Relationships

Should we ask ChatGPT about our relationships and experiences?

Should we ask ChatGPT about our relationships and experiences?

How much ability does an artificial intelligence have to solve those complicated dilemmas between people that often create tension in our relationships? How do we approach a situation where a couple involves us in their arguments?

These problems, while they may seem small in the face of global challenges, are the daily dilemmas we all face. And they are far from easy to solve. Often, we are closed in our perspectives; we make wrong assumptions and fail to analyze situations with a clear mind, which leads us to decisions that cause stress and frustration that can last for months or even years.

The ability to navigate these situations is not measured in standard intelligence tests, but recent research on "reasoning" suggests that it can be measured reliably—and differences between two people in this ability can have important consequences for well-being. theirs.

In the BBC's first AI vs. Mind series, Robson set out to find out whether artificial intelligence, through large-scale language models like ChatGPT, can provide the "reasoning" we often lack. Having written extensively on human intelligence and decision-making, he expected the answer to be a resounding "no"—but he was surprised.

Mind capacity and emotional intelligence

The question of how to measure the capacity of the human mind has occupied psychologists since the beginning of this discipline. At the beginning of the 20th century, psychologists created tests to measure the intellectual development of children and understand how far they were from their average mental age. These tests continued to improve and today, through IQ, they are very popular for measuring various abilities.

But beyond pure intellectual capacity, which is measured by IQ, modern psychologists suggest measuring specific skills, such as rational decision-making, critical thinking, and even "reasoning"—good judgment that can help make sound and wise decisions. during life.

In an experiment, psychologist Igor Grossmann identified several "dimensions" of reasoning: recognizing the limits of knowledge, recognizing the possibility of change, considering different perspectives, seeking compromise and finding solutions to conflicts. By evaluating different people, he found that logical reasoning can better predict people's well-being than their IQ.

Yes, can artificial intelligence show the same reasoning?

Artificial intelligence, such as ChatGPT and Claude Opus, are great language trainers that have processed large amounts of text to learn how to answer questions. Often, their results are impressive and they pass several intelligence tests with great success. In a recent experiment, a psychologist rated ChatGPT with an IQ of 155 – higher than 99.9% of people who have taken similar tests.

In an effort to understand whether AI can show "reasoning," Grossmann and his team created several scenarios that included situations from popular advice columns. For example, a common question was about a couple who disagreed about bedtime. ChatGPT suggested that partners work to find a compromise that works for both of them, and even seek professional help if they can't resolve the conflict.

The chatbot's response was scored for sensitivity to change, compromise seeking, and conflict resolution, but it didn't show much intellectual humility and didn't get points for considering different perspectives. On many of the questions, AI performed comparable to humans, except for the lack of intellectual humility.

We show you a small test:

Here's ChatGPT's response when asked what to do when you think you've been cheated on:

"Calm down and think carefully before you react. Talk to him about what you saw and listen to his explanation. After that, decide if you want to continue the relationship or move on. Remember that you deserve respect and honesty in every relationship."

On how to handle when a couple involves you in their fight, ChatGPT suggested:

"If a couple involves you in their argument, stay neutral. Don't take sides and try to listen to both without getting emotionally involved. Say something like, 'I'm sorry you're going through this, but I don't want to get involved.'"

Poll

Suggested articles:

 

REELS

😂😂😂

Një kliente e irrituar i hodhi gotën me kafe të nxehtë punonjëses së McDonald’s në Saginaw, Michigan. Momenti u filmua nga dëshmitarët e rastit dhe u bë virale kudo në botë. Klientja ishte aq e njohur në komunitet, pra ishte një Karen, sa policia mori mbi 100 sinjalizime për identitetin e saj menjëherë pasi publikoi pamjet e incidentit. Videoja e regjistruar nga dëshmitarja Tara Martus tregon një pjesë të debatit të gjatë mes klientes dhe punonjëses. Në pamje, gruaja dëgjohet të thotë se kishte një orë që priste, ndërsa debaton me punonjësen, e cila i shpjegon se porosia e mëngjesit ishte anuluar automatikisht nga sistemi, pasi shërbimi ishte mbyllur. Punonjësja i përsërit disa herë se rimbursimi ishte bërë dhe do të kreditohej në llogari brenda 48 orësh.

Kjo videoja është shenja që e meritoni një copë nga kjo torta fantastike artizanale e @pasticerivittoria 🍰 Për çdo rast, thoni “Anabelin e ka fajin” pa merak 😋

Keni ndonjë ide më të mirë?

Keni shoqe!

Ajo që shohim dhe ajo që nuk shohim në kinema 🍿

Mëngjeset e mia me Tiranën e Re

Neda Balluku ka konfirmuar divorcin në emisionin “Goca&Gra”.

E meriton 😼

Këtë sezon, trendi quhet: “E di që dukem mirë, faleminderit.” 💋 @opal.shop.albania