. Specifically, the ratio is 240 for CAES, compared to just 10 for Lithium Ion. That means that over the long lifetime of a CAES system, it'll store far more energy and require far less care than a battery-based system. Compared to other high-ESOI systems, CAES is the least geographically-sensitive system (since a suitable cavern is not actually required, as you can just use pressurized tanks).
In the current fossil fuel and grid based electrical system CAES doesn't make too much sense most of the time due to the low energy efficiency. If you compare storing fuel to be consumed later vs storing electricity is CAES, you're basically giving up 48-60% of your output by storing it in CAES. Even inefficient peaker plants do better than producing the energy more efficiently and storing it in CAES. Hence, the poor adoption of CAES today.
In a highly intermittent electrical system though, CAES goes from being one of the worst options to one of the best options, since we're now comparing the relatively lossy CAES system to losing the energy entirely, since you can't store the sun or wind for later consumption. In this case, the incredible ESOI makes a big difference. A battery-based system might store 90% of the excess collected energy from a solar plant during a peak generation time, but it has to be replaced after it's stored 10 times its input energy requirement. Meanwhile a CAES system only stores (let's say) 50% of the energy at a given moment, but it can store 24 times as much energy than the battery system over its extremely long lifecycle (so, for an input of x energy building the storage system you've stored at least 12*x more energy than the battery-based system, and it may even be 24*x if the lower efficiency was already factored in to the ESOI ratio). Since the alternative to a storage scenario is 100% loss, and the CAES system outperforms the LiIon system in total energy stored per energy invested, it's the more energy efficient system over the lifecycle of the system.
In addition, there are ways of improving the energy efficiency. Systems that store the heat generated by compression and use it to warm the system up during decompression can get the efficiency up to 70% (which would bring the above example numbers up to 16.8*x, assuming it's not substantially more costly to build and maintain). Another option is to use the heat and cooling usefully as part of another system (ex. using the waste heat to power an industrial process or heat a home and the waste cold for refrigeration).
As another plus, there's no need for Lithium or any other rare element.
It should be noted there is at least some promise for miniaturized CAES. It does have more severe disadvantages at a small scale, but the heat and cold generated as a byproduct is also more useful in a domestic environment. That said, I doubt it'll ever replace batteries at the smallest scales (ex. a compressed air laptop would be crazy for a variety of reasons).
For more information (especially about possible miniaturization), read these articles:
If you want some solid evidence that you don't need to clear out the past generations to progress a cause, consider how fast recent social shifts have occurred. From 2004-2017, opposition to gay marriage in the US went from 60% to 32% (http://www.pewforum.org/fact-sheet/changing-attitudes-on-gay-marriage/). The US didn't lose 28% of its population in 13 years, it lost around 10.9% of its population in that time frame (calculated from this https://tradingeconomics.com/u...). Even assuming that every old person was in opposition, there's still at least a 17% shift in public opinion that has nothing to do with the elderly dying out. While it is true that the slower demographic trend helped reduce opposition to gay marriage, the real change has been driven by people changing their minds.
If you break down the numbers further (https://en.wikipedia.org/wiki/Public_opinion_of_same-sex_marriage_in_the_United_States), you'll also see that which US state a person is in (or their political ideology) are better indicators than just age (although age is still a good indicator). States in the northeast and on the west coast have overwhelming support for this issue, while states in the midwest and southeast are on the fence about it, even though none of these states have a serious age skew. Democratic Silent Generation individuals are less likely to oppose gay marriage than their Republican Millennial Generation counterparts. Changing public opinions come from changing people's minds, not removing old people with outdated opinions that they refuse to change.
It's also quite possible that healthy old people would actually be more mentally flexible than today's crippled old people. By reducing the damage of aging to the brain, very old people would likely not get so set in their ways. In addition, a lot of very old people today are basically just waiting around to die, instead of interacting with the mainstream society, so this reduces the pressure to change they would normally have applied to them. As long as they can hole up in their paid-off home and live off their retirement money and/or social security and/or have their family care for them, what do they care what other people think? However, if they were healthy then they'd almost certainly prefer to get out and interact (or at least do other activities that end up with frequent social interaction), leading to them changing their minds to keep up and fit in.
People from hundreds of years ago who were kept alive and well by science would have eventually shifted their opinions to match the prevailing social climate, and today they wouldn't really be that different from the current population, other than having seen more history firsthand (well, and having had hundreds of years to accumulate wealth...I don't think it'd really have that big an impact on social progress for a variety of reasons, after all it's not like wealthy and powerful people are young today, and the wealthy very old folks would be massively outgunned by wealthy younger old folks from more recent generations that had been oppressed by the oldest folks in the past, and pretty much everyone alive today would have lived through past periods of extreme inequality and instability and realize that working together to create a relatively equitable and well functioning system is much better for all than the current dysfunctional short-range thinking winner-takes-all system (meanwhile, the current winner takes all system makes perfect sense for the powerful when they have limited lifespans and life is relatively cheap, as they need to get everything now while the getting is good)).
Personally, I believe the concept of "progress", especially social progress, is a modern myth. Various situations get better and worse, social fashions change, but humans are still the same humans they've always been.
The idea of progress assumes there's some sort of perfect state that we're moving towards, but I
All sophisticated modern AI systems (except possibly Deep Learning) suffer from the knowledge acquisition problem https://en.wikipedia.org/wiki/... Deep learning has its own different problems (ex. how did it get that answer? nobody knows!), but if things change you can basically just retrain it with the newest data.
As your system/model gets bigger and bigger with more and more moving parts (dimensions, sub-models, rules, algorithms, data, etc.), it becomes more and more brittle, because all these different parts are interrelated, so moving one part can cause many others to shift and stop working. In particular, in an active field like cancer research you may have sudden and massive shifts that invalidate or alter a lot of your previous rules and data, setting you back so much that it's basically like starting over. Worse yet, you need to figure out which parts were invalidated by the new change and which data is still relevant. There are often situations where the scale is such that you require millions of EXPERT man hours to make the required changes. Worse yet, the often intense interdependence between components means that a large team putting in these millions of hours will have a crushing communication overhead. At this point, you just don't have the manpower available, no matter who you are, and with the communication overhead it may simply be impossible to make the changes required in any reasonable timeframe, no matter how many resources you have available. Once you can't make changes in any reasonable timeframe, the world moves on and you have to give up eventually, especially when further massive shifts put you deeper and deeper into technical debt.
Comparatively, many human doctors can each independently learn these things. It's still expensive, but it's a smaller cost per doctor spread out over many doctors, with very little communication overhead. Humans are much better at learning from relatively small amounts of data and/or theoretical models presented to them. The main downside here of course is inconsistency, with some doctors not learning the new research properly, leading to bad outcomes. Also, if it were possible to make the change to the AI system, then it could be replicated at scale, including scaling up for demand. You can't suddenly train many new oncologists, it takes years.
Deep Learning has a similar but less severe problem, since now it only relies on the data, but it does rely on a massive quantity of data. If there is a massive shift in our understanding, you need to either root out data that teaches it the wrong thing (likely the label is wrong in light of new developments, in your typical supervised learning use case) or you need to overwhelm the incorrect data with right data. Either solution ALSO relies on experts, but unlike the knowledge acquisition problem that most models suffer from, they do not need to communicate (only focusing on whether each isolated piece of data is correct) and they don't need to be mythical doctor-engineers that can rewrite rules while deeply understanding the domain (or pairs of domain expert and engineer that work in conjunction) since the domain experts can use a tool created by the engineers to create the data, or the data can be scraped from other sources. Once your data is corrected, you simply need to retrain the system.
Alas, deep learning is far from a silver bullet. It often needs obscene amounts of data to perform acceptably, and the more data it needs to learn, the more expensive it is to build and brittle it is during maintenance. Plus, there seems to be some areas it just can't handle, even with sufficient data. It excels the most when the problem space doesn't change very quickly and you can get lots of data about it (ex. natural language translation, photo recognition, etc.). A cat isn't suddenly going to be something else tomorrow, and cat photos are plentiful, thus cats are easy to identify with deep learning.
I use plain old rsync on Android via Termux. I haven't set it up to sync automatically, but it looks like cron is available too.
For iOS, there seems to be similar tools like Termius.
Admittedly, this approach isn't very slick, there's no nice UI. However, rsync is fast, reliable, open source, highly customizable, secure, etc. and it's not that complicated to do the basics.
You realize that they can do those things without a cashless society, right? Just pass a law to tax savings or bail out the banks.
Simply moving savings to cash doesn't prevent these actions because it's illegal not to follow these rules and they're relatively well enforced. Your money is already their money.
Slashdot Mirror
Disney has been keeping Gravity Falls unreleased for years. Will it appear on this service?
We should have noon match up with the sun being directly overhead.
Then we could regularly adjust our schedules to make sense for the seasons.
Why is 9-5 so sacred in our society?
Oracle changed their licensing recently, you really need to look into it:
https://java.com/en/download/r...
https://www.oracle.com/technet...
Also, OpenJDK is now open source releases of the Oracle JDK except without the Oracle enterprise support.
What to teach depends on your end goal.
If you want to teach a practical language that your grads can immediately use in practice, and want it to also work with Java: teach Kotlin. The (Android) mobile space has gone this way, and Kotlin works (more or less) seamlessly with Java, so it's becoming more and more popular in the enterprise as well. I think it will eventually supplant Java as the most popular JVM language, beating out competitors like the more academic Scala.
If you want to teach a practical language that your grads can immediately use in practice, and don't care about it being on the JVM: teach Python. There's many good reasons it's becoming more and more popular as a first programming language. As an added bonus, you can later teach them C and have Python and C work together.
If you care more about concepts than immediate practicality, then I'd suggest Julia. It's a practical and high performance language with some key language features you won't find in other languages outside of Lisp. It also works great with C. That said, there's not too many Julia jobs around, though I think it'll grow in popularity over time, and it's actually pretty similar to other languages, so it'd be easy for a student who knows Julia to pick up Java, C, Python, etc.
I program in Java professionally, and I have to say you really did a good job explaining what's good about Java in your post. The multi-threading thing in particular is a huge deal, as I haven't found another language that truly excels at multi-threaded development like Java does, at least with the same quality of life as Java provides (languages like Python and Ruby provide better development quality of life in my opinion, but without the excellent multi-threading support...or a type system, and many other things that would take a while to get into).
I also feel like recent improvements to the Java language (basically Java 7 & 8) may give it a new lease on life, though I think the long term picture for Java is pretty grim, because it's held back by historical baggage and Oracle. In particular, even if Java doesn't lose ground to newer non-JVM languages like Python, then Java will lose ground over time to Kotlin and other JVM-based languages like Scala or Groovy. These newer JVM languages simply don't have the historical baggage of Java (and if you love Java but haven't tried out Kotlin you really should).
Also, another problem for all of these JVM-based languages is that Oracle's recent licensing of the JVM is troubling. That said, it can survive this thanks to OpenJDK and other open source JVM initiatives, but we may be in for some rocky times in Java-land.
I’ve found Java quite effective for writing short programs of less than 100 lines to explore mathematical ideas, like how many polygons can meet at a snugly at a point(see http://math.ucr.edu/home/baez/...). So Java does not need massive projects to be really useful.
I recently discovered and learned Julia, and I think it's the ideal language for these kinds of math exploration problems. It combines the raw firepower of languages like C and a developer quality of life that exceeds even the best languages I've seen previously. It's very popular with the High Performance Computing crowd. It has high level features that no other popular language has, such as multiple dispatch, though it manages to be a very practical language at the same time. It also has some surprisingly nice libraries in certain areas, like their graph theory library. I think it has a very bright future ahead of it. If you haven't tried it, you really should (or if you tried it a few years ago, you should give it another spin now that it's at version 1.0).
That said, even Julia isn't perfect (yet?). Right now the biggest problem is that while it has excellent HPC support, it's still lagging behind on its multi-threading support. Basically, it currently assumes that you are fine with running your additional threads inside other processes (possibly on other machines), and Julia makes this very easy for the developer to accomplish, though it's not quite the same as having cheap in-process threads like in Java. Another issue, specific to enterprise development, is that the ecosystem is still evolving VERY rapidly, even though the language has finally settled into version 1.0, so libraries shift under your feet all the time (though fortunately, they have a top notch package management system to handle library versioning). Oh, also start up times can be pretty bad due to how compilation works (basically, you can pre-compile code, but it's easy to end up having to compile at least some of the program on startup rather than ahead of time), but there are solutions to this most of the time and it buys a lot of positives (though it can be especially annoying when a library didn't optimize their startup times).
I'm also somewhat concerned about the use of garbage collection in Julia. Some recent results make me think that automatic reference counting may be the future in this area. While this concern applies to Java as well (there's a reason Java programs are so memory hungry!), it's compounded in Julia because it complicates the bridging of C and Julia, which would otherwise be seam
He's talking about his personal experience being inside greenhouses kept at 1500ppm, not how plants react to it.
Your source only suggests that action be taken in a domestic setting at 1000ppm to avoid higher levels, and that it causes slight drowsiness starting around 1000ppm. This hardly constitutes a serious danger to health.
That said, a world in which just going outside causes mild sick building syndrome is pretty messed up.
https://www.youtube.com/watch?...
I'm not sure why they selected that snippet of text as their prime example when the made up story about Brexit and the continued prose from Pride and Prejudice from the included video were both more impressive.
That said, I don't see why they think it's so dangerous that they need to keep it secret. People already know that everything that not everything they hear on the Internet is true (or if they do, they're already too far gone!).
Too late, they've already been doing it for the last month and nobody noticed!
Here's an interesting option for controlling cyber-weapons without taking them entirely off the table. Instead of banning them or allowing unlimited secrecy, instead the following rules have to be followed:
1. The cyber-weapon has to be completely declassified within 1 year of becoming operational. (Perhaps a somewhat longer time could be mandated, such as 3 years or 5 years, but if the countdown becomes too long then the situation becomes more and more like unlimited secrecy)
2. The cyber-weapon has to be declared when it becomes operational, so we know when to start the declassification countdown.
3. The cyber-weapon cannot be used against the populace of the country operating the cyber-weapon. If this is the case, the exploits involved have to be reported to vendors immediately, and it has to be declassified more quickly as the vendors fix the issue. (What constitutes being usable against the populace is an interesting question, as stricter interpretations of this may rule out cyber-weapons usable against any public software, also note that private/secret forks of public software used by specific countries for country-specific purposes would almost certainly count as country-specific).
The overall effect of this should be that cyber-weapons are short lived and limited in scope (mainly attacking the secret capabilities of other countries instead of public software/infrastructure). It incentivizes improvement of existing nationally-used public software by defence actors, as they can no longer exploit loopholes in the software used by their own nation. It also incentivizes other countries to use public software for their infrastructure, and increases the quality of said infrastructure dramatically as everyone would want said infrastructure to be of top notch quality. The relatively quick declassification time means that any scandalous abuse of the system can be detected quickly (such as if they ignored rule 3, or if they created a cyber-weapon that was brutal enough to cause war crimes). The cyber-weapons declarations also serve as a deterrent, indicating that such weapons exist without giving away details about who is targeted or what it's for (at least until it is declassified, at which point there should be new weapons in existence). If it ever got to the point where it'd be impossible to create a new weapon before the old ones expired, then there wouldn't be many vulnerabilities out there and so we'd live in a very safe cyber-environment, making cyber-warfare moot.
If there were reasons to classify some cyber-weapons for longer periods, then I would recommend that they at least be required have an accurate summary of their purpose and reason for the classification extension declassified after the normal period, and they should be subject to substantial court scrutiny, with an ultimate declassification required at some later date. If this is allowed at all, it should be rare.
As for your other two suggestions, I definitely agree that software used in weapon systems is important to keep classified. However, I strongly disagree that criminal justice software should be secret. The benefits of public review of criminal justice software outweigh the possibility that some genius could find an exploit that makes them harder to bring to justice. Also, such exploits are more likely to be detected in the first place.
It should also be noted that the complete "source code" of the law itself is already out in the public view, yet we don't worry about someone finding an exploit in the law, even though it happens from time to time, allowing some people to exploit the system. Clearly, having a transparent code of law is much more important than catching every criminal.
What about computers from before SSDs becoming that cheap?
Trucking it in via tiny bottles is very silly, but I drink mostly "bottled" water.
Where I live the water is extremely gross to drink, even though I'm fairly sure it's safe to drink. We use our tap water freely in any situations where taste doesn't matter, like cooking (the taste of the water is overwhelmed by the taste of the food), showers, washing things, etc. It's so bad tasting that I am willing to pay extra to get less terrible tasting water.
However, individually packaged bottled water is still extremely stupid even in this situation. Why? Because we have several 5 gallon jugs that we cycle through, and get them refilled at the grocery store. The water out of their machine is incredibly tasty and only costs $0.39/gallon. Even at a gallon per person per day, it's only $11.86/month per person. Although this is outrageously expensive compared to tap water ($0.01/gallon locally, or $0.30/month per person), this premium is definitely worth it to me. It's definitely much better cost-wise than bottled water, which in disposable 1 gallon jugs is around $1/gallon ($30/month per person) and in disposable bottles it's around $6.30/gallon ($189/month per person).
You don't need an expensive dispenser machine to get started doing this, I just bought a manual water pump that attaches to the top of the jug and pumps the water out with a few presses of my hand. It's inexpensive, portable, and requires minimal maintenance. I also bought proper reusable dew caps for the jugs, instead of the expendable ones they try to sell you in stores, so the jugs and caps basically last forever, only needing to be washed out every once in a while.
It's also relatively good for the environment. Unlike the disposable options, it doesn't produce a ton of plastic. I'm pretty sure the grocery store is able to produce the water very cheaply and so these machines are pretty much pure profit for them, hence why they still have them instead of trying to get you to buy their even more expensive bottled water options. I don't know exactly what kind of filters they use (I think it's based on reverse osmosis), but the disposable parts are probably based on carbon.
I looked into running my own filtration machine (not the little filters that don't really do anything, but the real deal), and it's even more cost effective, at least a quarter the cost of buying it from the store, even at the small scale I would have been operating it at. I bought a high quality gravity-based machine (good for camping), and the problem with it is that the water tastes terrible using that option as well (I think it's removing too much, so there's no remaining tasty minerals). If I owned a house in an area with crappy water and had higher water demand, I'd probably install a full-scale reverse osmosis filter, just like the store has. As long as I'm renting though, the extra cost of buying the water at the store and the inconvenience of lugging it home is just fine.
One last side note: it is nice having an emergency water supply. I usually have about 4-8 days of good quality water on hand. If there were ever a serious outage, we would hardly be impacted by the situation. We wouldn't need to start boiling our water, we would just avoid the use of tap water for a while.
tl;dr Don't buy individually bottled water! There's so many better options, even if you want bottled water.
If you assume that you need to cover all current demand for electricity (and more due to growth), then you're massively over-complicating things. Renewables can often be used directly. Renewables and fossil fuels lead to very different usage patterns as well.
For instance, instead of installing excess solar capacity for heating we can get most of our heating from direct solar flux, and use insulation and building materials that absorb heat to make the most of the available energy. For cooling, we can vent hot air as it conveniently separates from cooler air automatically. Obviously, this doesn't cover 100% of our energy needs, since we may be run out of solar energy after many cold, cloudy days, and these problems get worse at latitudes that get less sunlight. That said, the energy required to cover this gap is far lower than replicating our current fossil fuel system using renewables. All the energy generated through direct use is energy that doesn't have to be generated through solar panels and wind turbines, and direct use definitely doesn't require any rare earth metals.
All high energy household applications can be replaced with direct solar when it's available. Heating water with solar is well-trodden territory. Dryers can be replaced by hanging clothes on clotheslines. The sun can light the indoors during the day, and with modern LED lighting the remaining time doesn't really require that much energy. If energy really needs to be cut down, then solar cooking can take over for gas and electric, but at this point we're hitting diminishing returns.
Of course, this isn't quite as helpful in extremely northern climes, and we'll need to stop building such terribly inefficient homes and living such terribly inefficient lifestyles. However, most people live close enough to the equator for this to cover most of our energy needs, and the colder countries can import solar, use geothermal and hydropower, or build local wind and nuclear to cover the difference. They also still benefit, just less so.
In terms of industry, there's a lot more industries that are harder to make renewable. Insanely high temperatures, electricity being consumed directly, warm up and cool down cycles and chemical reactions that take a long time, carefully controlled environments, high labor and capital costs that need to be made the most of, etc. That said, industry can adapt to renewables (ideally using energy sources directly), and at least in manufacturing the finished products themselves are a form of energy storage (the energy you used to produce them doesn't need to be spent when energy is scarce).
For instance, why run a factory 24/7 if you only have energy 12/7? Running 24/7 makes sense in a fossil fuel world because you get more use out of your capital investments, and there's no need to warm-up and cool-down the factory if it's running continuously, but it's not the end of the world to have to do these things in most industries. This situation will continue to improve as computers and software improves, as we can use computer control to cope better with the complexities of manufacturing using intermittent power (with buffers and forecasts). We can also turn off many factories completely for several days in very low energy situations, such as multiple cloudy days in a row, so we can ensure that energy is available to homes.
Of course, much like homes this doesn't cover every possible industry. Some domain-specific high energy processes cannot be reasonably halted or require extremely precise environmental control over a long time, or the costs of halting may be extremely high or safety critical for a variety of reasons. These specific cases require a lot less energy to cover than all industry, and in some cases they may want to shut down for the winter if they're especially likely to run into issues. For critical industries that need to be 24/7/365, they can use base load or stored energy.
As we move more towards service and information oriented industries, the situation gets even better, as peopl
Unfortunately, the problems you talk about are just reflections of the wider society (and technology's impact upon it).
I recently ran across some websites on sustainable living, and I realized that we're doing it all wrong. As a society, we're not running at 100% efficiency, or even 50%, but likely less than 1% efficiency. Considering our current inefficiency, we could all be living a better-than-modern lifestyle with relatively little work (no more than 20-30 hours/week) and enjoy said work far more, all while minimizing our impact on the planet. It was truly an eye opener for me.
Don't get me wrong, the sustainability community has some half-baked ideas too, but after hearing some of them I can never see the world the same way again.
As a concrete example, modern western greenhouses are designed incredibly stupidly. They're impossible to insulate and lose their heat quickly at night, so they have to be propped up by burning fossil fuels to warm the greenhouse. This design only makes sense if labor is expensive and fossil fuel energy is cheap (even with cheap solar, it'd be better to utilize the solar directly). It also makes growing greenhouse crops incredibly risky, since by the time you're ready to harvest, you've already spent a ton of electricity on each crop and could lose it all if the crop fails due to a disease or some other problem. The better solution is to build a brick wall facing south and build the greenhouse on the south face of the wall. Over the day, the wall will soak up energy from the sun and radiate it back out at night. Brick walls are relatively cheap and simple and last a near eternity, and the wall doubles as protection against cold winds from the north and a serious insulating layer, so it makes an ideal energy storage system. Lastly, you can throw an insulating blanket over the greenhouse during the night to keep the warmth in even more effectively. At temperate latitudes, this creates an environment where you can grow temperature-sensitive crops all-year round without any heating, even in the depths of a midwestern winter (and at higher latitudes, it will drastically reduce the amount of heating required to sustain livable temperatures).
At first this might seem like a bit of trivia, perhaps only applicable to farmers. However, I don't see any reason why this wouldn't be applicable in a suburban setting. A greenhouse of this design about the same size as a typical midwestern suburban home can consistently feed dozens of people year-round. They are very popular in China, likely because they need the higher efficiency to feed their massive population and much of their population lives at latitudes where this design works ideally. In a sprawling US city, each home could have a relatively small greenhouse that would be able to easily feed the family living there and produce some excess for sale or storage, maintained by the family living there. Even in denser situations, a single properly-situated house-sized communal greenhouse could feed a dozen apartments year round (or the suburban ring could produce more food per family for sale to the city center). Setting this up and maintaining it isn't rocket science, and yet I have not seen such a greenhouse in the midwestern US, despite it being the perfect location.
Why do we not do this as individuals? I did a cost-benefit estimate, and a house-sized greenhouse would produce $60,000-120,000 in produce per year with an up-front investment of $30,000 and minimal maintenance costs thereafter (maybe $5,000 a year and likely less). In a suburban setting, there would be minimal transport costs, especially if food was transported to the suburb it was being grown in (perhaps being sold to neighbors). The labor costs are hard to figure out, but growing the produce would likely be no more than full time work for one person. $55,000 is a great wage for gardening/farming, and I do think the higher profit figures are more realistic. This work would also be good exercise.
I don't currently own a house, so I'm now looking fo
Too bad it's the young people who are needed to prop up the social security system who are dying in unusually large numbers...
https://nca2018.globalchange.g...
https://en.wikipedia.org/wiki/...
Let's take videogames, for example. No one has really figured out how to combine a popular consumer-goods-type product like that with the philosophy of free software. The common advice when asked how to make a living writing free software of "provide a service for hire to support your product" only works with some very narrow types of products, and never really with consumer-level products. So, essentially, videogames simply don't exist in Stallman's universe.
Why not use a Kickstarter or Patreon style funding model?
I know it's harder to extract money out of consumers after you're done building the project, but perhaps extracting the maximum profit out of consumers shouldn't be the point.
The "catch" is energy efficiency, running CAES as an isolated unit gives 40-52% electrical-to-electrical efficiency, which is pretty terrible. This is because (for high pressures) you create heat on compression and cold on decompression, and unless you can store the heat at compression and return it for decompression, you'll have to provide external cooling and heat (which brings down the efficiency to the figure given above). They're also large units, since the more you compress the air the more problematic these issues become (as well as increasing mechanical wear and safety concerns).
However, despite these problems CAES has some huge advantages that will become more and more important in the context of a high renewable energy mix. The biggest advantage is that they have an incredible ESOI ratio (Energy Stored On Invested), which means they can store a massive amount of energy over their lifecycle compared to the energy required to build and maintain them. Specifically, the ratio is 240 for CAES, compared to just 10 for Lithium Ion. That means that over the long lifetime of a CAES system, it'll store far more energy and require far less care than a battery-based system. Compared to other high-ESOI systems, CAES is the least geographically-sensitive system (since a suitable cavern is not actually required, as you can just use pressurized tanks).
In the current fossil fuel and grid based electrical system CAES doesn't make too much sense most of the time due to the low energy efficiency. If you compare storing fuel to be consumed later vs storing electricity is CAES, you're basically giving up 48-60% of your output by storing it in CAES. Even inefficient peaker plants do better than producing the energy more efficiently and storing it in CAES. Hence, the poor adoption of CAES today.
In a highly intermittent electrical system though, CAES goes from being one of the worst options to one of the best options, since we're now comparing the relatively lossy CAES system to losing the energy entirely, since you can't store the sun or wind for later consumption. In this case, the incredible ESOI makes a big difference. A battery-based system might store 90% of the excess collected energy from a solar plant during a peak generation time, but it has to be replaced after it's stored 10 times its input energy requirement. Meanwhile a CAES system only stores (let's say) 50% of the energy at a given moment, but it can store 24 times as much energy than the battery system over its extremely long lifecycle (so, for an input of x energy building the storage system you've stored at least 12*x more energy than the battery-based system, and it may even be 24*x if the lower efficiency was already factored in to the ESOI ratio). Since the alternative to a storage scenario is 100% loss, and the CAES system outperforms the LiIon system in total energy stored per energy invested, it's the more energy efficient system over the lifecycle of the system.
In addition, there are ways of improving the energy efficiency. Systems that store the heat generated by compression and use it to warm the system up during decompression can get the efficiency up to 70% (which would bring the above example numbers up to 16.8*x, assuming it's not substantially more costly to build and maintain). Another option is to use the heat and cooling usefully as part of another system (ex. using the waste heat to power an industrial process or heat a home and the waste cold for refrigeration).
As another plus, there's no need for Lithium or any other rare element.
It should be noted there is at least some promise for miniaturized CAES. It does have more severe disadvantages at a small scale, but the heat and cold generated as a byproduct is also more useful in a domestic environment. That said, I doubt it'll ever replace batteries at the smallest scales (ex. a compressed air laptop would be crazy for a variety of reasons).
For more information (especially about possible miniaturization), read these articles:
I'm pretty sure Niels Bohr didn't believe in the Niels Bohr model of the atom late in his life.
If you want some solid evidence that you don't need to clear out the past generations to progress a cause, consider how fast recent social shifts have occurred. From 2004-2017, opposition to gay marriage in the US went from 60% to 32% (http://www.pewforum.org/fact-sheet/changing-attitudes-on-gay-marriage/). The US didn't lose 28% of its population in 13 years, it lost around 10.9% of its population in that time frame (calculated from this https://tradingeconomics.com/u...). Even assuming that every old person was in opposition, there's still at least a 17% shift in public opinion that has nothing to do with the elderly dying out. While it is true that the slower demographic trend helped reduce opposition to gay marriage, the real change has been driven by people changing their minds.
If you break down the numbers further (https://en.wikipedia.org/wiki/Public_opinion_of_same-sex_marriage_in_the_United_States), you'll also see that which US state a person is in (or their political ideology) are better indicators than just age (although age is still a good indicator). States in the northeast and on the west coast have overwhelming support for this issue, while states in the midwest and southeast are on the fence about it, even though none of these states have a serious age skew. Democratic Silent Generation individuals are less likely to oppose gay marriage than their Republican Millennial Generation counterparts. Changing public opinions come from changing people's minds, not removing old people with outdated opinions that they refuse to change.
It's also quite possible that healthy old people would actually be more mentally flexible than today's crippled old people. By reducing the damage of aging to the brain, very old people would likely not get so set in their ways. In addition, a lot of very old people today are basically just waiting around to die, instead of interacting with the mainstream society, so this reduces the pressure to change they would normally have applied to them. As long as they can hole up in their paid-off home and live off their retirement money and/or social security and/or have their family care for them, what do they care what other people think? However, if they were healthy then they'd almost certainly prefer to get out and interact (or at least do other activities that end up with frequent social interaction), leading to them changing their minds to keep up and fit in.
People from hundreds of years ago who were kept alive and well by science would have eventually shifted their opinions to match the prevailing social climate, and today they wouldn't really be that different from the current population, other than having seen more history firsthand (well, and having had hundreds of years to accumulate wealth...I don't think it'd really have that big an impact on social progress for a variety of reasons, after all it's not like wealthy and powerful people are young today, and the wealthy very old folks would be massively outgunned by wealthy younger old folks from more recent generations that had been oppressed by the oldest folks in the past, and pretty much everyone alive today would have lived through past periods of extreme inequality and instability and realize that working together to create a relatively equitable and well functioning system is much better for all than the current dysfunctional short-range thinking winner-takes-all system (meanwhile, the current winner takes all system makes perfect sense for the powerful when they have limited lifespans and life is relatively cheap, as they need to get everything now while the getting is good)).
Personally, I believe the concept of "progress", especially social progress, is a modern myth. Various situations get better and worse, social fashions change, but humans are still the same humans they've always been.
The idea of progress assumes there's some sort of perfect state that we're moving towards, but I
Well, then how are you going to tell the Internet about how much energy they're using?
All sophisticated modern AI systems (except possibly Deep Learning) suffer from the knowledge acquisition problem https://en.wikipedia.org/wiki/... Deep learning has its own different problems (ex. how did it get that answer? nobody knows!), but if things change you can basically just retrain it with the newest data.
As your system/model gets bigger and bigger with more and more moving parts (dimensions, sub-models, rules, algorithms, data, etc.), it becomes more and more brittle, because all these different parts are interrelated, so moving one part can cause many others to shift and stop working. In particular, in an active field like cancer research you may have sudden and massive shifts that invalidate or alter a lot of your previous rules and data, setting you back so much that it's basically like starting over. Worse yet, you need to figure out which parts were invalidated by the new change and which data is still relevant. There are often situations where the scale is such that you require millions of EXPERT man hours to make the required changes. Worse yet, the often intense interdependence between components means that a large team putting in these millions of hours will have a crushing communication overhead. At this point, you just don't have the manpower available, no matter who you are, and with the communication overhead it may simply be impossible to make the changes required in any reasonable timeframe, no matter how many resources you have available. Once you can't make changes in any reasonable timeframe, the world moves on and you have to give up eventually, especially when further massive shifts put you deeper and deeper into technical debt.
Comparatively, many human doctors can each independently learn these things. It's still expensive, but it's a smaller cost per doctor spread out over many doctors, with very little communication overhead. Humans are much better at learning from relatively small amounts of data and/or theoretical models presented to them. The main downside here of course is inconsistency, with some doctors not learning the new research properly, leading to bad outcomes. Also, if it were possible to make the change to the AI system, then it could be replicated at scale, including scaling up for demand. You can't suddenly train many new oncologists, it takes years.
Deep Learning has a similar but less severe problem, since now it only relies on the data, but it does rely on a massive quantity of data. If there is a massive shift in our understanding, you need to either root out data that teaches it the wrong thing (likely the label is wrong in light of new developments, in your typical supervised learning use case) or you need to overwhelm the incorrect data with right data. Either solution ALSO relies on experts, but unlike the knowledge acquisition problem that most models suffer from, they do not need to communicate (only focusing on whether each isolated piece of data is correct) and they don't need to be mythical doctor-engineers that can rewrite rules while deeply understanding the domain (or pairs of domain expert and engineer that work in conjunction) since the domain experts can use a tool created by the engineers to create the data, or the data can be scraped from other sources. Once your data is corrected, you simply need to retrain the system.
Alas, deep learning is far from a silver bullet. It often needs obscene amounts of data to perform acceptably, and the more data it needs to learn, the more expensive it is to build and brittle it is during maintenance. Plus, there seems to be some areas it just can't handle, even with sufficient data. It excels the most when the problem space doesn't change very quickly and you can get lots of data about it (ex. natural language translation, photo recognition, etc.). A cat isn't suddenly going to be something else tomorrow, and cat photos are plentiful, thus cats are easy to identify with deep learning.
So, right now you have useful t
I use plain old rsync on Android via Termux. I haven't set it up to sync automatically, but it looks like cron is available too.
For iOS, there seems to be similar tools like Termius.
Admittedly, this approach isn't very slick, there's no nice UI. However, rsync is fast, reliable, open source, highly customizable, secure, etc. and it's not that complicated to do the basics.
You realize that they can do those things without a cashless society, right? Just pass a law to tax savings or bail out the banks.
Simply moving savings to cash doesn't prevent these actions because it's illegal not to follow these rules and they're relatively well enforced. Your money is already their money.