humans can create stories because they have real experiences. ML on the other hand does not. They can only lie about it. Sometimes humans are good at detecting lies.
So I cheated and asked chatGPT 4 your question verbatim. Here was the reply:
While I am designed to be an advanced language model that can perform a variety of tasks, there are certain tasks that I am not capable of performing as well as humans. Here are some examples:
Emotional support: While I can offer advice and suggestions, I do not have the ability to understand and empathize with human emotions in the same way that a human can.
Physical tasks: As an AI language model, I am not able to perform physical tasks such as cooking, cleaning, or driving.
Creative tasks: While I can assist with tasks such as writing or brainstorming, I do not have the same level of creativity and intuition as a human when it comes to generating new ideas.
Personalized healthcare: While I can offer general health advice and information, I am not a substitute for personalized medical advice from a healthcare professional who can take into account an individual's unique medical history and circumstances.
In summary, while I can perform many tasks and provide valuable assistance, there are certain tasks that require human capabilities and cannot be replaced by an AI language model like ChatGPT-10.
@alexia_georghiou Hopefully this is true but the current model doesn't know what GPT-10 is capable of.
Report
@singh_ankur yeah …I guess this is an example of the 10% hallucinating. It just went with what I asked and answered accordingly:)
Report
@alexia_georghiou If the bot would know enough about you (eg from your journaling) or have access to your medical history, those two could also be replaced.
Report
@veit_josef_schneider I wonder how much it covers so we won’t panic with the capabilities. I think you are right.
For example:
Machines can create thousands of poems, but they cannot pick out the one that universally moves humans.
Why are humans moved by a poem? I believe it comes from our ability to empathize, from our universal memories, experiences, and sympathy within our species.
However, the growth process of artificial intelligence determines that it cannot experience the sufferings and joys that humans go through, cannot truly identify with humans, and cannot truly become human.
I think this can be seen in many works of science fiction with philosophical implications.
@blueeon Bots already drive appreciation with likes, upvotes and comment.
We might have already lost an ability to appreciate someone with words.
letters, speaches and poems are a dying art.
Report
@blueeon@singh_ankur
Dr. Larson was hurt when the audience concluded that his piece -- a simple, engaging form called a two-part invention -- was written by the computer. But he felt somewhat mollified when the listeners went on to decide that the invention composed by EMI (pronounced ''Emmy'') was genuine Bach.
https://www.nytimes.com/1997/11/...
@singh_ankur Yeah, you're right. But I think it's the machines using our human sense of taste. Machines can learn from past preferences (labeled data), but can they discover new trends and lead human taste?
It is like asking how will people get from A-B now we have invented the wheel. We will just get there faster and easier.
Everything we have invented in the past has been to support human development. This is no different.
I think fact-checking and editing for style.
You can only verify so much information online—that's why major publishers and publications hire fact-checkers to actually call people (although less of an issue in saas and other industries that intentionally use spin).
Using AI to copy-edit text for grammatical issues is helpful. But style is often a personal choice that changes based on the audience. I think a lot of AI that tries to do it currently is too broad (professional, informal, academic, etc.)
That said, if AI could start using data from stylebooks to understand different rules/reasons behind editing decisions and do so effectively, that could be a total game-changer. I just think it's harder to train AI to do that than black-and-white grammar checking, but I hope someone proves me wrong some day!
Creative tasks. I think the human mind's creativity is limitless in comparison to what AI could achieve. However, if sentience is achieved, that's a different story
Replies
Vadoo AI
Chat with YouTube
Chat with YouTube
Vadoo AI
Vadoo AI
ReadPo
ReadPo
Vadoo AI
Vadoo AI
ReadPo
Conversa - Videos That Talk back
Conektto
Vadoo AI
Bloks
Moovby