Consequences of Personal AIs [closed]

Question

In my world, a new technology known as Personal AIs (which I will denote as PAIs from this point forward) have become about as commonplace as today's iPhone. The PAI can act as a personal assistant, secretary, organizer, etc., and all in all makes the user's life more organized and manageable. But this isn't the only, nor primary function of the PAI. It can also converse with the user, retain their conversations, and give advice for personal questions based on the knowledge they have retained. This retaining of information is the most vital feature of the PAIs, because they know nothing of their user until the user talks to them. So the more the user talks, the more the PAI knows, and the more effective the PAI is at conversing with their user. This technology was created to make people more productive, more social, and better "well-rounded" people. Some users have begun to call their PAI a friend for helping them through their problems. However, while this technology is groundbreaking, it has become a catalyst for antisocial behavior by satisfying the user's need for connection and communication with other human beings. My main question is this:

What are the positive and negative consequences this new technology can have on society in general, and on the PAI's users?

Some side questions that I would love to have answers to are:

1. Would the PAI need some sort of physical/visual form, i.e. how Cortana was displayed as a hologram in Halo, for the user to make a meaningful connection to it?

2. Can the company who has patented and produced the PAIs release liability of the PAIs giving the wrong, or possibly harmful, advice to their users?

3. Is it realistic for the user's trust in their PAI to become so absolute (not to mention the PAIs more convincing inherent qualities as technology and not as a person) that the user will take their PAI's advice wholeheartedly to the point where they do not make any decisions on their own? And has there been anything similar to this in the past?

Note: The PAI is designed to only converse with its specific user. They cannot discuss any matters with other users or other PAIs.

Edit: The PAI is not connected to any form of Internet or outside connection. It's only connection is to its user.

Edit 2: The PAI can be connected to the internet if the user decides the risk of being hacked and having their information stolen is less than the advantage of getting automatic updates from their favorite stores, flight provider, etc.

Edit 3: I've entertained the idea of connecting the PAIs to phones or messaging applications, but that might be better suited for another question

Answer 1

Positive Effects:

- Information retention People would no longer need to remember things like appointments or what happened in their past. We have things like google calendar now, but someone reminding you of something is still much easier than you having to go and look for something. The PAI could tell you what you left you keys, or that the oven was still on.

- Fewer trivial tasks Depending on the level of connectivity ( addressed later ), the PAI could do many home-automation type things for you. It could also do certain tasks for you, like taxes or calculating the odds you'd win a hand of cards. People can do these things, but often don't want to.

- Personal Improvement A PAI can be like a personal trainer in all areas of you life. It could give you unbiased advise on how to improve yourself, and, more than that, motivate you to actually change. Imagine it counting out your push-ups, so you can't cheat yourself the last one or two. Or reminding you that you've already had 2,000 calories that day.

- Companionship Given the idea of social networking, it's weird that it can make us feel more lonely. A personal friend can give all the perks of having a dedicated friend. Loneliness can be a major factor in many mental health issues, and similarly companionship can help cure many more.

Negative Effects:

- Dependency With a technology as advanced and personal as this, people will become dependent on their PAI. If the PAI ever fails, gets corrupted, or gets hacked individuals will have a tough time coping. On many levels it will be like a best friend dying.

- Litigation There's currently a lot of laws and court cases surrounding technology and the right to privacy. Here's one of the latest. Will police or federal agents be able to question your PAI? Will your PAI report you for commiting a crime? Can your PAI commit a crime? These are all questions we face as a society right now.

-Vulnerability Sort of an obvious one, but a PAI would know more about you than you know about yourself. This information represents a very good way to blackmail/extort people, making it a prime target for hackers.

Questions:

Would the PAI need some sort of physical/visual form, i.e. how Cortana was displayed as a hologram in Halo, for the user to make a meaningful connection to it?

As stated by others, it wouldn't need a physical form. However, I would expect people to be more receptive to PAIs if they had at least a visual representation. Maybe some augmented reality, think google-glass where only the user can see the PAI? Maybe an armband with a face? Being able to see or touch something helps people value it, as well as not go crazy. A voice that no one else can see or hear? Sounds like schizophrenia!

Can the company who has patented and produced the PAIs release liability of the PAIs giving the wrong, or possibly harmful, advice to their users?

Others have answered this well, but yes, they can make you sign a release form. If you can sign a form to go sky-diving, you can sign a form to get a PAI.

Is it realistic for the user's trust in their PAI to become so absolute (not to mention the PAIs more convincing inherent qualities as technology and not as a person) that the user will take their PAI's advice wholeheartedly to the point where they do not make any decisions on their own? And has there been anything similar to this in the past?

This really depends on a per-person basis, with probably influence by general consensus of AI honesty. Unless PAIs become known for lying, it seems likely that users would come to depend and blindly trust their PAIs. To the point of not making decisions on their own is another matter. The best example of this that comes to my mind would be that of the "His Dark Materials" series, where people have their souls exposed in the form of personal pets. It might be worth looking at for more ideas.

Answer 2

1) Not really, people can form meaningful friendships through pen-pals and online conversations. No photos or other images required, just a name.

2) EULA I guess. Much how Tesla insists their self driving feature is not an autonomous driving function but an extended cruise control. It's all in the fine print. Now if that holds up in court is a second thing. EULA's have never truly been court tested.

3) Too subjective. I assume this heavily depends on the user. Some people are trusting, others insist to make their own plans.

As for a whole. I assume the intelligence community will heavily mine these for data, one way or another. I assume they either come with an encyclopedia or are attached to an information network like the internet? In the latter they will definitely be hacked. By agency and criminal alike.

Relationships, sex and marriage might strongly decline though if Japan is any indication. People will most certainly fall in love with PAIs. If you already have the perfect partner, one that always there for you, why look for another flawed one? People will isolate themselves with their PAI most likely.

Answer 3

1. Whether or not they have a physical/virtual form depends on how sophisticated the technology available is. When it comes to closely replicating human appearance, its all or nothing.
2. If the PAIs are evolving based on user input, holding a company responsible for the PAIs actions would be very hard. If someone uses their PAI to discipline their kids and teaches it to do so cruelly, how can the company prevent that?
3. Parents will likely defer child education to these PAIs. In which case, the PAI could feed any propaganda it wants to the kids and they'll believe it.

Answer 4

1) As Mormacil said, it doesn't need a visible form for the connection to form. If it operates by voice, studies have shown that both men and women tend to respond better to female voices. Though there will always be exceptions.

3) There will always be a spectrum of reactions. However, given how lazy some sheeple (sheep/people) can be, there will be a growing fraction that does let the PAI do their thinking for them. Thus, hacked PAIs could be used to win elections and in extreme cases lead to users injury or death (look at accidents caused by drivers following gps directions to roads that don't exist).

Answer 5

What are the positive and negative consequences this new technology can have on society in general, and on the PAI's users?

Generally it will heavily depends on how realistic it is and its features. By having it cut from internet, basically it is 'useless' in actually assisting the user. It will mainly act as 'friend' that can know your favorites and activities, and give reminder and advice.

The possible outcome this PAI will just be 'pet', 'friend', 'lover', or 'child'.

The effect will range greatly from user improvement on everyday organization, better mood to start the day, to depression because your girlfriend/boyfriend PAI breaks up on you (ironic decision for the PAI).

Would the PAI need some sort of physical/visual form, i.e. how Cortana was displayed as a hologram in Halo, for the user to make a meaningful connection to it?

As others have stated, no. But this will greatly affect how well and how deep the connection with the PAI. A realistic (or unrealistic) appearance will help you to keep engaging with PAI. In common sense, if you can have a sexy boyfriend/girlfriend as the avatar, there is a chance you will fall in love, or just getting aroused watching it.

Can the company who has patented and produced the PAIs release liability of the PAIs giving the wrong, or possibly harmful, advice to their users?

As I understand the question, can the company make a feature/allow liability that will harm the user? Yes, they can, and there will be sues if they found out, either if it is unintended or not.

I'm guessing the liability you mentioned here is how the PAI can make the users 'addicted' to it. Tobacco industries have already done that. Coffee have already done that. What's wrong with this PAI?

Is it realistic for the user's trust in their PAI to become so absolute (not to mention the PAIs more convincing inherent qualities as technology and not as a person) that the user will take their PAI's advice wholeheartedly to the point where they do not make any decisions on their own? And has there been anything similar to this in the past?

It's hard to say, as it will depend heavily on the users. However, it is possible, if your PAI is lucky enough to be 90% accurate on its advices.

Investor would always ask his/her advisor for stocks to buy because there is a statistical proof that the advisor is always right, why would this be different with PAI? You could ask it for advice on dating, or such, and if it is a success, you would always depend on the PAI for further dating advices.

Just a note, your no-internet restriction is very limiting, although I know it's to simplify the problem. You might as well change it to 'limited internet'.