Theory of Mind in Algorithms: Achieved! 

Today, progress in technology is astonishingly fast. In my last post, I suggested that ChatGPT, especially version 4, could be perceived as a milestone, the first algorithm that we can classify as AGI, or non-narrow, specific intelligence (like a chess engine or a medical scan recognition engine), but general intelligence, “understanding” contexts, theories about theories, theories about theories about theories (“Nina thinks that Chris said that George saw…”) etc… 

 

I summarized it as follows: “Here I believe that ChatGPT and other similar tools have crossed the barrier that we can call “intelligence.” Furthermore, I think that ChatGPT might be a cornerstone of so-called AGI (Artificial General Intelligence). This might be even more true with ChatGPT 4 which is currently being released. I see it this way: in future AGI history books, chatGPT will be the first big milestone.” 

Check also the blog post

“What I talk about when I talk about… generative AI”

Emergent Theory of Mind in Large Language Models

Today I came across a paper that not only shook me but also directly confirms my thoughts (link here). Its title alone is very strong: “Theory of Mind May Have Spontaneously Emerged in Large Language Models.” 
Theory of mind is the ability to attribute oneself and others with mental states such as beliefs, desires, intentions, and emotions and to understand that these mental states can influence behavior. In other words, it is the ability to understand that other people have thoughts, feelings, and intentions that may differ from our own and to use this understanding to predict and explain their behavior. It is a fundamental cognitive ability that allows us to navigate social interactions and develop relationships with others. 

Implications and Limitations of AI Theory of Mind

Until now, it was believed that some animals (mainly mammals and birds) had these skills, and only humans had them in an advanced way. It turns out that classic tests typically applied to people to verify their ToM “level,” when applied to algorithms, correctly solved a maximum of about 40% of tests by 2020, which corresponds to the level of a 3.5-year-old child.  

After 2020, the situation changed dramatically, and ChatGPT version 3.5 already achieved 90% correctness, which is the level of a 7-year-old child. ChatGPT-4 reaches 95%… According to the authors, this indicates that an artificial algorithm has achieved a level of Theory of Mind comparable to that of humans. Moreover, and here I see confirmation of my assumptions, ToM can emerge spontaneously in the course of learning something else (in this case, a language model).  

You do not need to specifically program or design ToM “functionality” in the algorithm. It turns out that such “incidentally” created surprising “functionalities” have already happened. For example, during training image recognition, the network learned to count objects. Simply put, this activity turned out to be helpful in recognizing images, so its familiarity “evolutionarily” increased the value of the objective function. 

Theory of mind is the ability to attribute oneself and others with mental states such as beliefs, desires, intentions, and emotions and to understand that these mental states can influence behavior. In other words, it is the ability to understand that other people have thoughts, feelings, and intentions that may differ from our own and to use this understanding to predict and explain their behavior. It is a fundamental cognitive ability that allows us to navigate social interactions and develop relationships with others. 

The Relationship Between Theory of Mind and Consciousness: Defining Terms

Here is an important note. Theory of mind (in my opinion) is not synonymous with consciousness, but undoubtedly such an ability in AI algorithms is a powerful step forward. 

So, what about consciousness? Here, referring to my last article, a lot depends on the definition. Self-awareness refers to a human’s consciousness of their own thoughts, feelings, and experiences.  Conclusion: looking ahead to the future of AI and consciousness. 

Contact

Ready to meet the only technology partner you'll ever need?

Cloudica needs the information you provide to contact you about our services. You may unsubscribe from these communications at any time. For information on how to unsubscribe, as well as our privacy practices and commitment to protecting your privacy, please review our Privacy policy.

Once the above questions have been answered, a disaster recovery and backup readiness index can be calculated based on the following scale: 

  • Level 1: Inadequate – The organization has significant gaps in its disaster recovery and backup posture.  
  • Level 2: Developing – The organization has some disaster recovery and backup processes in place, but significant improvements are needed.  
  • Level 3: Mature – The organization has a mature disaster recovery and backup posture, but there is room for improvement.  
  • Level 4: Robust – The organization has a strong disaster recovery and backup posture and is well-prepared to address potential disruptions.  
  • Level 5: Exceptional – The organization has a comprehensive and mature approach to disaster recovery and backup. 

The disaster recovery and backup readiness index can be calculated by assigning a score of 1-5 to each question based on the level of readiness demonstrated. The scores are then averaged across all questions in each category to determine the readiness level for that category. The overall disaster recovery and backup readiness index is calculated by averaging the readiness levels across all categories. 

Level 1: Basic
You have minimal cybersecurity processes in place and face a high risk of cyberattacks. Immediate attention and significant improvements are necessary to enhance your security posture.

Level 2: Developing
You have some cybersecurity processes in place but require substantial improvements to reach a mature state. You should focus on strengthening your policies, procedures, and security controls.

Level 3: Mature
You have a solid cybersecurity posture, but there is still room for improvement. You should continue enhancing your processes, monitoring capabilities, and incident response practices.

Level 4: Advanced
You have a strong cybersecurity posture and are well-prepared to address potential threats. However, you should remain proactive and stay abreast of emerging threats and technologies to maintain your advanced level of security.

Level 5: Leading
You have a comprehensive and mature approach to cybersecurity. You are a leader in cybersecurity best practices and continually innovate to stay ahead of evolving threats.

Download E-book

"The best way to Outsource IT Staff"

A Game-Changing strategy for business success. How Outsourcing IT Staff Can transform your business.

Cloudica needs the information you provide to contact... Read more

Dziękujemy za rejestrację!

Link do webinaru otrzymają Państwo mailowo dzień przed spotkaniem.

23 Marca 2023

10:00 via MS Teams

Tomasz Woźniak

Thank you!

To download our e-book „The best way to Outsource IT Staff” click button below