AI as Normal Technology

We articulate a vision of artificial intelligence (AI) as normal technology. To view AI as normal is not to understate its impact—even transformative, general-purpose technologies such as electricity and the internet are “normal” in our conception. But it is in contrast to both utopian and dystopian visions of the future of AI which have a common tendency to treat it akin to a separate species, a highly autonomous, potentially superintelligent entity. 1. Nick Bostrom. 2012. The superintelligent will: Motivation and instrumental rationality in advanced artificial agents. Minds and Machines 22, 2 (May 2012), 71–85. https://doi:10.1007/s11023-012-9281-3; Nick Bostrom. 2017. Superintelligence: Paths, Dangers, Strategies (reprinted with corrections). Oxford University Press, Oxford, United Kingdom; Sam Altman, Greg Brockman, and Ilya Sutskever. 2023. Governance of Superintelligence (May 2023). https://openai.com/blog/governance-of-superintelligence; Shazeda Ahmed et al. 2023. Building the Epistemic Community of AI Safety. SSRN: Rochester, NY. doi:10.2139/ssrn.4641526.

The statement “AI is normal technology” is three things: a description of current AI, a prediction about the foreseeable future of AI, and a prescription about how we should treat it. We view AI as a tool that we can and should remain in control of, and we argue that this goal does not require drastic policy interventions or technical breakthroughs. We do not think that viewing AI as a humanlike intelligence is currently accurate or useful for understanding its societal impacts, nor is it likely to be in our vision of the future. 2. This is different from the question of whether it is helpful for an individual user to conceptualize a specific AI system as a tool as opposed to a human-like entity such as an intern, a co-worker, or a tutor.

The normal technology frame is about the relationship between technology and society. It rejects technological determinism, especially the notion of AI itself as an agent in determining its future. It is guided by lessons from past technological revolutions, such as the slow and uncertain nature of technology adoption and diffusion. It also emphasizes continuity between the past and the future trajectory of AI in terms of societal impact and the role of institutions in shaping this trajectory.

In Part I, we explain why we think that transformative economic and societal impacts will be slow (on the timescale of decades), making a critical distinction between AI methods, AI applications, and AI adoption, arguing that the three happen at different timescales.

In Part II, we discuss a potential division of labor between humans and AI in a world with advanced AI (but not “superintelligent” AI, which we view as incoherent as usually conceptualized). In this world, control is primarily in the hands of people and organizations; indeed, a greater and greater proportion of what people do in their jobs is AI control.

In Part III, we examine the implications of AI as normal technology for AI risks. We analyze accidents, arms races, misuse, and misalignment, and argue that viewing AI as normal technology leads to fundamentally different conclusions about mitigations compared to viewing AI as being humanlike.

Of course, we cannot be certain of our predictions, but we aim to describe what we view as the median outcome. We have not tried to quantify probabilities, but we have tried to make predictions that can tell us whether or not AI is behaving like normal technology.

In Part IV, we discuss the implications for AI policy. We advocate for reducing uncertainty as a first-rate policy goal and resilience as the overarching approach to catastrophic risks. We argue that drastic interventions premised on the difficulty of controlling superintelligent AI will, in fact, make things much worse if AI turns out to be normal technology— the downsides of which will be likely to mirror those of previous technologies that are deployed in capitalistic societies, such as inequality. 3. Daron Acemoglu and Simon Johnson. 2023. Power and Progress: Our Thousand-Year Struggle over Technology and Prosperity .PublicAffairs, New York, NY.

The world we describe in Part II is one in which AI is far more advanced than it is today. We are not claiming that AI progress—or human progress—will stop at that point. What comes after it? We do not know. Consider this analogy: At the dawn of the first Industrial Revolution, it would have been useful to try to think about what an industrial world would look like and how to prepare for it, but it would have been futile to try to predict electricity or computers. Our exercise here is similar. Since we reject “fast takeoff” scenarios, we do not see it as necessary or useful to envision a world further ahead than we have attempted to. If and when the scenario we describe in Part II materializes, we will be able to better anticipate and prepare for whatever comes next.

A note to readers. This essay has the unusual goal of stating a worldview rather than defending a proposition. The literature on AI superintelligence is copious. We have not tried to give a point-by-point response to potential counter arguments, as that would make the paper several times longer. This paper is merely the initial articulation of our views; we plan to elaborate on them in various follow ups.

Part I: The Speed of Progress

Figure 1. Like other general-purpose technologies, the impact of AI is materialized not when methods and capabilities improve, but when those improvements are translated into applications and are diffused through productive sectors of the economy. 4. Jeffrey Ding. 2024. Technology and the Rise of Great Powers: How Diffusion Shapes Economic Competition. Princeton University Press, Princeton. There are speed limits at each stage.

Will the progress of AI be gradual, allowing people and institutions to adapt as AI capabilities and adoption increase, or will there be jumps leading to massive disruption, or even a technological singularity? Our approach to this question is to analyze highly consequential tasks separately from less consequential tasks and to begin by analyzing the speed of adoption and diffusion of AI before returning to the speed of innovation and invention.

We use invention to refer to the development of new AI methods—such as large language models—that improve AI’s capabilities to carry out various tasks. Innovation refers to the development of products and applications using AI that consumers and businesses can use. Adoption refers to the decision by an individual (or team or firm) to use a technology, whereas diffusion refers to the broader social process through which the level of adoption increases. For sufficiently disruptive technologies, diffusion might require changes to the structure of firms and organizations, as well as to social norms and laws.

AI diffusion in safety-critical areas is slow

In the paper Against Predictive Optimization, we compiled a comprehensive list of about 50 applications of predictive optimization, namely the use of machine learning (ML) to make decisions about individuals by predicting their future behavior or outcomes. 5. Angelina Wang et al. 2023. Against predictive optimization: On the legitimacy of decision-making algorithms that optimize predictive accuracy. In Proceedings of the 2023 ACM Conference on Fairness, Accountability, and Transparency (Chicago, IL, USA: ACM, 2023), 626–26. doi:10.1145/3593013.3594030. Most of these applications, such as criminal risk prediction, insurance risk prediction, or child maltreatment prediction, are used to make decisions that have important consequences for people.

While these applications have proliferated, there is a crucial nuance: In most cases, decades-old statistical techniques are used—simple, interpretable models (mostly regression) and relatively small sets of handcrafted features. More complex machine learning methods, such as random forests, are rarely used, and modern methods, such as transformers, are nowhere to be found.

In other words, in this broad set of domains, AI diffusion lags decades behind innovation. A major reason is safety—when models are more complex and less intelligible, it is hard to anticipate all possible deployment conditions in the testing and validation process. A good example is Epic’s sepsis prediction tool which, despite having seemingly high accuracy when internally validated, performed far worse in hospitals, missing two thirds of sepsis cases and overwhelming physicians with false alerts. 6. Casey Ross. 2022. Epic’s Overhaul of a Flawed Algorithm Shows Why AI Oversight Is a Life-or-Death Issue. STAT. https://www.statnews.com/2022/10/24/epic-overhaul-of-a-flawed-algorithm/.

Epic’s sepsis prediction tool failed because of errors that are hard to catch when you have complex models with unconstrained feature sets. 7. Andrew Wong et al. 2021. External validation of a widely implemented proprietary sepsis prediction model in hospitalized patients. JAMA Internal Medicine 181, 8 (August 2021), 1065–70, https://doi:10.1001/jamainternmed.2021.2626. In particular, one of the features used to train the model was whether a physician had already prescribed antibiotics —to treat sepsis. In other words, during testing and validation, the model was using a feature from the future, relying on a variable that was causally dependent on the outcome. Of course, this feature would not be available during deployment. Interpretability and auditing methods will no doubt improve so that we will get much better at catching these issues, but we are not there yet.

In the case of generative AI, even failures that seem extremely obvious in hindsight were not caught during testing. One example is the early Bing chatbot “Sydney” that went off the rails during extended conversations; the developers evidently did not anticipate that conversations could last for more than a handful of turns. 8. Kevin Roose. 2023. A Conversation With Bing’s Chatbot Left Me Deeply Unsettled. The New York Times (February 2023). https://www.nytimes.com/2023/02/16/technology/bing-chatbot-microsoft-chatgpt.html. Similarly, the Gemini image generator was seemingly never tested on historical figures. 9. Dan Milmo and Alex Hern. 2024. ‘We definitely messed up’: why did Google AI tool make offensive historical images? The Guardian (March 2024). https://www.theguardian.com/technology/2024/mar/08/we-definitely-messed-up-why-did-google-ai-tool-make-offensive-historical-images Fortunately, these were not highly consequential applications.

More empirical work would be helpful for understanding the innovation-diffusion lag in various applications and the reasons for this lag. But, for now, the evidence that we have analyzed in our previous work is consistent with the view that there are already extremely strong safety-related speed limits in highly consequential tasks. These limits are often enforced through regulation, such as the FDA’s supervision of medical devices, as well as newer legislation such as the EU AI Act, which puts strict requirements on high-risk AI. 10. Jamie Bernardi et al. 2024. Societal adaptation to advanced AI. arXiv: May 2024. Retrieved from http://arxiv.org/abs/2405.10295; Center for Devices and Radiological Health. 2024. Regulatory evaluation of new artificial intelligence (AI) uses for improving and automating medical practices. FDA (June 2024). https://www.fda.gov/medical-devices/medical-device-regulatory-science-research-programs-conducted-osel/regulatory-evaluation-new-artificial-intelligence-ai-uses-improving-and-automating-medical-practices; “Regulation (EU) 2024/1689 of the European Parliament and o

[truncated for AI cost control]