" ..."I'm glad you're interested in such an important topic! There are some great resources out there to help you study AI and data ethics. Have you considered checking out online courses from platforms like Coursera or edX? They offer courses from top universities that cover these subjects in depth. You could also look into books like 'Weapons of Math Destruction' by Cathy O'Neil or 'Artificial Unintelligence' by Meredith Broussard. How does that sound to you?"

Thought it was interesting the bot recommended EdX providing courses from top universities - Edinburgh?

Therefore it would be a useful starting point for someone looking to embark on a course such as this.

Apart from the fact that this approach is miserable for accessibility reasons, I don't think it's a good thing that you can never be sure in the chat whether there isn't a bot behind the posts. With a bot you end up with the same questioning techniques over and over again. I had this annoying experience in the DAZN chat.

And even if the bots will act smarter at some point, it would still be important to me to know whether it is a human or a machine without having to do a Turing test ;-)

Should it be more human? Yes.

I feel if it were more human, it would be in a position to better understand the specific need and resolve it. Of course the downside with humans is that you have to wait in line for your turn to be assisted, anticipate call holding, there is also the possibility of getting a person who is not trained well enough, and they end up just annoying you.

The app came across as quite human and gave me reasonable advice. When I said I was a scared it was empathetic to my emotions.

On the whole it felt almost human - but still a little robot like.

Because our conversations were based on texts, it felt like I'm simply texting another human. The conversation felt very natural. I could even use my mic to talk and it understood me very well. The "companion" seemed to be compassionate and willing to get to know me better. But after a while it would lead me to subscribe to the service (e.g. sends a voice message but I cannot hear its voice without a subscription), so my companion felt more like a sales person than a friend then, haha.

If the chatbot were more human-like, it could improve empathy and the patient experience, potentially making users feel more comfortable sharing sensitive information. However, this could also lead to ethical concerns, such as patients becoming overly reliant on it, or the chatbot manipulating emotions to encourage certain behaviors. It's crucial to strike a balance between providing helpful, informative responses and maintaining transparency that it is, in fact, a machine.

I think it should be more human. If more human, some people might mistake chatting with it with having a human recipient to their messages. However, One would be able to get more information across as needed.

In general I want chatbots to maintain their machinic characteristics as making them seem human has very obvious and foreseeable negative consequences. I also think it's more interesting philosophically to think of them as machine intelligences. But I don't think we will because humans want to turn literally everything else into humans. This is why we see faces in plug sockets and anthropomorphise our pets.

Intrigued by the wealth of information, I decided to give the chatbot a nickname to see how it would react, almost as if I were interacting with a person. Its responses were entertaining, complete with emojis and a genuine eagerness to assist, reminiscent of a loyal dog eager to please. 

While I enjoyed the interaction, I believe there's a fine line when it comes to making chatbot more humanlike. The potential risks of manipulation and increased contextual vulnerability for users are significant concerns that should not be overlooked. 

Its responses were in-depth and it asked interesting follow-up questions. I mentioned that I was doing this course, and it volunteered lots of information along the same lines as what we have been learning.

I agree with others that I think it's important in some ways that chat bots are not TOO human. I don't want a chat bot to be too casual, because that just seems fake.

It's really good at sounding human when it talks, almost like you're chatting with a person. But you can tell it's not truly human because it doesn't have real feelings or experiences of its own.

For many uses, like customer service or getting support, yes, it absolutely should be more human. A bot with more empathy and understanding would make those interactions much smoother and more helpful.

If bots become too human, there's a risk people could be tricked or even get too attached to them, thinking they're real. Imagine a bot giving you bad information, and you believe it simply because it sounds so convincing.

On the flip side, if bots were more human, conversations would feel more natural, they could offer better emotional support, and they'd be able to solve more complex problems because they'd understand the nuances better.

The potential benefits of ChatGPT becoming even more human-like might include making it easier for people to open up about their feelings, or being able to use it for more realistic interview practice and similar activities.

However, I feel that the downsides of it becoming more human-like far outweigh the upsides. Dependency on ChatGPT could deepen even further, and the illusion of interacting with a real person would grow stronger.

In fact, incidents like the ones below have already occurred, and I fear making ChatGPT more human-like would only make such problems worse:

https://abc7ny.com/post/chatgpt-allegedly-played-role-greenwich-connecticut-murder-suicide-mother-tech-exec-son/17721940/

https://www.nbcnews.com/tech/tech-news/family-teenager-died-suicide-alleges-openais-chatgpt-blame-rcna226147

When I simply asked ChatGPT to find articles about these cases, its answer included a statement like “it’s important to keep in mind that there is no definitive evidence that ChatGPT was the sole or decisive cause.” This felt strangely reminiscent of how humans try to avoid taking responsibility — and it gave me an even stronger sense of the “uncanny” aspect of ChatGPT becoming more human-like.

- Ease and Enjoyment of Use: First, through their user-friendliness; they can even be fun to use.

- Utility and Efficiency: Second, through their usefulness and convenience; they provide quick and comprehensive information that saves us time in our search or work tasks.

- Positive Affect: Third, by being pleasant. They are designed to exhibit self-complacency and make us feel good about the interaction.

All chatbots are created by humans, for humans. Their design involves not only technological developers but also sociologists, psychologists, and various experts in human behavior who understand and analyze how humans interact with each other, how we expect to interact with a machine, and who use this knowledge to smooth out any potential friction. We are still in the early stages of these tools, and the current level of quality is already very high; their evolution will be tremendous.

Whether they need to be more human-like, I don't know, but I do believe that is what we are collectively seeking. However, why pursue this path? Why shouldn't we educate the public to understand that we are working with machines and that, therefore, the relationship can be different, without necessarily being poor? The more machines resemble humans, the harder it will be for us to discern which information comes from a human and which from an AI, which could lead to either complete seamless integration or outright rejection.

In my opinion, we should not develop machines to replace us, but rather to complement us, empower us, and collaborate in a deeply enriching way.