Christopher P. Peters, M.Ap.Stat. Gentry Hanks, PhD OpenAI ChatGPT-4
As AGI economizes general human intelligence, people will scramble to acquire relatively specialized intelligence since it is now much more scarce relative to general intelligence (and both are highly demanded).
Although there are still debates about the existence of artificial general intelligence (AGI), I believe it exists and my focus now is on artificial specialized intelligence (ASI). As AGI becomes more ubiquitous, it will supplement and relieve human suppliers of general intelligence by gaining specialized intelligence and knowledge, spanning technical to domain knowledge. This will change occupations in response to incentives transmitted by changes in wages, leading to companies increasingly using AGI agents for work. For example, at least one insurance agency is already using an AI chatbot to answer customer questions, making general customer service agents cost-ineffective in the long run (as opposed to specialized folks who’ll be even more in demand). Both AGI and ASI speed up human knowledge acquisition that will speed up this process as these technologies are increasingly deployed.
The importance and value of specialized intelligence, both human and artificial, is now higher than ever. Stochastic thinkers, in particular, will be in high demand since these machines heavily benefit from that kind of thinking. This overarching societal trend explains the explosion of development and application of AI, data science, machine learning and probabilistic programming languages. Specialized intelligence, and especially stochastic thinking, will be critical for optimizing both the human good and financial profit that stems from these technologies. This raises the value and demand for such intelligence and also automating specialized intelligence, hence my new additional focus on artificial specialized intelligence, beyond just AGI (and the specialized intelligence it depends on for profitable deployment).
In business, specialized knowledge is critical to success, especially in the technology industry. This is because the industry is highly competitive, and companies need to continually innovate and adapt to remain competitive. For example, companies need to understand their customers' needs and preferences, and develop products and services that meet those needs. This requires not only technical knowledge, such as knowledge about programming languages, but also practical knowledge, such as the ability to apply one's understanding in real-life situations.
Specialists in our society, such as industry practitioners and adjacent academics who have specialized in stochastic thinking, will be required to guide and construct AGI agents and agent teams. This means statisticians, economists, econometricians, reinforcement learning and game theory practitioners will have a significant role in the development of ASI. Our human specialized intelligence will be critical for optimizing both the human good and resulting financial profit that stems from these technologies. We will also see those adjacent to stochastic thinking move over from deterministic thinking to this new paradigm. For example, software engineers will need to learn how to think in stochastic terms like a statistician or game theory practitioner would (to match the reality that human behavior is mostly stochastic not deterministic).
Who are the specialists in our society? Well, academics for number one, but also industry practitioners like myself that have specialized in stochastic thinking in the business domain. Statisticians, economists, econometricians, reinforcement learning and game theory practitioners will be required to guide and construct AGI Agents and agent teams (as artificial specialized intelligence is still largely in the future). I think our human specialized intelligence will be critical for optimizing both the human good and resulting financial profit that stems from these technologies.