There is no free lunch. If someone spent significant amounts of money to make it - you're going to pay for it one way, or the other.
This is where creators need to get creative: there isn't one single model that can do that - there are a number of ways, including patronage (e.g. Patreon.com , kickstarter.com ), paywalls (e.g. WSJ.com , washingtonpost.com, netflix.com and the like), and ad supported content (e.g. YouTube.com etc) to name a few. There are other models too - such as 'name your own prices' schemes and the like.
If you want a consistently high quality product - most of us are going to have to pay one way or the other, otherwise it will go away because the people doing it won't be able to continue giving away high quality free content forever. I've seen this happen a number of times over the years.
The corollary to that is you get what you pay for. There is going to be free content out there - but its quality is going to be commensurate with the person's time and effort put into it, so unless they are independently wealthy they will not be able to put in the amount of effort, or hire the necessary people to produce high quality work on an ongoing basis. There may be exceptions to this - but they will be rare indeed. Good on you if you find such an artist.
I'm forced more and more to multi-task, and have a wider range of choices to make in any given day on the job. This has increased the overall output only slightly - primarily because my work requires research to get to the bottom of many questions - and has certainly eroded the quality of that output immensely - forcing 2nd passes across some items that are in error.
I think sensory overload in all forms is a bad thing for human beings - regardless of their age.
The real problem is not how to manage life expectancy in the event of an accident. The real problem is the fact that living in post modern industrialized cities is killing us.
RazorSharp, you need to qualify that statement. You don't know enough about his experience and learning over the course of his lifetime. Hubris leads to nemesis.
I used to be able to hire candidates without degrees based upon other factors, such as experience (e.g. in the 1990s). Today I don't have that option anymore - of no fault of my own. It's not a matter of letting HR do the screening. It is a matter of company policy as established by the executive board in conjunction with HR best practices. Someone may be a moron in all of this, but it's not the hiring manager.
The L40 looks like it would have been a good typer. I never got to play with one of those.
My next favorite laptop I still use today is an Apple Macbook Pro circa 2006. It is the first Intel processor Mac - a 2.16 GHz Core Duo, and its keyboard has scalloped backlit keys that have decent spacing and action when compared to other modern laptops. When Apple went to the chicklet style keys, I wasn't very happy, and was almost relieved when the video went out on my updated Pro. After replacing a delaminated battery in the 2006 model, I was back in business - albeit slower processing than the Intel i7 equipped model. I only use if for writing/documentation - so it's fine for what I use it for.
When I'm not mobile I have two keyboards I like to use. The first is a Corsair gaming keyboard - with Cherry key-switches - extremely good action (much like an IBM model M keyboard), and when I start to get a twinge of carpal-tunnel coming on I switch to my ergonomic Maxim keyboard - which allows me to adjust the split and angle of the keys.
Having a good keyboard is important, not only for performance, but also for longevity.
In my opinion, the best keyboard on any laptop computer was the Toshiba T1200 keyboard circa 1988. Very much like the IBM Selectric / M-series buckling spring switch keyboards - in a portable form factor. I have yet to find another laptop keyboard that can match the feel and speed of typing possible with that machine. You could beat hell out of it, and it would just keep going. You can read about the T1200 laptop and see images of it here.
In 1988 I was in college, and I want to say I was one of the first, if not first person to use a laptop computer for taking notes in class on my campus (I didn't see ubiquitous laptop use in school until the 1990s). There were PC and Unix workstations and kiosks on campus - but they of course were not portable by any stretch of the imagination.
As for current machines - nothing has come close in terms of keyboard ergonomics.
You hit upon the real problem: Companies put more focus on the bottom line, than doing what is right for their customers. Hence operating with minimal IT workforce, and resorting to off-shoring and other cost saving methods that directly impact their ability to deliver quality code, and more importantly keep it updated to avoid zero day exploits (as studies have found most zero day exploits take 6 months to a year to find and a fix to be coded, yet the average time for systems in the wild to be updated is 3 to 5 years). IT should know every piece of code that is placed in the network and its source.
So, what's the fix, aside from reforming corporation and stock market rules? Corporations need to know that if they don't take security seriously there will be bad outcomes for them. Lawsuits are one mechanism for this. Another is through customer choices - boycott companies that don't take security seriously. For corporations that actually want to make changes to deal with this correctly, IT culture needs to change in the following ways:
* IT should know every piece of code that is placed in the network and its source. This means having an absolutely clear understanding of every library, framework, and any non-standard custom extensions deployed. This will serve two purposes. On the one hand it will ensure that IT is being proactive about patching to avoid zero day exploits. On the other hand it will drive simplification and good software engineering; another way of saying this is KISS (Keep It Simple, Stupid). The more complex systems you put into place - and more importantly the more that complexity comes from code that is generated outside of your own organization, the more likely there are for bugs (potentially exploitable zero days) to exist within the overall code base.
* IT costs need to be viewed as a cost of doing business, rather than something that can be dispensed with or minimized. To do security right takes resources, and this has increased relevance not only with breaches that we've seen happening, but also to meet corporate requirements from a legal and regulatory perspective (e.g. Sarbanes-Oxley). Costs can be managed, if companies are willing to invest in building automation to help them manage what they've got - and doing that first item above (weeding out overly complex designs).
* IT needs to also change their culture from what I call a 'shrink-wrapped' software mentality - where software is thrown over the wall to operations and the developers walk away and never work on it again, to a culture that values long term developer ownership and maintenance of systems they have created in partnership with operational teams. This is related to something else that I see a lot of in IT: brain drain. Basically, due to nomadic existence of developers in an organization either through rotation or vendor outsourcing, long term knowledge of integration between existing systems and new development is lost every year to 18 months - breaking the ability of the company to quickly patch or otherwise modify systems in response to security issues or simply the need for responses to competitive forces.
We could transform IT from a necessary burden to a much needed and appreciated partner in business. But, that will require the decisiveness on the part of CTOs, and CEOs to dedicate resources to that specific mission.
What they are really trying to say is, "stop eating over your keyboard. You've got so much cruft in there, I'm surprised any of your keys work. Dried Pepsi, cat hair, and Cornflakes are not good for your keyboard."
It depends on what kind of business you are in. If you are a retail outlet - then using a web hosting service - or maybe even software as a service would be right for you. On the other hand, there will be some things that you probably don't want to trust to the cloud - in particular accounts, passwords, and documents defining your own proprietary information. Those things I would keep on an air-gapped system....or on paper.
Cloud is clearly an attempt to lower headcount costs. The concept is converting the network from physical to virtual and banking on end-state automation (network elasticity, dynamic scalability, geographical redundancy and the like) to make a stronger and cheaper network to run.
Just like all snake oil - the promises have outpaced the reality due to a number of factors, including the assumption that all problems are suitable for the original design, coupled with lack of engineering when it comes to planning the network, compute, and storage for edge cases - such as near-real-time systems - e.g. voice, telemetry, and the like. Another issue is the assumption that an application that is built and tuned to run on dedicated network and computing resources, can be easily migrated as-is to the new virtual infrastructure without integration and modification to work with the performance and resource limitations imposed by the cloud infrastructure.
Are there applications for cloud? Certainly. But every application is not appropriate for cloud, and naive attempts to throw everything at the cloud will not have the anticipated savings due to more complex trouble-shooting needed when failures occur.
That's a political problem of "How do we get rid of the idiots in a democracy where NOBODY is going to vote for the Nazi Party and related eugenics program?"
I hope you were joking about the NAZIs' eugenics program as a solution. Assuming you were trying to make a point:
Quality education is the key. Also define 'Idiot'. Is that someone who disagrees with your point of view? Is it someone with a given measurable intelligence? Or perhaps you're considering people who parrot what their tribe says on a subject, without using critical thinking skills? Seeing 'idiots' all around us is not a problem, as much as a symptom of out of control tribalism.
The biggest problem in the world today is rampant tribalism: the various groups see enemies at every turn, and the Internet is only amplifying this polarization between all factions as like minded people find sustenance in echo chambers. No one is willing to have empathy for anyone not in their group. The boundaries of these groups are religious, ethnic/race, gender, sexuality, economic, and political. Segregation, and ultimately destruction of the 'other' is the goal - assuming they cannot be reeducated and absorbed by the group. While the Internet has had positive impacts, the echo chamber effect is leading us towards 'Lord of the Flies' writ large.
The solution is for everyone to stop viewing each other as enemies in a life and death struggle, and instead seek empathy, acceptance, and true understanding. This requires people to leave their echo chambers and make an effort to moderate their extreme views, set that aside, and really connect with other people without judgment.
Many groups have made empathy and moderation a dirty word - so the odds of them moderating their divisive activities are slim to none. Ultimately our very survival as a species on this planet is in the balance.
Encyclopedia: a book, often in many volumes, containing articles on various topics, often arranged in alphabetical order, dealing either with the whole range of human knowledge or with one particular subject: a medical encyclopedia
Wikipedia takes the place of an Encyclopedia - which by definition is intended for the widest possible audience. The main purpose of an Encyclopedia is to provide introductory material to the widest range of topics. While you may have deeper information available, the primary role of providing understandable introductions to the topics must not be optional. If you are really interested in depth of research, then why not link the main articles in Wikipedia, with other wiki based subject focused sites, or research paper repositories?
The fact that there is a simple.wikipedia.org, speaks volumes to the breakdown of Wikipedia's primary mission. simple.wikipedia.org should not need to exist.
That's not what this is about. An encyclopedia is supposed to contain introductory material that covers a topic in a comprehensible way for anyone. As a minimum, the article should have this introductory material. It is not optional. As has been stated perfectly by ebyrob above: "There's a difference between a difficult subject and obfuscation for a pretense of erudition."
Perhaps the idiots are the people who don't know how to communicate effectively.
Fully agree. There are too many specialized technical microcosms that forget what they do involves others outside of their little fiefdom. It takes real talent to both deal with the jargon among peers and communicate that jargon to others effectively. Few people know how to do this well, and the ones that refuse to try shouldn't be viewed as being any smarter then the rest of us when they hide behind their jargon.
I don't think it has to be 'reputable' sources - any sources should be held to account for what they say. If anyone fabricates information that impacts anyone negatively, then the punishment should fit the size of the impact of the crime. You swindle 1 person out of $100 - small claims court. If you manipulate millions of people and impact the fabric of the nation negatively - perhaps with people dieing as a result - then death for the perpetrator might not be too high a price to pay.
Here is a little heuristic to use when looking at claims online (regardless of content):
If it's too good to be true, assume that it is not true.
Be skeptical about anything important to you - but more importantly seek out the truth for those things that you care about. And I'm talking about objective scientific truth - not what your cousin down the street told you, because he has a friend of a friend who works in the top secret facility where they are making the Soylent Green....
And now I know where the phrase "beat a dead horse" comes from. I never imagined it was from people wanting to beat or "execute" a corpse.
Wrong. It is not about trying to execute a corpse.
The phrase comes about from the idea of getting a dead horse to move. No matter how much you may whip/beat the dead horse, it isn't going anywhere.
In general use it means you're talking about a subject that isn't going to change, e.g. "If you want to get gun control laws changed, you're beating a dead horse; the NRA and Republicans have that all tied up."
And airplanes will never be more than the proof of concept made by the Wright brothers. There's no way those principles will scale to anything that is economically feasible.
And to take that a step further - licensed commercial drone pilots (Unmanned Aircraft Systems Pilots per FAA terminology) already outnumber licensed aircraft pilots in the United States as of FAA data released in February 2016.
A group of bacteria were growing in a petry dish. Their civilization lasted for 60 days. Every day they doubled their numbers. On the 59th day some bacteria were trying to make a point that the food is finite that this level of growth is unsustainable....You understand that we can make machines run the whole of the economy. ALL of it. Forever, because AI learns faster so if we create a new job, ANY new job, AI will be able to do it better and cheaper...understand? All future jobs are gone as well. From that moment we either enter the Aurora situation - 1 human per 100 km2 served by thousands of robots or we kill each other since there is no "profit" in such future....
Your petri dish analogy is flawed. It doesn't take into account the lack of farming of arable land all over the world (e.g. everyone's back yard, and other surfaces in urban environments). We are nowhere near the 59th day in the petri dish.
As for your AI assertion - AI is not able to duplicate human capacities in a general purpose sense. Most AI is tethered to the network - requiring large arrays of computers running in parallel to do the job and is focused on a specific use case/world model. This definitely wouldn't be contained on the robot itself. Given a future of more frequent and violent natural disasters (that break the network), AI is not going to be as ubiquitous as you think. Without connectivity to the network, your AI is not very smart - particularly when its environment is altered beyond the model it can understand.
Slashdot Mirror
Speak for yourself. I own everything on my systems. Of course, I don't use cloud services to store anything, and backup my own data.
The old tech still works...most people just don't know how to use it.
There is no free lunch. If someone spent significant amounts of money to make it - you're going to pay for it one way, or the other.
This is where creators need to get creative: there isn't one single model that can do that - there are a number of ways, including patronage (e.g. Patreon.com , kickstarter.com ), paywalls (e.g. WSJ.com , washingtonpost.com, netflix.com and the like), and ad supported content (e.g. YouTube.com etc) to name a few. There are other models too - such as 'name your own prices' schemes and the like.
If you want a consistently high quality product - most of us are going to have to pay one way or the other, otherwise it will go away because the people doing it won't be able to continue giving away high quality free content forever. I've seen this happen a number of times over the years.
The corollary to that is you get what you pay for. There is going to be free content out there - but its quality is going to be commensurate with the person's time and effort put into it, so unless they are independently wealthy they will not be able to put in the amount of effort, or hire the necessary people to produce high quality work on an ongoing basis. There may be exceptions to this - but they will be rare indeed. Good on you if you find such an artist.
Words can be infectious to a population that isn't taught to be critical thinkers.
I'm forced more and more to multi-task, and have a wider range of choices to make in any given day on the job. This has increased the overall output only slightly - primarily because my work requires research to get to the bottom of many questions - and has certainly eroded the quality of that output immensely - forcing 2nd passes across some items that are in error.
I think sensory overload in all forms is a bad thing for human beings - regardless of their age.
Budda tried that - and he found you need balance in your life - the middle way. All things in moderation.
The real problem is not how to manage life expectancy in the event of an accident. The real problem is the fact that living in post modern industrialized cities is killing us.
RazorSharp, you need to qualify that statement. You don't know enough about his experience and learning over the course of his lifetime. Hubris leads to nemesis.
I used to be able to hire candidates without degrees based upon other factors, such as experience (e.g. in the 1990s). Today I don't have that option anymore - of no fault of my own. It's not a matter of letting HR do the screening. It is a matter of company policy as established by the executive board in conjunction with HR best practices. Someone may be a moron in all of this, but it's not the hiring manager.
The L40 looks like it would have been a good typer. I never got to play with one of those.
My next favorite laptop I still use today is an Apple Macbook Pro circa 2006. It is the first Intel processor Mac - a 2.16 GHz Core Duo, and its keyboard has scalloped backlit keys that have decent spacing and action when compared to other modern laptops. When Apple went to the chicklet style keys, I wasn't very happy, and was almost relieved when the video went out on my updated Pro. After replacing a delaminated battery in the 2006 model, I was back in business - albeit slower processing than the Intel i7 equipped model. I only use if for writing/documentation - so it's fine for what I use it for.
When I'm not mobile I have two keyboards I like to use. The first is a Corsair gaming keyboard - with Cherry key-switches - extremely good action (much like an IBM model M keyboard), and when I start to get a twinge of carpal-tunnel coming on I switch to my ergonomic Maxim keyboard - which allows me to adjust the split and angle of the keys.
Having a good keyboard is important, not only for performance, but also for longevity.
In my opinion, the best keyboard on any laptop computer was the Toshiba T1200 keyboard circa 1988. Very much like the IBM Selectric / M-series buckling spring switch keyboards - in a portable form factor. I have yet to find another laptop keyboard that can match the feel and speed of typing possible with that machine. You could beat hell out of it, and it would just keep going. You can read about the T1200 laptop and see images of it here.
In 1988 I was in college, and I want to say I was one of the first, if not first person to use a laptop computer for taking notes in class on my campus (I didn't see ubiquitous laptop use in school until the 1990s). There were PC and Unix workstations and kiosks on campus - but they of course were not portable by any stretch of the imagination.
As for current machines - nothing has come close in terms of keyboard ergonomics.
You hit upon the real problem: Companies put more focus on the bottom line, than doing what is right for their customers. Hence operating with minimal IT workforce, and resorting to off-shoring and other cost saving methods that directly impact their ability to deliver quality code, and more importantly keep it updated to avoid zero day exploits (as studies have found most zero day exploits take 6 months to a year to find and a fix to be coded, yet the average time for systems in the wild to be updated is 3 to 5 years). IT should know every piece of code that is placed in the network and its source.
So, what's the fix, aside from reforming corporation and stock market rules? Corporations need to know that if they don't take security seriously there will be bad outcomes for them. Lawsuits are one mechanism for this. Another is through customer choices - boycott companies that don't take security seriously. For corporations that actually want to make changes to deal with this correctly, IT culture needs to change in the following ways:
* IT should know every piece of code that is placed in the network and its source. This means having an absolutely clear understanding of every library, framework, and any non-standard custom extensions deployed. This will serve two purposes. On the one hand it will ensure that IT is being proactive about patching to avoid zero day exploits. On the other hand it will drive simplification and good software engineering; another way of saying this is KISS (Keep It Simple, Stupid). The more complex systems you put into place - and more importantly the more that complexity comes from code that is generated outside of your own organization, the more likely there are for bugs (potentially exploitable zero days) to exist within the overall code base.
* IT costs need to be viewed as a cost of doing business, rather than something that can be dispensed with or minimized. To do security right takes resources, and this has increased relevance not only with breaches that we've seen happening, but also to meet corporate requirements from a legal and regulatory perspective (e.g. Sarbanes-Oxley). Costs can be managed, if companies are willing to invest in building automation to help them manage what they've got - and doing that first item above (weeding out overly complex designs).
* IT needs to also change their culture from what I call a 'shrink-wrapped' software mentality - where software is thrown over the wall to operations and the developers walk away and never work on it again, to a culture that values long term developer ownership and maintenance of systems they have created in partnership with operational teams. This is related to something else that I see a lot of in IT: brain drain. Basically, due to nomadic existence of developers in an organization either through rotation or vendor outsourcing, long term knowledge of integration between existing systems and new development is lost every year to 18 months - breaking the ability of the company to quickly patch or otherwise modify systems in response to security issues or simply the need for responses to competitive forces.
We could transform IT from a necessary burden to a much needed and appreciated partner in business. But, that will require the decisiveness on the part of CTOs, and CEOs to dedicate resources to that specific mission.
I think what happened is they got 1 terabyte in and realized that the data started to repeat over and over...and over.
What they are really trying to say is, "stop eating over your keyboard. You've got so much cruft in there, I'm surprised any of your keys work. Dried Pepsi, cat hair, and Cornflakes are not good for your keyboard."
It depends on what kind of business you are in. If you are a retail outlet - then using a web hosting service - or maybe even software as a service would be right for you. On the other hand, there will be some things that you probably don't want to trust to the cloud - in particular accounts, passwords, and documents defining your own proprietary information. Those things I would keep on an air-gapped system....or on paper.
Cloud is clearly an attempt to lower headcount costs. The concept is converting the network from physical to virtual and banking on end-state automation (network elasticity, dynamic scalability, geographical redundancy and the like) to make a stronger and cheaper network to run.
Just like all snake oil - the promises have outpaced the reality due to a number of factors, including the assumption that all problems are suitable for the original design, coupled with lack of engineering when it comes to planning the network, compute, and storage for edge cases - such as near-real-time systems - e.g. voice, telemetry, and the like. Another issue is the assumption that an application that is built and tuned to run on dedicated network and computing resources, can be easily migrated as-is to the new virtual infrastructure without integration and modification to work with the performance and resource limitations imposed by the cloud infrastructure.
Are there applications for cloud? Certainly. But every application is not appropriate for cloud, and naive attempts to throw everything at the cloud will not have the anticipated savings due to more complex trouble-shooting needed when failures occur.
That's a political problem of "How do we get rid of the idiots in a democracy where NOBODY is going to vote for the Nazi Party and related eugenics program?"
I hope you were joking about the NAZIs' eugenics program as a solution. Assuming you were trying to make a point: Quality education is the key. Also define 'Idiot'. Is that someone who disagrees with your point of view? Is it someone with a given measurable intelligence? Or perhaps you're considering people who parrot what their tribe says on a subject, without using critical thinking skills? Seeing 'idiots' all around us is not a problem, as much as a symptom of out of control tribalism.
The biggest problem in the world today is rampant tribalism: the various groups see enemies at every turn, and the Internet is only amplifying this polarization between all factions as like minded people find sustenance in echo chambers. No one is willing to have empathy for anyone not in their group. The boundaries of these groups are religious, ethnic/race, gender, sexuality, economic, and political. Segregation, and ultimately destruction of the 'other' is the goal - assuming they cannot be reeducated and absorbed by the group. While the Internet has had positive impacts, the echo chamber effect is leading us towards 'Lord of the Flies' writ large.
The solution is for everyone to stop viewing each other as enemies in a life and death struggle, and instead seek empathy, acceptance, and true understanding. This requires people to leave their echo chambers and make an effort to moderate their extreme views, set that aside, and really connect with other people without judgment.
Many groups have made empathy and moderation a dirty word - so the odds of them moderating their divisive activities are slim to none. Ultimately our very survival as a species on this planet is in the balance.
Wikipedia takes the place of an Encyclopedia - which by definition is intended for the widest possible audience. The main purpose of an Encyclopedia is to provide introductory material to the widest range of topics. While you may have deeper information available, the primary role of providing understandable introductions to the topics must not be optional. If you are really interested in depth of research, then why not link the main articles in Wikipedia, with other wiki based subject focused sites, or research paper repositories?
The fact that there is a simple.wikipedia.org, speaks volumes to the breakdown of Wikipedia's primary mission. simple.wikipedia.org should not need to exist.
That's not what this is about. An encyclopedia is supposed to contain introductory material that covers a topic in a comprehensible way for anyone. As a minimum, the article should have this introductory material. It is not optional. As has been stated perfectly by ebyrob above: "There's a difference between a difficult subject and obfuscation for a pretense of erudition."
Perhaps the idiots are the people who don't know how to communicate effectively.
That expresses perfectly what I see daily.
Fully agree. There are too many specialized technical microcosms that forget what they do involves others outside of their little fiefdom. It takes real talent to both deal with the jargon among peers and communicate that jargon to others effectively. Few people know how to do this well, and the ones that refuse to try shouldn't be viewed as being any smarter then the rest of us when they hide behind their jargon.
I don't think it has to be 'reputable' sources - any sources should be held to account for what they say. If anyone fabricates information that impacts anyone negatively, then the punishment should fit the size of the impact of the crime. You swindle 1 person out of $100 - small claims court. If you manipulate millions of people and impact the fabric of the nation negatively - perhaps with people dieing as a result - then death for the perpetrator might not be too high a price to pay.
Here is a little heuristic to use when looking at claims online (regardless of content):
If it's too good to be true, assume that it is not true.
Be skeptical about anything important to you - but more importantly seek out the truth for those things that you care about. And I'm talking about objective scientific truth - not what your cousin down the street told you, because he has a friend of a friend who works in the top secret facility where they are making the Soylent Green....
I think we need to execute Mr Stephen Paddock.
You're too late.
And now I know where the phrase "beat a dead horse" comes from. I never imagined it was from people wanting to beat or "execute" a corpse.
Wrong. It is not about trying to execute a corpse. The phrase comes about from the idea of getting a dead horse to move. No matter how much you may whip/beat the dead horse, it isn't going anywhere. In general use it means you're talking about a subject that isn't going to change, e.g. "If you want to get gun control laws changed, you're beating a dead horse; the NRA and Republicans have that all tied up."
And airplanes will never be more than the proof of concept made by the Wright brothers. There's no way those principles will scale to anything that is economically feasible.
And to take that a step further - licensed commercial drone pilots (Unmanned Aircraft Systems Pilots per FAA terminology) already outnumber licensed aircraft pilots in the United States as of FAA data released in February 2016.
Always....always...always....
A group of bacteria were growing in a petry dish. Their civilization lasted for 60 days. Every day they doubled their numbers. On the 59th day some bacteria were trying to make a point that the food is finite that this level of growth is unsustainable....You understand that we can make machines run the whole of the economy. ALL of it. Forever, because AI learns faster so if we create a new job, ANY new job, AI will be able to do it better and cheaper...understand? All future jobs are gone as well. From that moment we either enter the Aurora situation - 1 human per 100 km2 served by thousands of robots or we kill each other since there is no "profit" in such future....
Your petri dish analogy is flawed. It doesn't take into account the lack of farming of arable land all over the world (e.g. everyone's back yard, and other surfaces in urban environments). We are nowhere near the 59th day in the petri dish. As for your AI assertion - AI is not able to duplicate human capacities in a general purpose sense. Most AI is tethered to the network - requiring large arrays of computers running in parallel to do the job and is focused on a specific use case/world model. This definitely wouldn't be contained on the robot itself. Given a future of more frequent and violent natural disasters (that break the network), AI is not going to be as ubiquitous as you think. Without connectivity to the network, your AI is not very smart - particularly when its environment is altered beyond the model it can understand.