953 replies to this topic

[Mind]

Don't get your hopes up too high. It is likely that governments will ban or limit the use of highly-functional personal robots. The government will make up a bunch of reasons, safety, security, etc....

[adamh]

Mag1 wrote:

 

"with domestic tasks if you cut the lawn once you have pretty much got the task down. There is minimal change in the environment from one cutting to the next. This would apply to most features of household chores "

 

Yes and factory work will be even easier since nothing changes from one job to the next. They already have roombas to do vacuuming and I've seen robot mowers but they are still primitive. For the home, besides ordinary household tasks, the security features will be important. They will watch against fire, flood, burglars and other potential problems. 

 

Home security services will get upgraded. With ai, a bot will be able to tell the difference between a minor and major issue. In addition to summoning help and alerting the owner or occupant, it may be able to put out minor fires by itself and can tell firefighters exactly where the fire is. If a pipe bursts and floods the home, the bot could temporarily shut off the water

 

Most of those functions except cleaning and inspecting areas do not require movement. The brains of the unit will be a computer giving directions to the specialized bots which could do the moving and manipulating things. No need for a bot that looks and sounds human. That just makes it more difficult. And no, the govt will not make them illegal 

[QuestforLife]

There is a more general point here (regarding robots doing menial tasks),  (as I discussed earlier in this thread) that menial/unskilled tasks seem to have large escaped automation so far, whereas some skilled jobs are coming under threat. Perhaps with the improvement in domestic robots, we'll see unskilled work also getting  gradually replaced. 

 

I don't know what it is like in other countries, but in the UK we have the weird situation where a professional is hiring in electricians, plumbers, even gardeners, labourers and cleaners, who all earn as much (or more) than him.  This has been driven by the scarcity of people wanting to work in those roles and other things like a bottleneck on the supply of new houses, driving up the cost of home improvements. Can automation alleviate this? Or would that be a 'bad thing' if it destroys one of the last ways someone without decades of education can earn a decent living?  

[Mind]

There is a more general point here (regarding robots doing menial tasks),  (as I discussed earlier in this thread) that menial/unskilled tasks seem to have large escaped automation so far, whereas some skilled jobs are coming under threat. Perhaps with the improvement in domestic robots, we'll see unskilled work also getting  gradually replaced. 

 

I don't know what it is like in other countries, but in the UK we have the weird situation where a professional is hiring in electricians, plumbers, even gardeners, labourers and cleaners, who all earn as much (or more) than him.  This has been driven by the scarcity of people wanting to work in those roles and other things like a bottleneck on the supply of new houses, driving up the cost of home improvements. Can automation alleviate this? Or would that be a 'bad thing' if it destroys one of the last ways someone without decades of education can earn a decent living?  

 

Same situation in the US. People who do the hard work that holds society together are getting paid amazing amounts of money. Stone masons around my area can easily make $50-100 per hour, many times more. This is work that is not yet easily replaceable with robots. What work is replaceable with current AI? Everything digital, like programming, coding, digital media, a lot of legal work, etc...

[Danail Bulgaria]

don't count on the mental unskilled jobs and tasks. They are not taken AT THE MOMENT.

 

But their removal is also possible, and has never stopped being under development in many AI and non AI projects. You were giving an example with jobs in the field of building and construction.

 

 

By the way, stumbled upon today at an interesting advertisement:

 

[Mind]

don't count on the mental unskilled jobs and tasks. They are not taken AT THE MOMENT.

 

But their removal is also possible, and has never stopped being under development in many AI and non AI projects. You were giving an example with jobs in the field of building and construction.

 

 

By the way, stumbled upon today at an interesting advertisement:

 

 

Robots will find their way into the trades - no doubt, just not as soon and AI will replace coders, programmers, game developers, movie-makers, digital artists etc...

[Danail Bulgaria]

Sounds logical. We will see. Time will show.

[adamh]

Robots are doing lots of work right now, you just don’t think of them as robots. They don’t usually have human features. You may have a few in your home

[Mind]

I consider bots to be traditional robots with a physical embodiment - something that can move and react to the environment. I don't consider a smart refrigerator or TV to be a bot, just an appliance with a a more sophisticated computer program.

 

I only have one bot - the Turtle - a product which I inspired.

[Advocatus Diaboli]

Define "physical embodiment". Mind, your personal definition may not correspond to common definitions of what properties to ascribe to a robot.

 

Wikipedia:

 

"A robot is a machine—especially one programmable by a computer—capable of carrying out a complex series of actions automatically.^[2] A robot can be guided by an external control device, or the control may be embedded within. Robots may be constructed to evoke human form, but most robots are task-performing machines, designed with an emphasis on stark functionality, rather than expressive aesthetics."

 

Examples (not from wiki)

 

1. Robotic Vacuum Cleaners

Description: These are autonomous devices designed to clean floors without human intervention.

Examples: The iRobot RoombaiRobot Roomba is the most recognized model, but others include the Roborock S7 and Ecovacs Deebot. Many models now combine vacuuming and mopping functions, enhancing their utility .

 

2. Floor Cleaning Robots

Types:

Sweeping Robots: Such as the iRobot Dirt Dog, designed for dry cleaning.

Washing Robots: Like the iRobot Scooba, which can wet mop floors .

 

3. Lawn Mowing Robots

Description: These robots automate lawn care by mowing grass.

Examples: Models like the Husqvarna Automower are popular for residential use, capable of handling various lawn sizes and terrains .

 

4. Window Cleaning Robots

Function: Designed to clean windows autonomously, these robots navigate surfaces to remove dirt and grime.

Example: The WINBOT is a notable model that provides streak-free cleaning results .

 

5. Security Robots

Purpose: These robots monitor homes or properties, often equipped with cameras and sensors.

Example: The Knightscope, which can patrol areas and send alerts about suspicious activities .

 

6. Companion Robots

Role: Designed to provide companionship, especially for the elderly or those living alone.

Examples: The Wakamaru, developed by Mitsubishi, and the more recent ElliQ,

which oers conversational interaction and health monitoring .

 

7. Litter Cleaning Robots

Functionality: These self-cleaning litter boxes automate the process of managing pet waste.

Example: The Litter-Robot, which detects when a cat has used it and cleans itself afterward .

 

8. Kitchen Robots

Description: These robots assist in meal preparation or cooking tasks.

Examples: Devices like the Rotimatic, which can make various types of bread, are becoming more common in kitchens .

 

Edited by Advocatus Diaboli, 22 October 2024 - 07:50 PM.

[dlewis1453]

 

Define "physical embodiment". Mind, your personal definition may not correspond to common definitions of what properties to ascribe to a robot.

 

 

Lolol who is this guy? 

[Advocatus Diaboli]

Mind is Director, Moderator, and Treasurer at Longecity. (hint, it clearly so states under his pic)

[Mind]

 

Define "physical embodiment". Mind, your personal definition may not correspond to common definitions of what properties to ascribe to a robot.

 

Wikipedia:

 

"A robot is a machine—especially one programmable by a computer—capable of carrying out a complex series of actions automatically.^[2] A robot can be guided by an external control device, or the control may be embedded within. Robots may be constructed to evoke human form, but most robots are task-performing machines, designed with an emphasis on stark functionality, rather than expressive aesthetics."

 

Examples (not from wiki)

 

1. Robotic Vacuum Cleaners

Description: These are autonomous devices designed to clean floors without human intervention.

Examples: The iRobot RoombaiRobot Roomba is the most recognized model, but others include the Roborock S7 and Ecovacs Deebot. Many models now combine vacuuming and mopping functions, enhancing their utility .

 

2. Floor Cleaning Robots

Types:

Sweeping Robots: Such as the iRobot Dirt Dog, designed for dry cleaning.

Washing Robots: Like the iRobot Scooba, which can wet mop floors .

 

3. Lawn Mowing Robots

Description: These robots automate lawn care by mowing grass.

Examples: Models like the Husqvarna Automower are popular for residential use, capable of handling various lawn sizes and terrains .

 

4. Window Cleaning Robots

Function: Designed to clean windows autonomously, these robots navigate surfaces to remove dirt and grime.

Example: The WINBOT is a notable model that provides streak-free cleaning results .

 

5. Security Robots

Purpose: These robots monitor homes or properties, often equipped with cameras and sensors.

Example: The Knightscope, which can patrol areas and send alerts about suspicious activities .

 

6. Companion Robots

Role: Designed to provide companionship, especially for the elderly or those living alone.

Examples: The Wakamaru, developed by Mitsubishi, and the more recent ElliQ,

which oers conversational interaction and health monitoring .

 

7. Litter Cleaning Robots

Functionality: These self-cleaning litter boxes automate the process of managing pet waste.

Example: The Litter-Robot, which detects when a cat has used it and cleans itself afterward .

 

8. Kitchen Robots

Description: These robots assist in meal preparation or cooking tasks.

Examples: Devices like the Rotimatic, which can make various types of bread, are becoming more common in kitchens .

 

 

All of those examples are robots. I agree.

 

Here is a little opinion article about the relentless automation of everything.

[Advocatus Diaboli]

Thanks, Mind. I just wanted to confirm that your "physical embodiment" didn't mean to you only the type of "robots" that appeared in  Karel Čapek's play R.U.R which were "humanoid" in appearance.

Edited by Advocatus Diaboli, 22 October 2024 - 11:20 PM.

[Mind]

I know a lot of people say that a UBI will be the solution to unemployment. I have argued that it is just a path to dependency, not human fulfillment.

 

UBI supporters often use flowery talk about people going on a type of UBI - saying it will free up people to become artists, more educated, or be more creative. No UBI experiment thus far has shown that people will do these things. On average, people on a UBI become less productive and poorer in the long run (and also dependent). Here is another large study proving this. This is basically common sense - give people free money and most people will just spend it on leisure, drugs, and what-not. Eventually, they will not be able to provide for themselves. The people/owners and/or the government will be in total control. Do what they say or lose your UBI.

 

UBI is just the ultimate form of welfare and control. I don't think it is a very enlightened concept at all.

[QuestforLife]

I know a lot of people say that a UBI will be the solution to unemployment. I have argued that it is just a path to dependency, not human fulfillment.

 

UBI supporters often use flowery talk about people going on a type of UBI - saying it will free up people to become artists, more educated, or be more creative. No UBI experiment thus far has shown that people will do these things. On average, people on a UBI become less productive and poorer in the long run (and also dependent). Here is another large study proving this. This is basically common sense - give people free money and most people will just spend it on leisure, drugs, and what-not. Eventually, they will not be able to provide for themselves. The people/owners and/or the government will be in total control. Do what they say or lose your UBI.

 

UBI is just the ultimate form of welfare and control. I don't think it is a very enlightened concept at all.

 

I agree, and I'd go further and say that this is the inevitable result of the continued growth of technology. 

 

Sometimes I imagine I belong to a different species, one that is healthy, highly intelligent, and long-lived, but one that has no technology whatsoever. That seems like a true nirvana to me. 

[mag1]

This is my latest insight about the roborevolution.

 

Fatal "accidents" will soon be labeled as homicide -- Murder 1. How do humanoids lead us inevitably to this logic? As it is now, life is dangerous. It just is. You go out there into

reality and before you know it things go wrong. Life happens. This does not require some great malevolence -- no, at a certain level this is life; this is how life has been

for tens of thousands of years. You start doing things and tragedy can happen from time to time. This has not typically been described as some criminal intent; it is life.     

 

Humanoids likely will change this type of logic rapidly. Very rapidly. We might see a very rapid shift towards accidents becoming criminalized over the next year or two with the wide spread arrival of humanoids. Here's how that would work. So, you have some fairly dangerous task such as chainsaw operator, roofer, toxic waste handler,  etc. etc.. A humanoid can do these tasks or a human can do these tasks. What happens if there were an accident or exposure to some risk? The humanoid -- the scooterbot type at a cost of ~$2,000 -- can easily be replaced without criminal consequence. With the human, there really is no "replacement" if there is some serious harm or fatality that occurs. It then would seem highly reasonable to ask why did this tragedy even happen? Why is a human put at risk when a cheap robot could do the job for us? Why allow these preventable tragedies to happen. Ergo, the black hole of logic will drag us to the idea that "accidents"are no longer accidents -- they are crimes.

 

A public consensus could quickly take shape that those who act as agents in such harms such as employers etc. displayed criminal intent by not choosing the safer options of humanoids to do dangerous things. Anything that is in some way dangerous, annoying, stressful, etc.. would then be required to be done by humanoids. This formulation takes the job replacement hypothesis of technology one step further -- it will not even be entirely necessary for humanoids  to be that much more efficient than humans at a task, but more that they will be able to protect humans against the various harms of life. This line of thinking might dramatically turbocharge the rollout of roboworld. Roboworld could then become a moral imperative for all of those with a moral conscience -- technology could bring us a world in which humans largely would not be allowed t do anything due to the dangerous involved. A safety Uptopia? or Dsytopia?

 

The evaluation of human life is typically thought of in terms of so much value per quality year of life; so, perhaps as an example $50,000 per QUALY. Meaning that if a worker had a fatal injury on the job who  might have been expected to live 50 more years, then this would be a $2.5 million loss in welfare (though there would need to be time and mortality discounting of this amount).

 

The logic for a wave of safetism would seem unavoidable. Even if some might offer themselves to be put at risk for a price would be overruled by public sentiment. Risk is something that society will no longer allow others to assume within the context of work, social roles etc. We could soon be approaching a very different era. The epoch of living in some sort of jungle state of nature in which there are the tragedies of daily living could soon end. One might also expect that humanoids might be able to act as security screeners in bricks and mortar reality to reduce interpersonal violence. The changes we could see over the near term with humanoids could be profound.

Edited by mag1, 30 October 2024 - 02:33 AM.

[Mind]

This is my latest insight about the roborevolution.

 

Fatal "accidents" will soon be labeled as homicide -- Murder 1. How do humanoids lead us inevitably to this logic? As it is now, life is dangerous. It just is. You go out there into

reality and before you know it things go wrong. Life happens. This does not require some great malevolence -- no, at a certain level this is life; this is how life has been

for tens of thousands of years. You start doing things and tragedy can happen from time to time. This has not typically been described as some criminal intent; it is life.     

 

Humanoids likely will change this type of logic rapidly. Very rapidly. We might see a very rapid shift towards accidents becoming criminalized over the next year or two with the wide spread arrival of humanoids. Here's how that would work. So, you have some fairly dangerous task such as chainsaw operator, roofer, toxic waste handler,  etc. etc.. A humanoid can do these tasks or a human can do these tasks. What happens if there were an accident or exposure to some risk? The humanoid -- the scooterbot type at a cost of ~$2,000 -- can easily be replaced without criminal consequence. With the human, there really is no "replacement" if there is some serious harm or fatality that occurs. It then would seem highly reasonable to ask why did this tragedy even happen? Why is a human put at risk when a cheap robot could do the job for us? Why allow these preventable tragedies to happen. Ergo, the black hole of logic will drag us to the idea that "accidents"are no longer accidents -- they are crimes.

 

A public consensus could quickly take shape that those who act as agents in such harms such as employers etc. displayed criminal intent by not choosing the safer options of humanoids to do dangerous things. Anything that is in some way dangerous, annoying, stressful, etc.. would then be required to be done by humanoids. This formulation takes the job replacement hypothesis of technology one step further -- it will not even be entirely necessary for humanoids  to be that much more efficient than humans at a task, but more that they will be able to protect humans against the various harms of life. This line of thinking might dramatically turbocharge the rollout of roboworld. Roboworld could then become a moral imperative for all of those with a moral conscience -- technology could bring us a world in which humans largely would not be allowed t do anything due to the dangerous involved. A safety Uptopia? or Dsytopia?

 

The evaluation of human life is typically thought of in terms of so much value per quality year of life; so, perhaps as an example $50,000 per QUALY. Meaning that if a worker had a fatal injury on the job who  might have been expected to live 50 more years, then this would be a $2.5 million loss in welfare (though there would need to be time and mortality discounting of this amount).

 

The logic for a wave of safetism would seem unavoidable. Even if some might offer themselves to be put at risk for a price would be overruled by public sentiment. Risk is something that society will no longer allow others to assume within the context of work, social roles etc. We could soon be approaching a very different era. The epoch of living in some sort of jungle state of nature in which there are the tragedies of daily living could soon end. One might also expect that humanoids might be able to act as security screeners in bricks and mortar reality to reduce interpersonal violence. The changes we could see over the near term with humanoids could be profound.

 

One major assumption in this scenario is that robots will do everything perfect. Nothing will ever be able to perform perfectly with no errors in the "natural world". It would be an awful world where robots could create accidents and mistakes and never be prosecuted, yet if humans do the same thing they could get the death penalty.

[mag1]

mind, thank you for responding!

 

I think the logic that I have suggested is essentially unavoidable. we might now only be about 2-3 years away from a social Singularity that will inevitably follow from this logic.

When I actually thought about this I began to appreciate just how far reaching this new roboethics could become.You can start off with the fairly easy instances of chainsaw

operators and roofers; but there is so much more that can be done! Then you could think about those involved with hazardous chemical, or respiratory irritants ; then those

in transport ; then move towards those jobs that involve substantial psychological stress such as teachers and nurses; you might then think of those workers who must work

in physically demanding jobs such as in agriculture; then there are the dangerous jobs involving policing .... We have up till now on thread focused almost exclusively from

the point of view of how UBI might drive an existential crisis for people; here it is more about a push to enhance human quality of life by enhancing productivity and enhancing 

the well-being of workers. There is a substantial class based feature of the labor market in which there are a range of jobs that people will not do because they are dangerous

etc. So those who are willing to take such jobs then have a certain social advantage -- they will work the jobs no one else wants. Yet, if we were to then eliminate these jobs

because society found them to be too dangerous etc, then a wide range of people wold lose this social advantage and be at even more disadvantage..  

 

This is becoming scary. Over a 2-3 year time horizon we could start to see profound social change based upon the above logic.

[Mind]

Robots and AI will NEVER be able to operate perfectly in the environment, because nature is non-linear and chaotic.

 

Also, there are situations where there is no "good" solution/action. The self-driving vehicle accident is a good thought experiment in this regard. What you are suggesting is that AI could make poor decisions and kill how many ever people it wanted, yet face no penalty. However, if a person does the same thing, then they get prison/death penalty. I don't think many people will agree to this.

 

Also, if robots/AI do everything and humans are prevented from doing ANYTHING that might cause an accident, then humans will be relegated to doing NOTHING - because every action, no matter how trivial, has the potential to go awry.

[mag1]

mind thank you for responding! This thread has consistently through the years focused on the meat and potato issues of practical AI rollout into bricks and mortar reality, so we have been on the right wavelength and we are now in a great position to observe the emergence of humanoids and other applications.

 

One of the surprising updates I have noticed over the last few days is the mass scale roll-out of robotaxis. The latest volume that was mentioned was ~150,000 trips per week with Waymo -- it appears that this service has now entered a near vertical ramp-up. This is quite startling. Robotransport would massively redefine our economy; Once you can override 3D space, then a wide range of economic implications would seem to follow. For example, one might imagine that the robotaxis at some level might morph into robobuggies for retail shopping applications, etc.. The roborevolution has begun.

 

In terms of perfection, I suppose it is best to remember that perfection is the enemy of the good. With the latest news from Waymo, they reported with a very large volume of robotaxi trips they had had a few minor fender benders none of which caused serious illness. I am not entirely sure of the precise numbers of accidents with human drivers, though I can certainly imagine that it would likely be much higher. When you read of some of the accidents with human drivers you really can only gasp at the level of negligence that can be involved. There are drivers who cause accidents that are so drunk that they are simply unable to even maintain a standing balance -- people can be staggeringly drunk and still get into their car and put the community at risk. Indeed, automobile accidents are one of the leading causes of mortality in most Western nations. It is not that difficult to anticipate that this will not remain true for long. 10%o f the drivers likely cause 905 of the accidents. With Waymo, perhaps the drinking and driving crisis could end over the next year or two. 

 

The power of AI is so awesome. It has the ability to learn from every mistake and to avoid repeating the same mistake in the future. It is a learning machine! Every piece of information is fedback into its training set and can immediately be accessed by the entire global fleet! In comparison, we see humans who after a century of car driving constantly make the same mistakes over and over and this is causing carnage on our roadways. AI has this potential to approach [and potentially achieve] perfection, while there is no indication that humans will ever match such an outcome.  

 

Further, with respect to nature being non-linear and chaotic, I think it is interesting to remember that with robotransport there could be a way in which the technology could bring types of order to transport that would not be achievable with humans. For example, with robotransport one might simply get into the car and watch a movie or read a book; the need to rush here and there might no longer be present. You might even enjoy your "rush " commute tell AI not to get there too fast. with such logic; there might be a lowering of the speed limits as people preferred slow commuting to fast commuting. A slower commute would be a safer commute. Strangely, when traffic was flowing in this more orderly and safer way, then it might actually be possible to get where you are going quicker because traffic jams due to accidents etc. would no longer occur.

 

In terms of nature being non-linear and chaotic -- there is also the application of the humanoids to domestic tasks.What we have noticed is that typical domestic type tasks are much less chaotic, then what you might expect a robo-taxi might experience. With a domestic humanoid, you might ask it to cut the lawn, yet every time one cuts the lawn it is a nearly identical process. There is not that much variation from one time to the next. One might imagine that they would bring a humanoid home from the store and then for the first time or two it might be teleoperated. Yet, on the third run they could simply push a button and tell it to reenact the motions (with potential for some variation) from a previous lawn cutting and that would be it -- it would be largely programed to cut the lawn. After many instances of cutting the lawn, one could imagine that the basic program for lawncutting would be well-established and it would take very minimal effort to program a new instance.

 

Mind, I was thinking more about dangerous things that people do and the consequence that then ensue for them and others. For example, going out there with a chain saw. That can be a remarkably dangerous activity. After watching a few videos of what can go wrong when cutting down a tree with a chain saw I have a much better understanding of how dangerous this can be. Humanoids would seem to be a perfect substitute for human chain saw operators. This is nearly a perfect example, where you could replace the harms that now occur with humans and substitute these harms with at worst harms to a humanoid. That simply seems to me to be too easy of a choice to make. My hunch is that such substitution is now on the short-to-medium time horizon. It is hard to believe that such a substitution would not occur. So, what I am suggesting here is more that humans will be prevented from harming themselves by having humanoids taking over their roles. Now consider all the harms that can occur for humans in a wide range of tasks, teaching, nursing, various physical hazards etc.. Humanoids might have far-reaching effects over the near- medium term.

 

Yes, and your projection of this to its logical endpoint in which humans would no longer be allowed to do anything because there is at least theoretical danger in any activity (for some profoundly incompetent people there is probably real and not merely theoretical risk in EVERYTHING. A world in which humans are waited upon hand and foot would then seem a logical consequence.  Have any sci-fi writers ever explored this possibility?

 

The Social Singularity is now approaching!

[Mind]

Another problem with more automated functions of society - ceding all responsibility and control of running human society over to AI - is that there is no recourse for when the AI makes mistakes. There is no world in which robots and AI will never make mistakes. When society is run by algorithms and people have no input, a lot of people are going to be harmed.

 

Debanking is a good example. Once the algorithm flags you as problematic, you are pretty much screwed. You have to move heaven and earth to reverse the action. You go to the bank and they say - "we don't know how this happened, sorry, go find another bank"

[mag1]

Mind, this is becoming scary. Ten years ago they achieved superhuman ability in chess/Go etc., yet these are well structured games that have a solution that can be calculated. 

With more human type tasks such as driving, it almost felt as if such problems could not be solved -- They were too much embedded in the complexity of the human way. The fact

that driving has been unlocked would suggest that most other tasks such as mowing the lawn, making scrambled eggs, etc. will not be that far off either. The launch of the wide-scale

roll-out of robo-taxis now suggests that we truly are approaching some sort of Social Singularity.

 

I have suggested that this might be approaching pretty much consistently over the last 10 years with the robobuggies, but with robotaxis it truly is going large right now. Substantial

social disruption might occur over the next year or two. My hunch has been that we will know something is truly happening when total fertility rates begin to collapse. To some extent

this has already happened -- perhaps over the next year TFRs will REALLY collapse and people will become extremely worried about the long term viability of our species. For many

the prospects of bringing a life into this world who would exist on lifelong UBI and then being under the thumb of a totalitarian thought police would be intolerable. Making some tangible

contribution to the economic and social affairs of the community for many is deeply deeply hardwired into their core being. Such a life would for many seem unlivable. Many of the leading

technologists do not appear to reject such an assessment.

[Mind]

Here Google Gemini tells someone that they "are not needed" and "please die".

 

Similar to what some people have mentioned in this discussion - that human actions will be made illegal - after which humans will be "useless".

 

Thankfully, we hope, LLMs are not really thinking up this stuff on their own, but just using their training data and info that can be found on the web. If this is the case, then there is another problem - that too many humans talk openly about how humans are useless, destroying the planet, etc...

 

Now...could you imagine if you had a robot in your house and it looked you in the eye and said "you are not needed...please die". You would not sleep that night - not without unplugging/powering-down the robot, destroying it, locking it up, or something similar.