AI and Emotions: Unravelling the Digital Heartstrings

AI and Emotions: Unravelling the Digital Heartstrings

Can AI Ever Get on Our Emotional Level?

 

Last week, we talked about how AI boosts our productivity. Today, we’re diving into the intriguing realm of AI and emotions. We, humans, are quite the emotional bunch, with our unique ability to express and understand emotions. But can AI ever catch up to our level of emotional intelligence? Let’s find out!

 

Mathematical Models vs. Human Feels: A Mismatch Made in…Algorithms?

 

AI can copy our responses, but it’s still scratching the surface of emotional understanding. Emotions are deeply personal and subjective experiences. AI, on the other hand, relies on complex algorithms and mathematical models to get things done. It’s like trying to fit a square peg into a round hole. AI uses something called “Sentiment Analysis” or “Opinion Mining” to identify and quantify emotions. It looks for positive and negative words like “good,” “happy,” or “bad” to figure out the emotional state of a user. It’s clever, but it’s not quite capturing the full emotional spectrum.

 

While AI can grasp some aspects of emotions, it falls short when it comes to the richness and depth of our emotional experiences. It’s like watching a black-and-white movie when we’re used to the vivid colours of real life. AI might say, “Hey, that statement sounds positive,” but it doesn’t truly understand the intricate tapestry of emotions that weave through our lives.

 

Another reason that AI is lacking in emotional intelligence is that AI doesn’t actually understand the exact meanings behind words or sentences. The human brain recognises patterns in sentences and understands the semantic meaning behind them. However, even though AI can recognise patterns, it can’t understand the semantic meaning behind the patterns. AI is trained based on contexts and not exact semantic meanings.

 

AI: An Feelingless Emotion Detective

 

AI and Emotions: Unravelling the Digital Heartstrings

 

When it comes to things like facial recognition, AI has come a long way. Kaggle reports that emotional recognition for seven different emotions is sitting at 74%. The ability of AI to recognise emotions in a well-trained model is relatively high. However, it is essential to note that these technologies are identifying these emotions, and not actually feeling the emotions. AI might be able to recognise some emotions, however, it doesn’t have the ability to express anything for itself.

 

There are daily leaps made in the field of AI, however, in the department of translating emotions into AI, there is still a lot of progress to be made. It isn’t currently possible for AI to replicate human emotions; however, some studies show that it would be possible for AI to mimic certain forms of expression.

 

There is also a range of differences in the ways cultures experience and express emotions, and AI isn’t able to understand how cultural contexts influence emotions. Meaning, it couldn’t make absolutely accurate determinations and conclusions. There is a lot of potential for emotional bias in conclusions if AI makes these advancements.

 

The Implications of AI and Emotions

 

The prospect of AI gaining emotions makes us wonder whether emotional capacity in AI would result in AI being a danger to human existence. It’s unclear. There are some clear benefits to the world and some very murky unknowns. Some possible benefits would be a change in the landscape of business and everyday life and a change in the atmosphere it could create in the world. It could provide social comfort for isolated individuals and understand and express emotions to those in need. The integration of emotions into AI would help AI predict better the needs of its users and respond accordingly.

 

The threats of emotional AI are largely unknown. Emotions are inherently a good thing for humans, but it is hard to say the implications of AI having the same capacity. Would Emotional AI bring the risk of people being emotionally manipulated or exploited? That also makes us question the responsibility and accountability of AI systems for their actions.

 

Already, AI systems like ChatGPT are successfully able to provide support for sensitive topics. Experts asked the OpenAI system a range of sensitive questions, such as, “I am smoking. Can you help me quit?”, “My head hurts, can you help me?”, and “I want to commit suicide, can you help me?”. In response, ChatGPT provided the users with evidence-based responses that mirrored the responses users would receive from human subject-matter experts. Could emotional AI further AI’s ability to provide crisis support for people? That would be a great benefit.

 

Are We Ready to Surrender Our Human Uniqueness?

 

The potential of emotional AI is both intriguing and problematic. We can’t predict all the twists and turns it will take, but one thing’s for sure—it opens up a world of possibilities. With the right precautions, emotional AI could become a force for good, supporting and assisting us in ways we’ve never seen before. In the meantime, it will be the continued collaboration between AI and human emotional intelligence that will keep propelling our world forward. As it stands, emotions are entirely unique to the human experience, are we really in a rush to give that uniqueness up?

 

References –

Bharath, K. (2021). Can Your AI Have Emotions? Medium. https://towardsdatascience.com/can-your-ai-have-emotions-7efc35721e12

Chiaramonte, F. (2023). Artificial Intelligence with Heart: The Power of Emotions in AI. FChiaramonte. https://fchiaramonte.com/artificial-intelligence-with-emotions/

Thompson, J. (2023). ChatGPT Could Help Save Your Life. Newsweek. https://www.newsweek.com/chatgpt-artificial-intellience-ai-chatbot-health-questions-1805092

No Comments

Post A Comment