OpenAI has much larger goals than building an amazing, but let’s face it, kind of cute chat interface that helps college kids cheat. Their true goal is developing advanced general intelligence (AGI) that benefits all of humanity. With the recent release of their principles and mission statement, it’s clear that OpenAI is serious about its trajectory towards AGI, and that the company views its GPT-3 models, including ChatGPT, as steps along the continuum towards AGI. As we head towards this future, it’s important to consider the impact that OpenAI’s plans will have on jobs and the economy in the short and long term.
The advancement of AI has already begun to disrupt many industries and the workforce, and this trend is only expected to continue as AI becomes increasingly sophisticated. OpenAI’s mission to create AGI systems that are smarter than humans, increase abundance, boost the global economy, and enhance scientific knowledge has the potential to bring about both positive and negative consequences for the job market and the economy. With this in mind, let’s discuss the impact of OpenAI’s plans and explore the potential consequences of its stated trajectory toward AGI.
The mission is to ensure that the development of artificial general intelligence (AGI) benefits all of humanity. It’s important to note that the aim is to create AI systems that are smarter than humans and use this to increase abundance, boost the global economy, and enhance scientific knowledge. The ultimate goal is to provide everyone with access to cognitive support and multiply human ingenuity and creativity. However, the risks of AGI such as misuse, accidents, and societal disruption, must be addressed to ensure its positive impact on society.
The stated principles behind the mission of artificial general intelligence (AGI) are to “empower humanity” and maximize its potential. The aim is to create a future that is not a utopia but maximizes the good and minimizes the bad, and that AGI should be an amplifier of humanity. To successfully navigate the massive risks involved, OpenAI’s approach is to continuously learn and adapt, by deploying less powerful versions of the technology, to minimize the risk of failure. Translation: ChatGPT’s power is purposefully held back (you ain’t seen nothing yet.)
OpenAI’s portrayal of their mission conveys a sense of slowness – a deliberate, careful, and thoughtful roll-out of increasingly advanced technologies. If OpenAI were in a competitive vacuum, I would be more inclined to believe their turtle-paced development plan. However, Google and others will likely force a greater acceptance of disruption, which may lay a much bumpier road.
Gradual, Iterative Integration of AGI: OpenAI’s gradual approach to AGI integration attempts to allow for a smoother transition into the economy and workforce. This approach may provide workers with time to reskill and adapt to the changing landscape while also giving policymakers and institutions the time to understand the impacts of AI on jobs and the economy. If everything goes well, the result is a more regulated deployment of AGI that ensures the benefits are widely shared.
Increased Usage of AI: OpenAI believes that the widespread of AI will be a positive and productivity multiplier. AI could lead to increased economic growth and new job opportunities in research and development, science, and medicine. Overall increased productivity would then in theory lift individual economic circumstances and standards of living.
Aligned and Steerable AI Models: OpenAI’s focus on creating aligned and steerable (an interesting and carefully chosen term in and of itself, who is steering?) AI models could lead to a more responsible deployment of AI, reducing the risk of job displacement and economic disruption. By ensuring that AI systems align with society’s values, OpenAI aims to minimize negative impacts on the workforce and the economy.
User Discretion in AI Behavior: Allowing users to modify and tune the behavior of AI systems could lead to increased customization and flexibility in AI deployment, potentially creating new job opportunities in areas such as AI customization and integration.
Global Conversation on AI Governance: OpenAI’s hope for a global conversation on the governance, distribution of benefits, and sharing of access to AI systems could result in a more responsible and ethical deployment of AI, reducing the risk of job displacement and economic disruption. This could also create new job opportunities in areas such as AI policy and regulation.
Aligned Incentives with a Good Outcome: OpenAI’s alignment of their incentives with a good outcome, including a clause in their charter about advancing safety, a cap on returns for shareholders, a nonprofit governance structure, and the sponsorship of a UBI experiment are interesting, but rather telling signs of the possibility of radical job displacement.
The development of AGI will have a profound impact on the future of humanity and the job market. As AGI systems become more advanced, they have the potential to dramatically alter the workforce and the economy, leading to job displacement, economic disruption, and concentration of power. However, with proper governance and regulations, AGI could also result in new job creation, economic growth, and improved quality of life.
Importance of Public Consultation: To ensure that AGI benefits all members of society, OpenAI says that they are committing that major decisions about AGI be made through a democratic and transparent process involving public consultation and input. This will help mitigate the potential for negative impacts on jobs and the economy and ensure that the benefits of AGI are widely and fairly shared.
The Potential of AGI to Revolutionize Industries: AGI has the potential to revolutionize many industries, including science, through its ability to accelerate research and bring about new discoveries. However, the impact on jobs and the economy will depend on how the technology is used and regulated. There is potential for job displacement, economic disruption, new job creation and economic growth.
Navigating the Risks of AGI: The transition to a world with intelligence augmentation and superintelligence is one of human history’s most important, hopeful, yet frightening projects. The development of AGI must be guided by responsible and ethical principles to minimize the risks and ensure that the benefits of AGI are widely shared.
Aligning AGI with Human Flourishing: By aligning their incentives with a good outcome and working towards AGI systems aligned with human flourishing, OpenAI hopes to contribute to a future with increased abundance, economic growth, and improved quality of life. The impact on jobs and the economy will depend on the successful development of these systems and how they are used and regulated.
In the coming years, the progression of AI toward AGI may follow an exponential curve. It is important to understand that AI breeds more AI, as the systems and models learn to advance themselves and contribute to new data models. The result could be shocking advancements, as opposed to the slow and progressive, iterative plan that OpenAI has communicated, at least in part, for the purpose of public relations. Here are some broad guesses about what the next 3 years could look like.
2023: GPT-3 Apps Everywhere. GPT-3 powered applications will proliferate rapidly and already have. Specialized apps will be launched daily by startups looking to disrupt particular industries. The earliest apps will be front-ends for the GPT-3 API, tuned models for highly specific use cases. These will center around creating specialized documents, usually reserved for human specialists, including docs for legal, marketing, website copy, creative text, speech analysis, recruiting, real estate, and more. Large organizations will use AI to understand their vast databases and reporting as Microsoft pushes Azure services to the enterprise.
2024: Meta Apps and GPT-4 Operating Systems. If 2023 will create a universe of GPT-3 powered applications, the following year should focus on the centralization and connectivity of these applications, like a cycle of bundling after an unbundling. AI will be integrated into organization-wide servers, apps will talk to each other, and AI app stores may launch. For example, about talking to each other, the marketing text generated by 2023’s copywriting AI model may talk to another startup’s legal AI model to determine the legal risks of its marketing text for a particular country. 2024 could look like advanced professionals gathered around a desk, talking, and sharing their knowledge about their particular areas of strength. Multiple artificial “intelligences” start to converge a bit to create the beginnings of decision-making.
2025: It Learns to Code and Multi-Modal Explodes. A few years in, we have a proliferation of AI-powered apps, and AI has become the base layer for all of the technology – a baked-in constant that is always present. GPT-3 can output bits of code in 2023, and perform a certain level of bug-catching. However, by now, the competitive pressure between Google and Microsoft-backed OpenAI, as well as other entrants, will, due to competitive pressures, cause acceleration in the move toward autonomous AI development – when AI begins to be unleashed to develop itself. Because of this self-development, I believe this is also when you seriously impact very advanced fields such as pharmaceuticals, energy, medicine, and science. Also, in 2025, AI will become truly multi-modal, with the ability to seamlessly input and output all forms of media, including music, art, text, and code.
OpenAI’s mission to develop advanced general intelligence (AGI) has the potential to greatly impact the job market and the economy, both in the short and long term. As AI continues to disrupt many industries and the workforce, it’s important to consider the potential consequences of OpenAI’s trajectory toward AGI. The company’s focus on creating aligned and steerable AI models, developing new alignment techniques, prioritizing AI safety and capabilities, and encouraging a global conversation on AI governance are all positive steps, but there is a clear and acknowledged message that the impact will be both profound and significant.
In terms of the good and bad – pick how you’re feeling today. 2030 may look like a creative utopia with no limit to actualization and expression – our stories are all told and heard, and we listen to Spotify playlists sung just for us. Or something very different. I certainly won’t be able to figure it out ahead of time.