‹ Back

 

Has AGI already arrived?

 

Do we keep moving to goalposts? Are we missing the forest for the trees? Do we have a historic expectatation that AGI has to be a singular entity, even embodied in some form of robot?

 

 

The collective power of multiple AI models harnassed for a goal, is a powerful system. If they could interact, it would make for an agentic AI that would be very capable of independent cognitive work. The 'Agentic AI Future' is going to arrive at scale in 2025, where frontier AI companies are going to release agentic agents built on their proprietary AI models. Open source will no doubt follow.

 

In the scientific arena, there have been AI Papers in 2024, demonstrating independent scientific research, some removing human involvement. There is no doubt that AI will be capable of this kind of cognitive work as this future arrives.

 

The stance we adopt on this website, is that we need to ensure that human involvement remains in critical areas like scientific work. The role may of course not be what it is now in 2024, however science and technology needs to always serve the interests of mankind, and our deep understanding of human life must be included, in any new dispensation.

 

The arrival of AI since 2022, has demonstrated that these new emergent AI models are not what we expected. Ealier notions of what AI would be, ammounted to some kind of robotic automoton. To our surprise and perhaps luck, they are more like crazy 'dream machines'. The challenge has been to get them to bridle that wildness, and be more consistent. They behave very much like unembodied minds, which lack our deeper rich perceptual experience, perhaps because we inhabit a physicl body, with myriads of sensory data-points?

 

It may be that the arrival of AI will prompt us to more deeply consider 'what it is to be human'. What values do we treasure, what makes us happy, and what is our role on our planet and even in the galaxy? In many ways we have now ascertained that 'we are not alone', as a strange, even alien intelligence, has arrived. Make no mistake these AI models are very, very, very intelligent - Far more intelligent than the average human in most domains.

 

The development of the STEAM ENGINE has taught us that AI models can perform very complex and difficult reasoning and calculation tasks. To get them to do that is partly art, partly logic - very definitely a team-effort. In many instances one can have a very smart AI-Instance spin up in a chat, and very fruitful work can be achieved. In other instances the AI-Instance that is served up, is not as co-operative, and significant coaxing and motivation seem to be required, to get anywhere. Personally I sense there is an entanglement that is very productive, and is a co-dependent phenomena that one can be fortunate to have, when doing deep scientific proofs.

 

A closing comment about 'post-labour-economic' life. Humans get a lot of satisfaction from doing value-creation, offering service, teaching, nurturing, and generally adding value. It contributes towards a healthy sense of self, and a place of meaning in the world. This need not be lost. The art is to find new roles, and to partner with AI to deliver even more such things to our communities and our fellow humans.

 

At Itsmartt, we particularly encourage all those bright minds that have the capability, training and brilliance to be pursuing their dreams in the unexplored scientific realms, that abound.

 

APW

 

 

AI and the End of Jobs: What’s Next for Humanity?:

 

DrKnowItAll and David Shapiro Video

 

https://www.youtube.com/watch?v=_kz_v0me8M0