I had a similar thought upon seeing the demo presentation...
You could build that interaction with a Genesys IVR system.... heck, you could have built it 10 years ago... and it wouldn't have a lick of AI involved. It would be extremely cool and a great use of IVR, but it wouldn't be groundbreaking.
Building that interaction without any specific question/response trees and completely using AI to understand the conversation is a completely different animal and is amazing. It isn't clear exactly where this falls on the spectrum, nor is it clear that this is a voice assistant that goes beyond what google can do or is more widely available than google's (or Amazon, or Apple, or...)
I'm not saying it isn't either..... these are huge companies with huge customer bases and extremely large incentives to get natural language interactions automated in order to serve their customers better, so it stands to reason that they'd make this sort of breakthrough. But that demo wasn't really proof of anything earth shattering in AI. The fact that these are very task-specific agents suggests that they are closer to a very sophisticated IVR than they are to a complete natural language AI that can handle a wide array of requests in natural language.
In fact, the article pretty much says this... they have "shipping agents" that handle scheduling package delivery. And "scheduling agents" that they are working on for making reservations.
Building this with IVR technology would be a hard and complex task... but it doesn't require a full AI to do something like that demo.
He is right about working in blocks. Some jobs are like that - you are more productive if you have long periods of interrupted time working on it with intense focus. There is a startup cost to a lot of work - putting it down and picking it up takes time. Analytic work like data mining, SQL programmer, writer, graphic artist, etc. would probably fit this model.
Other jobs are more intense in a different way, and reducing the consecutive hours actually increases productivity. Things like a factory line inspector or a data entry clerk or call center work might fall into this category.
It is almost as if a top-down, one-size-fits-all mandate isn't such a great idea.
Yeah, the Tesla Roadster is the vehicle that changed the perception of electric cars.
Before that they were slow, light and almost concept-car in appearance and finish - GM EV-1. Everyone was thinking super-econo-box.
The Roadster was specifically designed for the task of demonstrating that electric cars can be cool. And having ridden in one at the time they came out, I can confirm that it was an experience like no other. The huge amount of torque and acceleration made them immensely fun, despite their shortcomings.
Then they came out with the Model S, which demonstrated just how sporty and cool could be translated to luxury sedan and really cool.
The stated mission of Tesla was to act as a demonstration of how electric vehicles that would appeal to a mass market could actually be built. I'd say they accomplished that mission in spades. Of course, once they got a little taste of success, they redefined their goals upward. But that doesn't mean they didn't already accomplish what they set out to do.
They have been comparing the size of it to an A380 airplane (which seats 800). Were they to decide to send a tourist trip with 10 people aboard, the passengers would likely feel less cramped than the ISS astronauts. Estimating the cost of launch is tough, with the high-reusability goal designed to bring the costs down into the single digit millions per launch. So let's say they can sell it for $100 million - that's only 10 million per passenger. 10 passengers in that large ship would be pretty luxurious, as spaceflight goes.
As for the time... yeah, you can't cut the ride short. But there's plenty of vacation spots here on earth where people are happy to go knowing that rescue is days away at the earliest.
One man's gentrification is another man's urban renewal.
To me they're fighting the wrong battle - they should be fighting for higher wages so that as gentrification occurs they're not pushed into ghettos but move laterally, maintaining their standard of living.
That is absolutely the wrong direction to push.
If you want to keep rents (and housing prices) affordable, you have to remove barriers to affordable (and not affordable) housing.
Where are housing prices the highest? Cities like San Francisco, New York City, etc. that have strong controls in place that make developing new housing very difficult.
If you removed these restrictions and let developers build loads of new, expensive, luxury housing, you'll find that the value of the old housing stock won't appreciate like it does if you restrict these developers. In fact, you will likely find that the older, less luxuriously appointed housing drops in price.
There are some pretty strong forces that oppose looser zoning restrictions - including existing homeowners who would be very happy to have their buildings double and triple in value. But the ones that make absolutely no sense are folks who think they are advocating for more affordable housing by advocating rent controls and set-asides for affordable housing in any new project. You don't get lower prices by limiting supply and capping prices. You get shortages.
I doubt they deleted the app. Because most people buy their phones from the carriers, and Facebook comes as a part of the bloatware package that you can't delete at all. At least not without root access... and these days even most slashdot readers don't root their phones.
In academic circles it is commonplace to have a journal club organized around your specialty. A group of researchers and graduate students meets to discuss journal articles relevant to their field. Sometimes it is exciting because you get to talk about someone opening a new area of inquiry or finding an interesting new fact. Other times it is exciting because you pick apart a paper that is completely unsupported by the results included in the paper.
Those are actually the best. There are a lot of articles that get published that never should have been published... even by the most prestigious journals.
Journal club is perhaps the best place to learn how to understand good experimental design and clear reasoning based on results.
The best club I was ever in did it in a very interesting way. We would bring in cutouts of just the materials and methods and the results. Then everyone had to analyze the results and draw our own conclusions as a group. Only then would we unveil the paper's conclusions section. It made for some very clear-headed analysis. Very often the authors of a paper don't include enough data to support their conclusions, even if they actually do have other results (not published) that further bolster their conclusions.
Anyway, the whole process is a lot of fun and it often looks just like carysub's take above.
Because they know that they are on the side of everything that is true and just and right. And they know that nobody will ever exercise this control who isn't true and just and right, just like they are. It is impossible that someone could ever decide that their ideas were unacceptable. Because reality has a liberal bias. And they are on the right side of history. You'll see....
I'm really surprised at the slashdot take on this.
We had this discussion here a long time ago. It was argued endlessly and Slashdot as a community was strongly convinced that all information should be free. And online services are not responsible for user content. Remember that? It was very, very important that online services like Slashdot avoid getting in the business of taking editorial control over user generated content. Because right-wing types kept threatening to either regulate offensive speech or hold internet companies liable for offensive and/or slanderous content.
Apparently the Slashdot community has collectively lost their minds. Are we really forgetting the "hands off" tradeoff that was made to keep these companies exempt from being held responsible for user-posted content?
And yet here we are, with politicians loudly agreeing that the big user-generated-content companies should exercise editorial control over the content on their platform. In many cases these are the exact same politicians who crafted the policy that allowed the internet to have Facebook, Twitter and Youtube. People like Chuck Schumer, Nancy Peloci, etc. were there when the policy was created. Their fingerprints are all over it, so they know full well that deciding which political views are acceptable negates the deal, opening these platforms up to liability for the user generated content they publish.
When they started with their "truth commission" ideas back in 2015/16 I thought this was obvious. It should be painfully obvious by now. But somehow the passage of time has allowed everyone to forget the terms of the deal. I wonder if they'll rue the day they chose to follow this path, or if everyone will somehow forget the deal and craft a new world where you can exercise editorial control over content and yet not be liable for that content.
Copyright protected images retain their copyright? Even if someone publishes them on a website that is accessible to the public?
I'm shocked at this disturbing turn of events. Next they'll be claiming that news articles retain their copyright even after they've been published online!
How is this even news? We have a lot of anti-IP folks around here, but even they have to acknowledge that this has been settled law... well, as long as there has been an internet. Longer, I guess. I mean, you couldn't just cut a photo out of a magazine and use it in your own ad copy either.
I support your take, but I'll add that computer learning programs are already very good at basic skill building.
My kids do a couple of math programs - Reflex math is used to build specific skills in basic math operations. It measures how quickly and well they do and keeps feeding them practice until they master each "math fact". It moves them along at their own pace. So one kid might be working on 2 digit multiplication while the next kid is still trying to master subtraction.
They do the same for reading. It starts out really simply, but as they move forward the computer parses out the difference between reading, reading comprehension and various forms of advanced comprehension (reading between the lines). There is often a big gap between the level a student is mastering in the various areas of reading. The computer works to focus on their weaknesses while continuing to move their strengths forward. This doesn't work quite as well as the math, but it still is way, way ahead of our old readers and quiz packets.
Yeah, that was basically what I was about to write. Try "hands on" calculus. That really makes sense. And everyone should make a pilgrimage to Philadelphia when studying the constitution in history class. And Greece to learn about the history of western civilization. It is more engaging of the entire body to actually visit Sparta.
This opinion piece seems like it was written by someone who has no idea of how computers are being used in schools.
He's a poli-sci professor. I wonder how he engages the whole body in his sophomore political science class. Sure, you could act out democratic processes... but exactly how would you make something meaningful and efficient about city-states, or monarchies? And political science is probably one of the easier ones.
Science and engineering have always been taught "hands on", and today they are increasingly "hands on" at the elementary and secondary levels, even as we introduce more computer-based learning.
Education today is becoming increasingly customized. Teachers used to move the entire class along as a block - maybe breaking off a couple of advanced kids for independent study. Today they have individual plans for each student and they juggle many groups simultaneously. Computers greatly aid this process. (the only downside being that kids and teachers spend way more of their time dealing with testing than they probably should.)
Computers have made math a much more "hands on" experience than it used to be. And they teach each kid at a customized pace, emphasizing areas of weakness until they have mastered important skills. It is still fairly primitive, but it is a major improvement over rows of kids reciting their times tables as a group.
it's not nothing to do with moral policing. Credit card transactions are effectively loans. In large parts of the world you have a legal right to dispute any charge on your card as a result. Adult content has a high percentage of disputes (probably from guys who's wives/girlfriends notice the charge). Even if you can prove the charge is valid it's still expensive to do so. Hence why nobody wants to be involved in it.
With corporations always, always, always follow the money. Anything bigger than a leomonade stand is completely amoral.
This is a great take.... and unfortunately it is just not correct.
The US Federal government began threatening banks back in 2013 that if they did business with disfavored industries, they risked being taken down by the Feds. It is often sold as being about "money laundering", but it targeted legal business that were in disfavor with the administration like firearms dealers, check cashing services and payday lenders. Along with this other groups were impacted like adult entertainers like porn actresses and producers.
These are perfectly legal businesses that the government decided they wanted to run out of business by threatening anyone who does business with them.
And that is why Patreon is even in the conversation. Because they can have trouble even getting bank accounts. So adult entertainers and others have been forcibly "unbanked". Now they have to hunt around for other ways to move money and get paid.
Even though the official program was recently ended, the effect lives on.
The last 18 years has seen a massive shift away from civil liberties in many ways (USA PATRIOT Act, warrantless wiretapping, metadata collection, etc.) and things like Operation Choke Point have changed the culture to the point where people actually see this sort of thing as acceptable. It isn't.
No, he's just using an example that people might be familiar with.
There are many niche packages out there that used to be sold as software that you'd have a VAR set up at your site. Then they started offering a hosted service. Then they started offering "cloud based" service.
And now they've started discontinuing the "you set up your own server in your datacenter" version.
I'm sure most of them would still negotiate with you and set it up for a price, but the business model is definitely moving to hosted services. Support is just so much cheaper if you don't have to deal with hundreds of different individual deployments with their own hardware and software idiosyncrasies.
At least with AWS or Azure, you can get a real human to talk to when shit goes wrong, you just have to pay for the privilege. Not sure Google even has the concept of customer service, but their competitors do.
I think this is the salient point.
If you are dealing with an enterprise service, you have a dedicated customer support rep and team. And you pay for this.
Getting the "google chat was not available" problem means that you didn't have an enterprise level contract.
Law firms are notorious for being behind the times on tech (and most other business needs).
Once you get to be a midsized company, most companies will have generators to keep the doors open in an extended outage.... at least if you are located in an area where this is a reasonable possibility.
I was with a company in hurricane territory, so I not only developed fully redundant data centers across 3 locations (one being a colo provider), we also had a huge generator in place to run most of the company in extended power outages. It worked great when we were out of power for 2 weeks after a hurricane. I also had contracts in place for emergency office space for critical employees if the building were to be destroyed.
But our costs were over a quarter million a day if we were down (not including payroll). And if we were down for more than 4 days we risked violating contracts that would have placed tens or hundreds of millions at risk. So the cost of setting up a $120k generator system wasn't that big of a problem.
It all depends on what your costs of down time are. For BankAmerica or AT&T, there can be no downtime, so they pay through the nose for super-hardened systems. For a 2 man law firm, a battery backup that shuts the system down while they go hit the links waiting for the power to come back on might be sufficient.
It is funny to read the "puny $6k/mo" denigration. When you are not exposed to these enterprise level vendors, you can really get taken aback by the scale of the thing.
I had my first experience with this in negotiating package shipping. I was dealing with UPS and FedEx, trying to get contracts in place and looking for better pricing. In Telecommunications I had really good success, with lots of players competing for my business. We were spending $50k per year (and growing) with FedEx, so I expected a little attention from our UPS rep when I started looking for competing quotes. But they were treating us like we weren't all that important.
When I told the story to my wife, she was really surprised. She dealt with UPS all the time and they were very responsive. Then she dropped the bomb.... "We have 3 reps from UPS on site in our warehouse"....
Uh, what? How much do you guys spend on UPS? "Oh, it depends.... three to five million dollars a month...".
HA! So that's what a big account looks like. Well.... no wonder they weren't knocking my door down.
Enterprise data is the same. If you "maintain your servers" all by yourself... you really don't understand enterprise. And even enterprise is peanuts next to the really big boys.
If you get the chance, everyone should go visit a real, hardened, world-class data center. It is amazing. Even fairly large corporate data centers are quaint in comparison. I got a tour of the NAP of the Americas in Miami as my first experience with this level of infrastructure. They spend more on power redundancy than most large corporate data centers will spend in total. They have these huge flywheels to handle the intermediate load while their enormous diesel generators crank up, instead of batteries. They have fully redundant, 2x(N+1) cooling and air handling systems. Each floor is independent and redundant. Security is beyond the pale. These are the locations that hold things like the root servers to the internet DNS system. They are not your "but we have a generator and a raised floor" data center. They are built to "even if the city is destroyed" level. Very cool to see.
Yeah, he is right about that "single point of failure".
There are a huge number of companies that fall into that hole between "too small to have our own IT department" and "big enough to have a large IT department with formal procedures and personnel redundancies".
I was that "one guy who does everything" for a long time. It is easily the most efficient way to run a company. Even when we were a small department with a half-dozen really sharp guys, it was still a high-wire act, dependent on having A players who knew the business and systems well.
It was only after we hit maybe 15 or so IT employees that we really had full redundancy in personnel and reasonably well documented procedures. And even that was pretty small... we didn't have dedicated QA employees, for example.
It all depends on what your needs are and what your risk tolerance is. For most growing small companies, the reward of having the "one really smart guy" as a key point of failure is worth the risk. Later, when it comes time to sell out to an investor or go public, that risk/reward equation flips and you have to figure out how to eliminate that red flag on the audit.
As many here have said, going "cloud" isn't a panacea to fix that problem.
There are a few examples of cars making it over 300k miles and even to 400k miles on the original battery with only minimal degradation. And there are others expressing astonishment at these numbers. I've seen plenty of discussions where 150k miles is supposed to land you with significant battery degradation.
I suppose there's a bit of speculation involved in the lower numbers and a bit of careful handling of the equipment in the higher numbers.
Still, it doesn't alter the conclusion, it just makes the numbers closer, with a break-even point of a half-million miles at an expensive price point for gas and zero cost electricity.
Or maybe this means he works in QA. Somebody has the keys to the kingdom.
Actually, they said he used other people's logins. So that's pretty interesting.
Maybe he's a network admin? Who else could gain access to accounts like that? Maybe the QA guys left their passwords on sticky notes?
Ok, maybe you are right. Maybe they need to examine their controls.. maybe "as actually implemented" instead of "as written on paper". You shouldn't be able to gain access to multiple people's logins.
Remember when one of SpaceX's rockets exploded, and they thought one of their nearby competitors sabotaged it?
No, I don't remember that at all.
I do remember that the engineers investigating the incident included sabotage in their fault tree. And I remember a couple of execs being asked specifically about that and saying that yes, it was being considered.
But no, I don't remember them actually thinking that a competitor sabotaged their rocket.
I do remember a lot of speculation on the internet about it. Lots of people reading things into random blips on video feeds. Lots of people opining that a well-placed rifle shot could trigger such an explosion. But "random guys on the net" is not the same thing as "SpaceX thought".
Slashdot Mirror
It's probably more mechanical turk than AI bot.
I had a similar thought upon seeing the demo presentation...
You could build that interaction with a Genesys IVR system.... heck, you could have built it 10 years ago... and it wouldn't have a lick of AI involved. It would be extremely cool and a great use of IVR, but it wouldn't be groundbreaking.
Building that interaction without any specific question/response trees and completely using AI to understand the conversation is a completely different animal and is amazing. It isn't clear exactly where this falls on the spectrum, nor is it clear that this is a voice assistant that goes beyond what google can do or is more widely available than google's (or Amazon, or Apple, or...)
I'm not saying it isn't either..... these are huge companies with huge customer bases and extremely large incentives to get natural language interactions automated in order to serve their customers better, so it stands to reason that they'd make this sort of breakthrough. But that demo wasn't really proof of anything earth shattering in AI. The fact that these are very task-specific agents suggests that they are closer to a very sophisticated IVR than they are to a complete natural language AI that can handle a wide array of requests in natural language.
In fact, the article pretty much says this... they have "shipping agents" that handle scheduling package delivery. And "scheduling agents" that they are working on for making reservations.
Building this with IVR technology would be a hard and complex task... but it doesn't require a full AI to do something like that demo.
Exactly my first reaction.
He is right about working in blocks. Some jobs are like that - you are more productive if you have long periods of interrupted time working on it with intense focus. There is a startup cost to a lot of work - putting it down and picking it up takes time. Analytic work like data mining, SQL programmer, writer, graphic artist, etc. would probably fit this model.
Other jobs are more intense in a different way, and reducing the consecutive hours actually increases productivity. Things like a factory line inspector or a data entry clerk or call center work might fall into this category.
It is almost as if a top-down, one-size-fits-all mandate isn't such a great idea.
Yeah, the Tesla Roadster is the vehicle that changed the perception of electric cars.
Before that they were slow, light and almost concept-car in appearance and finish - GM EV-1. Everyone was thinking super-econo-box.
The Roadster was specifically designed for the task of demonstrating that electric cars can be cool. And having ridden in one at the time they came out, I can confirm that it was an experience like no other. The huge amount of torque and acceleration made them immensely fun, despite their shortcomings.
Then they came out with the Model S, which demonstrated just how sporty and cool could be translated to luxury sedan and really cool.
The stated mission of Tesla was to act as a demonstration of how electric vehicles that would appeal to a mass market could actually be built. I'd say they accomplished that mission in spades. Of course, once they got a little taste of success, they redefined their goals upward. But that doesn't mean they didn't already accomplish what they set out to do.
Just to underline the massive size of the BFS, I looked up the numbers.
The ISS has an internal pressurized volume of 931.57 m3. That's not the usable space, just the pressurized volume.
The BFR/BFS is supposed to be 825m3.
So one BFS is in the same ball park as the entire ISS in terms of pressurized volume. Smaller, but in the same general ballpark.
The BFS is pretty big. It could conceivably provide a less claustrophobic space for a week long pleasure cruise than the ISS.
Take a look at this rendering of the BFS docked to the ISS for a sense of the scale of the thing.
They have been comparing the size of it to an A380 airplane (which seats 800). Were they to decide to send a tourist trip with 10 people aboard, the passengers would likely feel less cramped than the ISS astronauts. Estimating the cost of launch is tough, with the high-reusability goal designed to bring the costs down into the single digit millions per launch. So let's say they can sell it for $100 million - that's only 10 million per passenger. 10 passengers in that large ship would be pretty luxurious, as spaceflight goes.
As for the time... yeah, you can't cut the ride short. But there's plenty of vacation spots here on earth where people are happy to go knowing that rescue is days away at the earliest.
One man's gentrification is another man's urban renewal.
To me they're fighting the wrong battle - they should be fighting for higher wages so that as gentrification occurs they're not pushed into ghettos but move laterally, maintaining their standard of living.
That is absolutely the wrong direction to push.
If you want to keep rents (and housing prices) affordable, you have to remove barriers to affordable (and not affordable) housing.
Where are housing prices the highest? Cities like San Francisco, New York City, etc. that have strong controls in place that make developing new housing very difficult.
If you removed these restrictions and let developers build loads of new, expensive, luxury housing, you'll find that the value of the old housing stock won't appreciate like it does if you restrict these developers. In fact, you will likely find that the older, less luxuriously appointed housing drops in price.
There are some pretty strong forces that oppose looser zoning restrictions - including existing homeowners who would be very happy to have their buildings double and triple in value. But the ones that make absolutely no sense are folks who think they are advocating for more affordable housing by advocating rent controls and set-asides for affordable housing in any new project. You don't get lower prices by limiting supply and capping prices. You get shortages.
I doubt they deleted the app. Because most people buy their phones from the carriers, and Facebook comes as a part of the bloatware package that you can't delete at all. At least not without root access... and these days even most slashdot readers don't root their phones.
You would enjoy journal club.
In academic circles it is commonplace to have a journal club organized around your specialty. A group of researchers and graduate students meets to discuss journal articles relevant to their field. Sometimes it is exciting because you get to talk about someone opening a new area of inquiry or finding an interesting new fact. Other times it is exciting because you pick apart a paper that is completely unsupported by the results included in the paper.
Those are actually the best. There are a lot of articles that get published that never should have been published... even by the most prestigious journals.
Journal club is perhaps the best place to learn how to understand good experimental design and clear reasoning based on results.
The best club I was ever in did it in a very interesting way. We would bring in cutouts of just the materials and methods and the results. Then everyone had to analyze the results and draw our own conclusions as a group. Only then would we unveil the paper's conclusions section. It made for some very clear-headed analysis. Very often the authors of a paper don't include enough data to support their conclusions, even if they actually do have other results (not published) that further bolster their conclusions.
Anyway, the whole process is a lot of fun and it often looks just like carysub's take above.
. One unfair block and you're out forever?
Do they really think that's a good idea?
Yes. Yes they do.
Because they know that they are on the side of everything that is true and just and right. And they know that nobody will ever exercise this control who isn't true and just and right, just like they are. It is impossible that someone could ever decide that their ideas were unacceptable. Because reality has a liberal bias. And they are on the right side of history. You'll see....
I'm really surprised at the slashdot take on this.
We had this discussion here a long time ago. It was argued endlessly and Slashdot as a community was strongly convinced that all information should be free. And online services are not responsible for user content. Remember that? It was very, very important that online services like Slashdot avoid getting in the business of taking editorial control over user generated content. Because right-wing types kept threatening to either regulate offensive speech or hold internet companies liable for offensive and/or slanderous content.
Apparently the Slashdot community has collectively lost their minds. Are we really forgetting the "hands off" tradeoff that was made to keep these companies exempt from being held responsible for user-posted content?
And yet here we are, with politicians loudly agreeing that the big user-generated-content companies should exercise editorial control over the content on their platform. In many cases these are the exact same politicians who crafted the policy that allowed the internet to have Facebook, Twitter and Youtube. People like Chuck Schumer, Nancy Peloci, etc. were there when the policy was created. Their fingerprints are all over it, so they know full well that deciding which political views are acceptable negates the deal, opening these platforms up to liability for the user generated content they publish.
When they started with their "truth commission" ideas back in 2015/16 I thought this was obvious. It should be painfully obvious by now. But somehow the passage of time has allowed everyone to forget the terms of the deal. I wonder if they'll rue the day they chose to follow this path, or if everyone will somehow forget the deal and craft a new world where you can exercise editorial control over content and yet not be liable for that content.
Copyright protected images retain their copyright? Even if someone publishes them on a website that is accessible to the public?
I'm shocked at this disturbing turn of events. Next they'll be claiming that news articles retain their copyright even after they've been published online!
How is this even news? We have a lot of anti-IP folks around here, but even they have to acknowledge that this has been settled law... well, as long as there has been an internet. Longer, I guess. I mean, you couldn't just cut a photo out of a magazine and use it in your own ad copy either.
At the risk of feeding a troll.... Physics is not hands-on calculus.
You can use calculus as a tool for doing physics, just like you can use other mathematics for doing physics. But it is not physics.
I support your take, but I'll add that computer learning programs are already very good at basic skill building.
My kids do a couple of math programs - Reflex math is used to build specific skills in basic math operations. It measures how quickly and well they do and keeps feeding them practice until they master each "math fact". It moves them along at their own pace. So one kid might be working on 2 digit multiplication while the next kid is still trying to master subtraction.
They do the same for reading. It starts out really simply, but as they move forward the computer parses out the difference between reading, reading comprehension and various forms of advanced comprehension (reading between the lines). There is often a big gap between the level a student is mastering in the various areas of reading. The computer works to focus on their weaknesses while continuing to move their strengths forward. This doesn't work quite as well as the math, but it still is way, way ahead of our old readers and quiz packets.
Yeah, that was basically what I was about to write. Try "hands on" calculus. That really makes sense. And everyone should make a pilgrimage to Philadelphia when studying the constitution in history class. And Greece to learn about the history of western civilization. It is more engaging of the entire body to actually visit Sparta.
This opinion piece seems like it was written by someone who has no idea of how computers are being used in schools.
He's a poli-sci professor. I wonder how he engages the whole body in his sophomore political science class. Sure, you could act out democratic processes ... but exactly how would you make something meaningful and efficient about city-states, or monarchies? And political science is probably one of the easier ones.
Science and engineering have always been taught "hands on", and today they are increasingly "hands on" at the elementary and secondary levels, even as we introduce more computer-based learning.
Education today is becoming increasingly customized. Teachers used to move the entire class along as a block - maybe breaking off a couple of advanced kids for independent study. Today they have individual plans for each student and they juggle many groups simultaneously. Computers greatly aid this process. (the only downside being that kids and teachers spend way more of their time dealing with testing than they probably should.)
Computers have made math a much more "hands on" experience than it used to be. And they teach each kid at a customized pace, emphasizing areas of weakness until they have mastered important skills. It is still fairly primitive, but it is a major improvement over rows of kids reciting their times tables as a group.
it's not nothing to do with moral policing. Credit card transactions are effectively loans. In large parts of the world you have a legal right to dispute any charge on your card as a result. Adult content has a high percentage of disputes (probably from guys who's wives/girlfriends notice the charge). Even if you can prove the charge is valid it's still expensive to do so. Hence why nobody wants to be involved in it.
With corporations always, always, always follow the money. Anything bigger than a leomonade stand is completely amoral.
This is a great take.... and unfortunately it is just not correct.
This move is the direct result of Operation Choke Point.
The US Federal government began threatening banks back in 2013 that if they did business with disfavored industries, they risked being taken down by the Feds. It is often sold as being about "money laundering", but it targeted legal business that were in disfavor with the administration like firearms dealers, check cashing services and payday lenders. Along with this other groups were impacted like adult entertainers like porn actresses and producers.
These are perfectly legal businesses that the government decided they wanted to run out of business by threatening anyone who does business with them.
And that is why Patreon is even in the conversation. Because they can have trouble even getting bank accounts. So adult entertainers and others have been forcibly "unbanked". Now they have to hunt around for other ways to move money and get paid.
Even though the official program was recently ended, the effect lives on.
The last 18 years has seen a massive shift away from civil liberties in many ways (USA PATRIOT Act, warrantless wiretapping, metadata collection, etc.) and things like Operation Choke Point have changed the culture to the point where people actually see this sort of thing as acceptable. It isn't.
No, he's just using an example that people might be familiar with.
There are many niche packages out there that used to be sold as software that you'd have a VAR set up at your site. Then they started offering a hosted service. Then they started offering "cloud based" service.
And now they've started discontinuing the "you set up your own server in your datacenter" version.
I'm sure most of them would still negotiate with you and set it up for a price, but the business model is definitely moving to hosted services. Support is just so much cheaper if you don't have to deal with hundreds of different individual deployments with their own hardware and software idiosyncrasies.
At least with AWS or Azure, you can get a real human to talk to when shit goes wrong, you just have to pay for the privilege. Not sure Google even has the concept of customer service, but their competitors do.
I think this is the salient point.
If you are dealing with an enterprise service, you have a dedicated customer support rep and team. And you pay for this.
Getting the "google chat was not available" problem means that you didn't have an enterprise level contract.
It depends on how much you are losing.
Law firms are notorious for being behind the times on tech (and most other business needs).
Once you get to be a midsized company, most companies will have generators to keep the doors open in an extended outage.... at least if you are located in an area where this is a reasonable possibility.
I was with a company in hurricane territory, so I not only developed fully redundant data centers across 3 locations (one being a colo provider), we also had a huge generator in place to run most of the company in extended power outages. It worked great when we were out of power for 2 weeks after a hurricane. I also had contracts in place for emergency office space for critical employees if the building were to be destroyed.
But our costs were over a quarter million a day if we were down (not including payroll). And if we were down for more than 4 days we risked violating contracts that would have placed tens or hundreds of millions at risk. So the cost of setting up a $120k generator system wasn't that big of a problem.
It all depends on what your costs of down time are. For BankAmerica or AT&T, there can be no downtime, so they pay through the nose for super-hardened systems. For a 2 man law firm, a battery backup that shuts the system down while they go hit the links waiting for the power to come back on might be sufficient.
It is funny to read the "puny $6k /mo" denigration. When you are not exposed to these enterprise level vendors, you can really get taken aback by the scale of the thing.
I had my first experience with this in negotiating package shipping. I was dealing with UPS and FedEx, trying to get contracts in place and looking for better pricing. In Telecommunications I had really good success, with lots of players competing for my business. We were spending $50k per year (and growing) with FedEx, so I expected a little attention from our UPS rep when I started looking for competing quotes. But they were treating us like we weren't all that important.
When I told the story to my wife, she was really surprised. She dealt with UPS all the time and they were very responsive. Then she dropped the bomb.... "We have 3 reps from UPS on site in our warehouse"....
Uh, what? How much do you guys spend on UPS? "Oh, it depends.... three to five million dollars a month...".
HA! So that's what a big account looks like. Well.... no wonder they weren't knocking my door down.
Enterprise data is the same. If you "maintain your servers" all by yourself... you really don't understand enterprise. And even enterprise is peanuts next to the really big boys.
If you get the chance, everyone should go visit a real, hardened, world-class data center. It is amazing. Even fairly large corporate data centers are quaint in comparison. I got a tour of the NAP of the Americas in Miami as my first experience with this level of infrastructure. They spend more on power redundancy than most large corporate data centers will spend in total. They have these huge flywheels to handle the intermediate load while their enormous diesel generators crank up, instead of batteries. They have fully redundant, 2x(N+1) cooling and air handling systems. Each floor is independent and redundant. Security is beyond the pale. These are the locations that hold things like the root servers to the internet DNS system. They are not your "but we have a generator and a raised floor" data center. They are built to "even if the city is destroyed" level. Very cool to see.
Yeah, he is right about that "single point of failure".
There are a huge number of companies that fall into that hole between "too small to have our own IT department" and "big enough to have a large IT department with formal procedures and personnel redundancies".
I was that "one guy who does everything" for a long time. It is easily the most efficient way to run a company. Even when we were a small department with a half-dozen really sharp guys, it was still a high-wire act, dependent on having A players who knew the business and systems well.
It was only after we hit maybe 15 or so IT employees that we really had full redundancy in personnel and reasonably well documented procedures. And even that was pretty small... we didn't have dedicated QA employees, for example.
It all depends on what your needs are and what your risk tolerance is. For most growing small companies, the reward of having the "one really smart guy" as a key point of failure is worth the risk. Later, when it comes time to sell out to an investor or go public, that risk/reward equation flips and you have to figure out how to eliminate that red flag on the audit.
As many here have said, going "cloud" isn't a panacea to fix that problem.
That's a pretty interesting link.
There are a few examples of cars making it over 300k miles and even to 400k miles on the original battery with only minimal degradation. And there are others expressing astonishment at these numbers. I've seen plenty of discussions where 150k miles is supposed to land you with significant battery degradation.
I suppose there's a bit of speculation involved in the lower numbers and a bit of careful handling of the equipment in the higher numbers.
Still, it doesn't alter the conclusion, it just makes the numbers closer, with a break-even point of a half-million miles at an expensive price point for gas and zero cost electricity.
Yeah, I think in British courts even the truth is not always a defense.
Or maybe this means he works in QA. Somebody has the keys to the kingdom.
Actually, they said he used other people's logins. So that's pretty interesting.
Maybe he's a network admin? Who else could gain access to accounts like that? Maybe the QA guys left their passwords on sticky notes?
Ok, maybe you are right. Maybe they need to examine their controls.. maybe "as actually implemented" instead of "as written on paper". You shouldn't be able to gain access to multiple people's logins.
Remember when one of SpaceX's rockets exploded, and they thought one of their nearby competitors sabotaged it?
No, I don't remember that at all.
I do remember that the engineers investigating the incident included sabotage in their fault tree. And I remember a couple of execs being asked specifically about that and saying that yes, it was being considered.
But no, I don't remember them actually thinking that a competitor sabotaged their rocket.
I do remember a lot of speculation on the internet about it. Lots of people reading things into random blips on video feeds. Lots of people opining that a well-placed rifle shot could trigger such an explosion. But "random guys on the net" is not the same thing as "SpaceX thought".