AI technology is advancing rapidly, but will it ever truly grasp the complexity of human emotions?
|
AI technology is advancing rapidly, but will it ever truly grasp the complexity of human emotions?
Simple answer, no.
Things such as facial expressions, tone of voice, other outside sources and emotions of others are often deciding factors, and AI can only understand previous data absorbed from outside sources most likely not involved in the situation.
Additionally, without emotions themselves, they will never understand true emotions, and can only form an emotional opinion based on (data) that is formed by the input or emotions of others, and therefore can not form their own opinions.
The lack of having emotions themselves, they also lack a key factor in forming an opinion f emotions of others, and anything otherwise is still just statistical data.
Emotions are something living beings have not machines.
Rick
Universal4
NoDepositCasinos (26 July 2024)
Agree with Rick, when you add in humour, irony, sarcasm etc - which are often misunderstood between humans when communicating directly with each other, let alone AI trying to interpret them - then i can't seen AI getting that close.
For the latest bookmaker new customer offers visit https://www.newcustomeroffer.co.uk/
I believe its a little bit early to comment about this. Let's have this discussion in a year time.
I believe they will understand human emotion in 5-10+ years; they will understand but not feel it - of course. And will be smarter and faster than human's in every sense of the word.
For now, it's all fun, "wow" & nice with ChatGPT and other AI tools..... for now. But I believe if humanity doesn't limit AI's knowledge, we as human's will not have a bright future.
Looking to buy or sell iGaming sites?
Get in touch today!
www.CasinosBroker.com
WhatsApp: +40-757-953-988
Yes, I agree, today AI is elementary. I know OpenAI is working on Artificial General Intelligence (AGI). There is also a race between countries to be the first to achieve AGI and Superintelligent AI (ASI). To be honest, this is scary, and we need some kind of regulatory body overseeing this.
With enough training, AI will understand human emotions, in a way. I read once that by mapping brain scans and feeding it to an AI, it can also read your thoughts if you think at certain things the AI already knows (while you are plugged into an MRI).
So to understand emotions means to read your face and your thoughts at the same time. Kind of complicated to do without a lot of radiation, right?
So I think it can, but it will never be properly done. Not in the upcoming decades, maybe never.
Hume AI is already starting doing that...
I think in 2-3 years AI will fully understand emotions of humans
I think the tricky part of this question is the word "understand." In a sense, it already does. AI can define emotions, detect them quite well in text/images/videos, infer what emotion a given situation might evoke, and adjust its responses based on the emotional tone you ask for.
However, on the other hand, I don't believe it can feel (yet) or empathize with the human experience, which doesn't mean it won't be able to imitate or influence it.
"It" can never feel.
It may get better at being able to "emulate" or "copy" or falsely respond to emotions making the other party "think" it is presenting a genuine emotion.
It can be programmed to present empathy, but this will not be a real "feeling".
Rick
Universal4
Cash Bonus (30 July 2024)
The current AI is nothing more than expert systems - it's not even attempting to understand the world.
ChatGPT and its ilk are pretty good at understanding English linguistics breaking phrase down to adjectives and nouns, and having a pool of words and phrases it use to identity object and intent of the human typer.
But there is no pretense of understanding what the concepts embody.
To create a true intelligence (or an AGI) that has physical world concepts requires a completely different approach, and I am not sure that training on billions and trillions of pages of the internet will ever provide a valid basis for that.
gm2891 (28 July 2024), universal4 (27 July 2024)
What complexities of human emotions? Something like how i feel right now and what I'm potentially up to? They won't know that. From written texts, easily. Google has had sentiment analysis in their algo for years.
You can test it yourself.
1: Write some 300-500 words yourself on any topic and throw you're text below this prompt.
2: Give it to a bot that helped you to write a sentence below your initial question...
3: Ask follow up questions, if needed, and you'll find more about yourself then you can think of.
PROMPT: Look hard at the words. Find the feelings, both clear and hidden. Note the tone and strength. Use Plutchik or Ekman to sort them. Don't miss the mixed ones. Show the main feelings and the ones underneath. Track how they shift. Paint the overall mood. Back it all up with examples. Say how sure you are. Point out what's fuzzy, and prove every point.
universal4 (27 July 2024)
There are people who want to make sure all this doesn't go nuts:
Safe Superintelligence Inc
Ilya Sutskever
I would tend to think this guy is right about AI :
https://www.tiktok.com/@octoberthe28...46537487977761
It's just a way to googling things super fast and then rearrange elements to make them look good.
agree, when I need to use google for research, like before I buy a product, using ChatGPT, really help me.
"Understand"... never. Only humans can understand. It's about consciousness and conceptual thinking. Read, interpret, figure out, ..., sure...
Serving gambling as a scientist: Providing gambling sites with expert content, advice, authority, and valuable exposure https://probability.infarom.ro/services.html
No, I don't think AI will ever fully understand human emotions. While AI can recognize and respond to emotional cues, truly grasping the depth and nuance of emotions—like feeling empathy, understanding the subtle context, or experiencing emotions—is beyond its capability.
I did it, and I found it to be a very interesting exercise. It reminded me of my work analyzing data when I used to do qualitative research. I've been away from that area for a while, but the potential of AI for this type of analysis is incredible. While I wouldn't replace direct contact with data for anything, the time-saving aspect is phenomenal, and I will be doing more experiments.
Thanks for suggesting it!
gm2891 (30 July 2024)
Stop questioning the power of AI. It will evolve rapidly once human-brain chips are connected to the AI. Then AGI would be a few months away.
Kodak believed digital cameras were a fad, now it's dead.
Nokia made fun of tap-phones and now it's nowhere.
Never question Progress.