Singularity: Are we getting closer to it or drifting away from it?
(image generated with DALL-E)

Singularity: Are we getting closer to it or drifting away from it?

(Note: I have not yet read the recently published book "The Singularity Is Nearer: When We Merge with AI" by Ray Kurzweil. However, the title alone inspired me to write this short article)

The concept of technological singularity, proposed by mathematician and science fiction author Vernor Vinge, refers to a hypothetical future point at which artificial intelligence surpasses human intelligence, leading to rapid and unpredictable technological growth. Many have speculated that recent advances in artificial intelligence, particularly in areas such as deep learning and natural language processing, indicate that we are approaching this tipping point.

However, a more nuanced look at our progress reveals a paradox: the deeper we delve into artificial intelligence, the more we discover the enormous complexity of natural intelligence and cognition. Our current AI systems, impressive as they may be, rely heavily on statistical patterns and brute computation rather than demonstrating accurate understanding or general intelligence.

Consider the following thoughts:

"In our quest to emulate intelligence, we have created great imitators - systems that can perform certain tasks convincingly. But, as in a hall of mirrors, these creations merely reflect our limited understanding of intelligence rather than illuminating the nature of cognition itself. The further we progress, the more we realize how far we really are from grasping the fundamental nature of mind and consciousness."

Our current approach to AI, which relies primarily on vast amounts of data and computing power, may be diverging from the nature of general intelligence.

With its remarkable efficiency, adaptability, and generalization ability, the human brain operates on principles we are only beginning to understand. While biological brains inspire our artificial neural networks, they are crude approximations that lack the sophistication and emergent properties of their natural counterparts. The sum is more, and more importantly, different than its parts.

This realization raises essential questions:

  • Are we on the right path to creating accurate artificial general intelligence, or are we merely perfecting a limited AI?
  • How would our understanding of intelligence and consciousness need to evolve to make significant progress toward the Singularity?
  • What alternative approaches or paradigms might we need to explore to bridge the gap between artificial and natural intelligence?
  • Could pursuing Singularity with our current methods lead us further away from our goal and create a technological dead end?

While our AI capabilities have grown impressively, our approach may be fundamentally limited. Singularity is not getting closer; it may only get closer as we grapple with the true complexity of intelligence. To make real progress, we may need to radically rethink our understanding of intelligence and consciousness, moving beyond brute computation to understand the underlying principles that lead to general intelligence and self-awareness.



Murat



If you are interested in more thought-provoking thoughts:



Arnaud Q.

Gestionnaire de Projets - IA Responsable et Gouvernance des Données

8 个月

I think this article can enrich the reflection: its about the analogy of “hallucinations”, or rather “confabulations” in LLMs. https://www.ncbi.nlm.nih.gov/pmc/articles/PMC10619792/

回复

I posted on this subject earlier today and have since rad a newspaper interview with Dr Kurzweil. My instinct is to laugh, as I can identify the very sci-fi inspirations behind all of the "immortality" concepts he cites. In fact, I don't think it's at all funny. His thinking suggests 1st that a lot of wealthy corporations and individuals are moving at speed to develop artificial evolution techniques that will have vast social implications, virtually without oversight. The idea that massively higher amounts of computational power is both necessary and infinitely available seems dubious, or even false- with the concept of hybrid consciousness, though beloved of high concept speculators, does not seem obviously rooted in reality. Is there anyone who will read this book, understand it and publish a detailed critique? I am not a scientist so do not consider myself qualified. I do know about social structures, power relationships, environmental pressures and politics- so very much have a stake in the outcome- as do we all. As I said, my instinct is not to take these speculations seriously. If it IS serious, we as a society are in very serious trouble. I wonder nobody seems as concerned as I think they should be.

回复
Grant Castillou

Office Manager Apartment Management

9 个月

It's becoming clear that with all the brain and consciousness theories out there, the proof will be in the pudding. By this I mean, can any particular theory be used to create a human adult level conscious machine. My bet is on the late Gerald Edelman's Extended Theory of Neuronal Group Selection. The lead group in robotics based on this theory is the Neurorobotics Lab at UC at Irvine. Dr. Edelman distinguished between primary consciousness, which came first in evolution, and that humans share with other conscious animals, and higher order consciousness, which came to only humans with the acquisition of language. A machine with only primary consciousness will probably have to come first. https://arxiv.org/abs/2105.10461

回复
Jacob Sten Madsen

??Recruitment/talent/people/workforce acquisition evolutionary/strategist/manager ??Workforce/talent acquisition strategy to execution development/improvement, innovation, enthusiast ??

9 个月

Does this matter if reality is that as species and how things are done most by majority of people is little more advanced than computational actions/reactions. What I mean is that whether singularity or not, mere current AI abilities enough to create huge upheaval in every possible structure and solutions having been used to date, 3/4 of what humans do and consist of is in AI/computational terms really simple.

回复
Theresa Semler PhD

Excellence and Innovation Driver | Strategic Transformation Consultant | Leadership and Performance Coach

9 个月

Here are some further questions: - What do you mean by "natural intelligence"? Human, animal? Vast fields with many different perspectives (biological, chemical, psychological, linguistic, ethical ...) on what intelligence is. And many question marks. - Is singularity a goal worthwhile pursuing? Or, just a fantasy from a mechanistic age? (Notably, the idea that man is a machine and processes of consciousness can be described as such is Cartesian...)

回复

要查看或添加评论,请登录

Murat Durmus的更多文章

  • Bias is a Mirror, Not a Bug

    Bias is a Mirror, Not a Bug

    In pursuing justice, perfection is an illusion, while participation is a moral imperative. Fairness requires more than…

    1 条评论
  • On Writing

    On Writing

    Writing is increasingly becoming a ritual, a performance rather than a personal expression. It's a kind of…

    3 条评论
  • The Gentle Cage of Certainty

    The Gentle Cage of Certainty

    Refusing to ask questions means shutting ourselves away. Not into an iron cage, but into a gentle, familiar silence…

    3 条评论
  • Me Against an Armada of AI Agents

    Me Against an Armada of AI Agents

    A single, defiant human..

    2 条评论
  • The Silent Surrender

    The Silent Surrender

    The future of AI ethics will not be decided by committees drafting guidelines. It will be decided by the silent…

    13 条评论
  • The Slow Theft of Self

    The Slow Theft of Self

    I feel exposed, like my mind is shedding layers I never agreed to lose. I am naked, not in body but in thought.

  • The Abyss of Knowing or When Expertise Feels Like Guesswork

    The Abyss of Knowing or When Expertise Feels Like Guesswork

    Many know that AI is not just a technological development but an epistemological crisis. It forces us to question not…

    2 条评论
  • The Difference Between AI Safety, AI Ethics, and Responsible AI

    The Difference Between AI Safety, AI Ethics, and Responsible AI

    Some worry about existential threats, others about fairness, and others want to avoid bad PR. AI security, AI ethics…

    5 条评论
  • From AI-Ethics to Algorithmic Conscience

    From AI-Ethics to Algorithmic Conscience

    An algorithm can calculate probabilities, but a conscience weighs consequences. If we insist on calling it 'AI ethics,'…

    8 条评论
  • From AI Agents to AI Monads

    From AI Agents to AI Monads

    The path from AI agents to AI monads reflects the philosophical development from Cartesian dualism to Leibnizian…

社区洞察

其他会员也浏览了