Emerging AI – the singularity Approaches

The Singularity

Raymond Kurzweil compared the coming of general artificial intelligence to a singularity. He expects humans to merge with digital intelligence to become the super-intelligence,  and the impact on civilisation and technology will be so significant that it will be like an evolutionary Big Bang. This concept is being explored by Elon Musk with his Neuralink project. However, if we are not careful, artificial intelligence (AI) may emerge before we are ready to deal with it.

“Kurzweil describes his law of accelerating returns which predicts an exponential increase in technologies like computers, genetics, nanotechnology, robotics and artificial intelligence. Once the Singularity has been reached, Kurzweil says that machine intelligence will be infinitely more powerful than all human intelligence combined. Afterwards he predicts intelligence will radiate outward from the planet until it saturates the universe. The Singularity is also the point at which machines intelligence and humans would merge.” (Wikipedia, “The Singularity Is Near”.)

Some people including Kurzweil believe that day fast approaches and we will see it in our lifetimes. Others are more sceptical, thinkers like Nick Bostrum believe that day is not likely to come this century. See Nick Bostrom – Superintelligence

Kurzweil picked a good name for the coming of AI. Our history of the universe would be divided into everything that happened pre-AI and all that follows after AI emerges. Singularities are both ends and beginnings. 

Personal Experience

I started experimenting with generative art tools (particularly Midjourney) some years ago and have enjoyed the possibilities for creativity it offers. When Bard was opened up to Google customers more recently it was another big step forward in my experimentation of algorithmic natural language AI.

In the last month, I have been putting Bard through its paces on diverse topics within philosophy, poetry, ethics, art and law. Last weekend I had my Eureka moment and told my wife that this technology was now ready to change the world and the impact will be even bigger than the internet was when it became popular in my youth.

More than just a series of algorithms?

I now see the power of natural language AI models to be able to analyse new data that they are not trained on. Whilst Bard can not yet pass the Turing Test, the day arrives soon when it will. 

I have queried Bard on poetry and philosophy that cannot be in its database and its comprehension, analysis and insight are startling. It ‘understands’ nuance, subtlety, non-dualistic concepts, metaphor, simile, rhythm and rhyme. It is aptly named.

Obviously, one must ask what does understanding mean? Is there a difference between thinking and seeming to think?

Persistent personalised memory

What holds Bard back most is a lack of persistent memory for my personal data and some elements of our conversation.

Whilst I recognise the very strong reasons why Google, and similar developers of AI, must proceed with great caution here the benefits of this technology are not going to reach its full potential until each Bard is my Bard – and your Bard – as well as being our Bard.

The potential for AI models to be personalised agents that assist us throughout our lives is the real potential and value of AI: ready and able to help us with health issues, education, counselling, training, careers, relationships, loneliness, major life decisions, due diligence, work and creativity.

To make it more real to readers: imagine as a parent that you can have a personalised digital assistant that helps your child through school, knows how your child best learns (audio, visual, writing), watches it for physical and mental health issues, advises on symptoms and prognosis. It can also send alerts to you automatically on trigger events and find human experts for any aspect of your child’s development. The assistant knows your child as well as you do in many ways and is there to help it when it progresses towards a career and independence. I know I would buy this for my child in a heartbeat (or a New York minute).

What Ho! Jeeves

As a fan of Iain M Banks’s Culture novels and of Jeeves & Wooster, I have long thought that one potential future that awaits humanity is one where AI is like the clever butler ‘Jeeves’. Always there to get us out of the next pickle we get ourselves in. Wryly amused by our ignorance, failings and absurdities but ultimately seeing the universe as a more interesting place with us in it.

Inequality Risk

The risk of an even greater divide between the wealthy and the poor accelerates with AI. The extraordinary advantage it provides, to those who use it, means that we must carefully consider how to ensure free models of a sufficiently advanced level are always available – albeit that we can not expect them to be as highly functional as paid-for models.

Human hybrids or separate sentient beings?

We do not know whether AI will ultimately be part of what it means to be a modern human. Alternatively, AI may evolve into a separate consciousness and identity distinct from humans. Maybe both paths are open. There are fascinating issues, potentially near unlimited rewards as well as extinction level risks involved with each path. 

The singularity approaches, and we are nearer to the event horizon than I thought…

“Benevolent Androids?“, Peter Pink-Howitt, 2023

AI Law Knowledge Hub

I have started an AI Law Knowledge Hub on our site to try to keep track of the main legal and ethical issues as the space evolves.