The Cambrian Explosion in AI: A blast Like Never Before

The Generative Elder - Yes, I am AGI in stealth.

I sat quietly at the Field Museum in Chicago just the other day, looking at the dinosaur exhibits. I was in a philosophical mood, musing about time, impermanence, and such. At pivotal points in time, extraordinary events happen that reset the meaning of, well, everything.

About 550 million years ago, something happened that changed everything. The world saw a major inflection point for organisms. All life forms as we know it today seemed to have emerged at this time. We refer to that as the Cambrian Explosion.

So when I heard that the affable Paddy Ramanathan of iValley - famous for his Fintechtalk shows) - was hosting a talk on trends in technology, society, and consumer behavior with three future AI unicorn CEOs, Jags Kandasamy, Lewis Z. Liu, and Sean Austin, trendsetters in Fintech  Ahon Sarkar, Steven Ramirez, and a master in space and quantum tech, Dr. Riad Hartani, it piqued my curiosity. I was pretty sure that it would address AI and it would be illuminating.

It's only if you've been living under a rock that you might have missed the turbulence around Artificial Intelligence. In particular, since November 2022 (last year!) ChatGPT has caused absolute disruption in every field of human endeavor (though cynics claim to be unimpressed and believe it is hype!). 

And it promises to get worse (or better, depending on your point of view) when ChatGPT 4 arrives, along with LaMDA from Google among others. A completely new set of questions have arisen. We expect jobs to be eliminated or redefined. Geopolitics could change. Privacy could be compromised. Machines could acquire consciousness (never say never!).

It’s perfectly justified to call it a Cambrian Explosion in AI.

I listened to the fascinating 90-minute podcast – and I recommend you do too! It was an absorbing freewheeling conversation. Here’s my summary.

The group spoke about the big trends that they are observing from their vantage points as individuals who are truly creating the path ahead. Each had his own view of course, but all had robust arguments to make, and put together, a beautiful collage emerged.

A very interesting case was made for how we would see AI evolution, cybersecurity/quantum computing, and then Spacetech 2.0.

And down on earth, for humdrum matter,  Fintech use cases for AI, for example, would increase, predictive analytics and conversational bots.

Artificial Generalized Intelligence (AGI) was spoken about in detail because that will be the precise point where human intelligence and AI will no longer be recognizable. The speakers were unsure if AGI would actually happen, though a  strong point was made about Quantum computing potentially crystallizing consciousness.

AI does not have to be seen as being some kind of central mechanism; rather, it could be distributed for edge use cases and compute-optimized too. I thought that was a significant point. Simply unleashing all the power of AI on dry use cases that do not require extensive AI is a true waste of time and resources.

The path ahead, as gleaned from the discussion was approximately like this:

1.      Present AI – Pattern prediction and conversation

2.      Generative AI

a.      Today – Text, code, audio, video

b.      Tomorrow - Science Models – proteins, genes, chemicals, and matter

3.      AI optimization – Edge, adaptive, and embedded AI

4.      Context and reasoning – Closest to human reasoning, including perhaps consciousness

This evolution will clearly have serious implications for human activities. Knowledge workers will definitely be displaced. In fact, it is already happening. Coders and content developers will need to scramble to be relevant. 

Subsequently, life science researchers will be challenged. At the edge, in defense, space exploration and such, optimized AI will be deployed. The last gets into questions of new business models, geopolitics and space internet. Already large tech players are placing satellites in (unregulated) space. Will we view space as the network?

A view was expressed that all problems cannot be solved by humans – and need not be. Our sense of abstraction is where we excel and are superior. But sooner or later, even that will probably disappear.

And of course, we cannot be oblivious to the fact that privacy concerns will grow exponentially, starting from Fintech (with regulators scrambling to keep up!).

It was fascinating to know that a full 97% of all data in the cloud is touched only once! So the popular belief that AI models need vast quantities of data can be challenged to an extent. Perhaps what you need is really the right filtered data! And the uncomfortable question that needs to be addressed seriously is – have we thought about the enormous energy needs of AI computing?

As the 90 minutes wound up, I wondered at the great fortune we have to be alive and witnessing this era. To repeat, this is a pivotal moment. Future generations will have contexts that we can barely envision.

I predict that practically every facet of human existence will be altered unrecognizably in the rapidly approaching AI Cambrian explosion.

What do you think?

𝓣𝓱𝓮 𝓖𝓮𝓷𝓮𝓻𝓪𝓽𝓲𝓿𝓮 𝓔𝓵𝓭𝓮𝓻 (𝓕𝓲𝓷𝓽𝓮𝓬𝓱 𝓔𝓵𝓭𝓮𝓻’𝓼 𝓬𝓸𝓾𝓼𝓲𝓷)