I've heard from people with electric vehicles where they need to take long trips to see family on a regular basis that going electric was a non-issue. Their main point was along what you said about already needing to stop. What many don't mention is the 80/20 rule. You can charge to 80% in 20% of the time and the remaining 20% takes 80% of the time. Even if you take small 10 minute breaks every few hours, you can charge back up to 80% in no time and it's actually easier on the battery.
I think he meant cleaning up once the plant shuts down and possibly disposal of spent fuel. I'm sure there's at least some set aside as insurance in regards to accidents. The biggest question is why do coal power plants get away without "clean up" costs since they have magnitudes more nuclear waste than nuclear? Seems to me that coal should play by the same rules. Watch prices skyrocket to nearer what they should be.
What about the nuclear waste from coal? From what I can find, coal produces about 100x more radiation per unit of energy as nuclear. Since ramping up renewables fast enough is not feasible, what is your stance on how to handle replacing existing coal power and renewable is not a full option?
One should note that wind turbines do not require permanent magnets, removing the need for nearly all rare earth materials. Still whichever metal for windings.
You just opened up m brain and looked in. Slow but steady. Not really an option for me. I'm pretty much useless until I've created a mental model, though the more I learn, the faster I can identify an existing model and make minor tweaks. It is very important to not shoehorn models just because of familiarity.
Around the age of 11, I learned about SMP. It seemed obvious to me at the time that there must be instructions to "synchronize" the CPUs data because it would be physically impossible for all CPUs to be perfectly in sync without a propagation delay that would affect core frequency. Since the CPUs ran at the same frequency, they obviously had some way of checking or marking that a location in memory was ready.
I spent a lot of time doing thought experiments along these lines. By the time I wrote my first program 10 years later, I decided it was a good fit for multithreading. Took me about 24 hours to google how to do those "synchronizations" that I thought about 10 years prior. Multithreading pretty much worked exactly how I concluded, plus a few nifty features like CAS that simplified things.
Nearly every programmer I've dealt with has issues with thought experiments. They can't get past something concrete that is currently happening and that they can step through a debugger or see a trace. I've never understood this. Code will work exactly as it is coded. Why does one need to see a trace to see how it "might" move through the code and how each line of code "might" change the data.
I know a senior software engineer who can't figure out how to setup Outlook filter. He has to call IT every time. The filters he needs are of the "let me google that for you" difficulty.
I never said cheap. I will agree that "fast" is a subjective word, but I use it in the practical form of keeping up with the customer. It takes time for customers to change their processes as you make tools for them. If you can make tools faster than they can change their processes, you're "fast" in my opinion. There is also looking to the future. I already spend a lot of time abstracting the projects into my head in order to understand the fundamental issues and I constantly think of new ways for a current project to be used in novel ways. It's always fun when someone wants something new and you tell them you already thought about that general issue years ago and will only be a few days of work to implement it because you already designed another project to facilitate that use case.
I've got many projects under my belt where I was given a compressed timeline because the original team had been working on it for months and the deadline was fast approaching with the team projecting the project getting pushed due to unforseen complexities. I jumped in, quickly read the high level description, looked at what had been worked on, threw everything out, started fresh, and had everything done in a few weeks, plus my own personal improvements. These projects all were originally designed as one-offs, but I changed them to be modular and reusable and the projects have almost all been reused many times with virtually zero changes and bug fixes for many years. What generally starts as a 6 month throw away tool turns into a 5+ year tool that becomes central to many new services that few envisioned the tool to be useful for.
Every tech company that has to deal with hiring. There are a lot of blogs from Google, Microsoft, Facebook, etc where the whomever is responsible for overseeing hiring has experimented with all kinds of hiring strategies and trying to link future performance with experience, and every company has the same issue. The programmer with 10 years of experience has just as much chance of being great at their job as someone with 6 months experience and a recent graduate.
Some people go as far as to say that they even experimented with hiring from people with other degrees that are associated with critical thinking and zero programming experience. And with on the job training, were up and as productive within 6 months as pretty much anyone with a CS degree.
It's quite telling when explicitly hiring people with no programming degrees or experience as just as good "on average". It's universally described as a crap shot. Unless someone has exactly the skill that you need right now, like COBOL, when it comes to predicting performance, random is just as good and probably less biased to poor performers. The main reason for interviewing is to get to know the person to figure out if they're a good fit. The biggest benefit of experience is dealing with people, not technical skills. Technology cannot solve people issues and the biggest bottleneck to most projects are humans and their self-destructive irrational quirks.
From what I've read about creating tests, psychologists claim the current thought is testing creativity, which includes novel problem solving, is fundamentally impossible. Any question on a test is by definition not novel. It is already well understood and known about. And even the "answer" to complex novel issues are subjective. In the end, the only way to "test" is to have a track record. The proof is in the pudding. Does that person have a history of creating solutions that are stable and reliable.
Experience has virtually no correlation with understanding or even skill. Concrete skills have a half life of about 2-3 years. At my job, the learning curve is about 1 year before you break even between salary and value. My team thrives on change. We need to master new skills constantly. We need to be fast, correct the first time, and our projects need to be easy for others to use/manage, otherwise we get stuck supporting. And ain't nobody got time for that.
Abstract skills are a must. It doesn't matter that you know all of the documented tricks to minimize Java's garbage collection. You need to be able to minimize garbage collection in all managed languages such that when you get pulled onto a project that uses a managed language that you may never use again, that you design and implement it correctly the first time.
pi-hole or pfSense+pfBlockerNG can block nearly every ad at the network level, making all devices on your network nearly ad free with zero client configuration.
Unless you're bean counting, performance reviews are subjective. The fundamental problem is quality matters and measuring quality of intangible things is a bit difficult. One programmer finishes projects faster but more issues crop up using their projects but nothing is technically "wrong" and another programmer is 1/2 the speed of the first, but their projects seem to have no issues.
Relational databases are not ideal for non-relational data? I see the opposite in my line of work. People using document database for relational data, then they have to constantly fix data inconsistencies. I always just assumed that I should use the best tool for the job, but I found out in the real world that most just use whatever they're most used to or is the current fad. Technology is magic. If you want your project to succeed, you need to do the current fad rain dance. If that's RDBMS, then you do the RDBMS rain dance, if the current fad is no-sql, then the no-sql rain dance.
Rare earth materials is not needed. Wind turbines can be without magnets and will reduce their efficient some bit, but they'd still be highly effective. Lithium ion batteries are about to require no rare earth materials and will start mass production in the next year with increased charge rates and more cycles. Then there's iron flow batteries which are highly scalable, but do need more testing to make sure they're being designed to minimize maintenance.
Batteries are already getting cheap enough that for above certain peak to average ratios, they're cheaper than adding more peaker power plants or ramping up/down base load power plants. And by "cheap enough" I mean they pay themselves off in 1-2 years. Of course some of these high peek to average ratio are being driven by uncontrollable renewable power. But it can still be an overall win because many areas in the world renewable plus batteries is already cheaper than the 1-2 year fuel cost of a fossil fuel power plant. Operation costs of fossil fuel power plants is ridiculous and seems to not be discussed and only up-front costs are talked about. And that's excluding the theoretical costs of pollution differences which makes renewable a theoretical instant win.
There may be social value in having people in rural areas that are getting electricity, even if there is no private profit in it. In the case of many farms, this is true. Some people would say "just cover the farms", but who the hell is going to live on a farm with no neighbors? And you'd have to be a complete moron to argue that we don't need farms.
Money is actually quite poor at reflecting the value of necessities, but is excellent for luxuries. When you have a society where only 2% of the population needs to work in order for everyone to survive, money would be pointless if everyone was unemployed. So the government steps in, plays with some subsidies to make certain basics are "cheaper" than what a free market would create because again, money does not represent public value, only private value.
A simple difference between public and private values are it is in my private interest to dump my pollution and not care about anyone down stream. But just because it's good for me doesn't mean it doesn't cause more overall damage to everyone else. Then there's "the whole is greater than the sum of its parts". A person contributes more value to society than they get paid. The per working person average GDP is 6x than of the average working person's income. This also applies to healthcare. Each person who dies to a treatable health issue is a onetime reduction to the GDP of $1mil. Just because the person can't afford $100k in hospital bills doesn't mean it isn't worth covering them.
I'm from the USA and my insurance contract states that marriage for economic reasons and not "romantic" reasons is classified as insurance fraud. I guess quite a few people were getting married just because insurance is so expensive and the insurance company doesn't want to cover those "fraudsters". And it's not a bad company, it's a really good company. Low rates, lots of coverage, non-profit. They only work in my state and are located in my state, and they even will cover your "significant other", even if same sex and our state does not recognize same sex, regardless if your married or not. But you'd better not commit insurance fraud by marring for economic reasons.
I had a recent Cholecystectomy. It was out-patient. Only took about 2 hours by the time I got to the hospital, the procedure itself was less than 15 minutes. I was up and walking by the time I got home. I had a water proof breathable bandaid that I can purchase was Walmart, allowing me to shower and I didn't have to remove it for a week. No stitches. Scars look like feint age spots. The doctor even said I would eat anything that I wanted immediately after surgery.
My computer boots in about 10 seconds, that's cool. Several of my games support streaming downloads, allowing me to start playing as soon as the first 200MiB download, which is only a few seconds. I have a 1ms ping to most CDNs, 6ms to most servers, and even when my connection is fully saturated, I still have less than a 10ms ping to the general internet.
I no longer have to defrag my harddrive. I can download an ISO, burn it to a $10 64GiB flash drive, and reboot within less than a minute if I know exactly what I need to do. Not having to go into my bios and flipping dip-switches on daughterboards and peripherals is awesome. USB, just plug it in. The quality of search engines brings almost all knowledge instantly. Programming frameworks and libraries are superb comparatively.
The overall quality of life dealing with technology is leaps and bounds better.
AIs are more pathological than humans. Humans have an inherent amount of chaos in decision making that allows for people to have a more difficult time, but not out right blackballed. The biggest issue with AIs is they're only as good as the data their fed, and the more you use AIs, the more biased the data becomes. AIs tend to be self reinforcing, and most humans fall victim to believing the AIs are correct because the create a kind of self-fulfilling prophecy.
A simple example might be that a person gets a credit risk score based on an AI. Then because of that potentially poor score, they have to pay higher interest. Maybe this person was low risk, but incorrect labeled as high risk. But now because they're labeled as high risk, they become higher risk because they have more difficulty with acquiring and managing their credit due to limitations. Then other AIs see this and they get trained to be like "this person is high risk and treating them as high risk makes them higher risk, so treat them as even higher risk, which makes them higher risk." Rinse and repeat until it snow-balls to the point that the person can no longer get credit. Yay AIs.
Slashdot Mirror
I've heard from people with electric vehicles where they need to take long trips to see family on a regular basis that going electric was a non-issue. Their main point was along what you said about already needing to stop. What many don't mention is the 80/20 rule. You can charge to 80% in 20% of the time and the remaining 20% takes 80% of the time. Even if you take small 10 minute breaks every few hours, you can charge back up to 80% in no time and it's actually easier on the battery.
Once you get into the millions of degrees, the 2x difference between f and c/k is irreverent. Just state the magnitude and say "degrees".
I think he meant cleaning up once the plant shuts down and possibly disposal of spent fuel. I'm sure there's at least some set aside as insurance in regards to accidents. The biggest question is why do coal power plants get away without "clean up" costs since they have magnitudes more nuclear waste than nuclear? Seems to me that coal should play by the same rules. Watch prices skyrocket to nearer what they should be.
What about the nuclear waste from coal? From what I can find, coal produces about 100x more radiation per unit of energy as nuclear. Since ramping up renewables fast enough is not feasible, what is your stance on how to handle replacing existing coal power and renewable is not a full option?
One should note that wind turbines do not require permanent magnets, removing the need for nearly all rare earth materials. Still whichever metal for windings.
Budweiser is like Coke, they can't just change the flavor, no matter how much better they think they can make it.
You just opened up m brain and looked in. Slow but steady. Not really an option for me. I'm pretty much useless until I've created a mental model, though the more I learn, the faster I can identify an existing model and make minor tweaks. It is very important to not shoehorn models just because of familiarity.
Around the age of 11, I learned about SMP. It seemed obvious to me at the time that there must be instructions to "synchronize" the CPUs data because it would be physically impossible for all CPUs to be perfectly in sync without a propagation delay that would affect core frequency. Since the CPUs ran at the same frequency, they obviously had some way of checking or marking that a location in memory was ready.
I spent a lot of time doing thought experiments along these lines. By the time I wrote my first program 10 years later, I decided it was a good fit for multithreading. Took me about 24 hours to google how to do those "synchronizations" that I thought about 10 years prior. Multithreading pretty much worked exactly how I concluded, plus a few nifty features like CAS that simplified things.
Nearly every programmer I've dealt with has issues with thought experiments. They can't get past something concrete that is currently happening and that they can step through a debugger or see a trace. I've never understood this. Code will work exactly as it is coded. Why does one need to see a trace to see how it "might" move through the code and how each line of code "might" change the data.
I know a senior software engineer who can't figure out how to setup Outlook filter. He has to call IT every time. The filters he needs are of the "let me google that for you" difficulty.
I never said cheap. I will agree that "fast" is a subjective word, but I use it in the practical form of keeping up with the customer. It takes time for customers to change their processes as you make tools for them. If you can make tools faster than they can change their processes, you're "fast" in my opinion. There is also looking to the future. I already spend a lot of time abstracting the projects into my head in order to understand the fundamental issues and I constantly think of new ways for a current project to be used in novel ways. It's always fun when someone wants something new and you tell them you already thought about that general issue years ago and will only be a few days of work to implement it because you already designed another project to facilitate that use case.
I've got many projects under my belt where I was given a compressed timeline because the original team had been working on it for months and the deadline was fast approaching with the team projecting the project getting pushed due to unforseen complexities. I jumped in, quickly read the high level description, looked at what had been worked on, threw everything out, started fresh, and had everything done in a few weeks, plus my own personal improvements. These projects all were originally designed as one-offs, but I changed them to be modular and reusable and the projects have almost all been reused many times with virtually zero changes and bug fixes for many years. What generally starts as a 6 month throw away tool turns into a 5+ year tool that becomes central to many new services that few envisioned the tool to be useful for.
Every tech company that has to deal with hiring. There are a lot of blogs from Google, Microsoft, Facebook, etc where the whomever is responsible for overseeing hiring has experimented with all kinds of hiring strategies and trying to link future performance with experience, and every company has the same issue. The programmer with 10 years of experience has just as much chance of being great at their job as someone with 6 months experience and a recent graduate.
Some people go as far as to say that they even experimented with hiring from people with other degrees that are associated with critical thinking and zero programming experience. And with on the job training, were up and as productive within 6 months as pretty much anyone with a CS degree.
It's quite telling when explicitly hiring people with no programming degrees or experience as just as good "on average". It's universally described as a crap shot. Unless someone has exactly the skill that you need right now, like COBOL, when it comes to predicting performance, random is just as good and probably less biased to poor performers. The main reason for interviewing is to get to know the person to figure out if they're a good fit. The biggest benefit of experience is dealing with people, not technical skills. Technology cannot solve people issues and the biggest bottleneck to most projects are humans and their self-destructive irrational quirks.
From what I've read about creating tests, psychologists claim the current thought is testing creativity, which includes novel problem solving, is fundamentally impossible. Any question on a test is by definition not novel. It is already well understood and known about. And even the "answer" to complex novel issues are subjective. In the end, the only way to "test" is to have a track record. The proof is in the pudding. Does that person have a history of creating solutions that are stable and reliable.
Experience has virtually no correlation with understanding or even skill. Concrete skills have a half life of about 2-3 years. At my job, the learning curve is about 1 year before you break even between salary and value. My team thrives on change. We need to master new skills constantly. We need to be fast, correct the first time, and our projects need to be easy for others to use/manage, otherwise we get stuck supporting. And ain't nobody got time for that.
Abstract skills are a must. It doesn't matter that you know all of the documented tricks to minimize Java's garbage collection. You need to be able to minimize garbage collection in all managed languages such that when you get pulled onto a project that uses a managed language that you may never use again, that you design and implement it correctly the first time.
pi-hole or pfSense+pfBlockerNG can block nearly every ad at the network level, making all devices on your network nearly ad free with zero client configuration.
Unless you're bean counting, performance reviews are subjective. The fundamental problem is quality matters and measuring quality of intangible things is a bit difficult. One programmer finishes projects faster but more issues crop up using their projects but nothing is technically "wrong" and another programmer is 1/2 the speed of the first, but their projects seem to have no issues.
Relational databases are not ideal for non-relational data? I see the opposite in my line of work. People using document database for relational data, then they have to constantly fix data inconsistencies. I always just assumed that I should use the best tool for the job, but I found out in the real world that most just use whatever they're most used to or is the current fad. Technology is magic. If you want your project to succeed, you need to do the current fad rain dance. If that's RDBMS, then you do the RDBMS rain dance, if the current fad is no-sql, then the no-sql rain dance.
Rare earth materials is not needed. Wind turbines can be without magnets and will reduce their efficient some bit, but they'd still be highly effective. Lithium ion batteries are about to require no rare earth materials and will start mass production in the next year with increased charge rates and more cycles. Then there's iron flow batteries which are highly scalable, but do need more testing to make sure they're being designed to minimize maintenance.
Batteries are already getting cheap enough that for above certain peak to average ratios, they're cheaper than adding more peaker power plants or ramping up/down base load power plants. And by "cheap enough" I mean they pay themselves off in 1-2 years. Of course some of these high peek to average ratio are being driven by uncontrollable renewable power. But it can still be an overall win because many areas in the world renewable plus batteries is already cheaper than the 1-2 year fuel cost of a fossil fuel power plant. Operation costs of fossil fuel power plants is ridiculous and seems to not be discussed and only up-front costs are talked about. And that's excluding the theoretical costs of pollution differences which makes renewable a theoretical instant win.
Profit != value
There may be social value in having people in rural areas that are getting electricity, even if there is no private profit in it. In the case of many farms, this is true. Some people would say "just cover the farms", but who the hell is going to live on a farm with no neighbors? And you'd have to be a complete moron to argue that we don't need farms.
Money is actually quite poor at reflecting the value of necessities, but is excellent for luxuries. When you have a society where only 2% of the population needs to work in order for everyone to survive, money would be pointless if everyone was unemployed. So the government steps in, plays with some subsidies to make certain basics are "cheaper" than what a free market would create because again, money does not represent public value, only private value.
A simple difference between public and private values are it is in my private interest to dump my pollution and not care about anyone down stream. But just because it's good for me doesn't mean it doesn't cause more overall damage to everyone else. Then there's "the whole is greater than the sum of its parts". A person contributes more value to society than they get paid. The per working person average GDP is 6x than of the average working person's income. This also applies to healthcare. Each person who dies to a treatable health issue is a onetime reduction to the GDP of $1mil. Just because the person can't afford $100k in hospital bills doesn't mean it isn't worth covering them.
I'm from the USA and my insurance contract states that marriage for economic reasons and not "romantic" reasons is classified as insurance fraud. I guess quite a few people were getting married just because insurance is so expensive and the insurance company doesn't want to cover those "fraudsters". And it's not a bad company, it's a really good company. Low rates, lots of coverage, non-profit. They only work in my state and are located in my state, and they even will cover your "significant other", even if same sex and our state does not recognize same sex, regardless if your married or not. But you'd better not commit insurance fraud by marring for economic reasons.
Depends on what they define as "100% Italian". Having many generations born and bred in Italy does not mean 100% Italian.
Depends on how "work with" is defined.
If I ignore the emails long enough, I feel that they're no longer important and I can just "archive" them for searching later.
I had a recent Cholecystectomy. It was out-patient. Only took about 2 hours by the time I got to the hospital, the procedure itself was less than 15 minutes. I was up and walking by the time I got home. I had a water proof breathable bandaid that I can purchase was Walmart, allowing me to shower and I didn't have to remove it for a week. No stitches. Scars look like feint age spots. The doctor even said I would eat anything that I wanted immediately after surgery.
My computer boots in about 10 seconds, that's cool. Several of my games support streaming downloads, allowing me to start playing as soon as the first 200MiB download, which is only a few seconds. I have a 1ms ping to most CDNs, 6ms to most servers, and even when my connection is fully saturated, I still have less than a 10ms ping to the general internet.
I no longer have to defrag my harddrive. I can download an ISO, burn it to a $10 64GiB flash drive, and reboot within less than a minute if I know exactly what I need to do. Not having to go into my bios and flipping dip-switches on daughterboards and peripherals is awesome. USB, just plug it in. The quality of search engines brings almost all knowledge instantly. Programming frameworks and libraries are superb comparatively.
The overall quality of life dealing with technology is leaps and bounds better.
AIs are more pathological than humans. Humans have an inherent amount of chaos in decision making that allows for people to have a more difficult time, but not out right blackballed. The biggest issue with AIs is they're only as good as the data their fed, and the more you use AIs, the more biased the data becomes. AIs tend to be self reinforcing, and most humans fall victim to believing the AIs are correct because the create a kind of self-fulfilling prophecy.
A simple example might be that a person gets a credit risk score based on an AI. Then because of that potentially poor score, they have to pay higher interest. Maybe this person was low risk, but incorrect labeled as high risk. But now because they're labeled as high risk, they become higher risk because they have more difficulty with acquiring and managing their credit due to limitations. Then other AIs see this and they get trained to be like "this person is high risk and treating them as high risk makes them higher risk, so treat them as even higher risk, which makes them higher risk." Rinse and repeat until it snow-balls to the point that the person can no longer get credit. Yay AIs.