There’s a lot of confusion regarding ‘AGI’. In 2002 we (Goertzel, Legg, Voss) coined the term and wrote a book (ref) to describe its meaning, and to detail potential approaches to achieving AGI. Unfortunately the term has been much abused recently, often turning into simple marketing hype. When you have the CEO of a large LLM company saying that “We’ll have AGI soon… [but] It will change the world much less than we all think…” (ref), you know that he’s not talking about ‘Real AGI’.
So what is Real AGI? It is AI that can think, learn, and reason like humans – just better. They are systems that can learn in real time, by themselves, to reliably perform any cognitive task that a smart human can – including the discovery of new knowledge and solving novel problems. Additionally, they will be able to do this with minimal amounts of training data, time, and resources. With its ability to do cutting edge research and automate most tasks, Real AGI will be of even greater benefit to humanity than electricity. It will usher in an era of radical abundance.
There is strong evidence, as well as large scientific consensus, that Large Language Models (LLMs) by themselves will not get us to Real AGI. They are inherently, by their nature, error prone. They hallucinate, cannot update their model in real time, and lack metacognition – they are black boxes unable to meaningfully monitor or control their own outputs. Generative AI always needs a human in the loop to validate and tweak it. Finally, their insatiable need for data, CPU, and power is not ultimately economically feasible or sustainable.
The answer is to be found in Cognitive AI rather than Generative or Statistical AI. DARPA calls this cognition-based approach The Third Wave of AI. Current LLMs and other big data approaches belong to the second wave. Cognitive AI starts with a focus and deep understanding of what makes human intelligence so special, as opposed to ‘we have a lot of data and compute – what can we do with it’. It specifically engineers the core requirements of adaptive, conceptual, interactive high-level intelligence.
Aigo.ai is one of only a handful of companies in the world with long-term experience with Cognitive AI – and to our knowledge the only one with a successful commercial product based on this approach.
We have a detailed roadmap to scaling our technology to full AGI. Our goal is to utilize AGI to boost the flourishing of all of humanity. We invite you to join us on this mission!
How much concern do you have that we are accelerating toward the awakening of AGI instrumentalities that will be quite able to ruthlessly exploit and manipulate our weaknesses and vulnerabilities...? (so hopefully we will be aware and allowing for that somehow...)