Where Americans stand on generative AI

Since ChatGPT’s initial release in late 2022, the impact of generative AI models has fallen somewhere between completely fizzling out and radically changing the world.

It saw an extremely rapid uptake, becoming the fastest application to reach 100 million users when it was released. But in the time that’s followed, some of the most optimistic (and most dire) predictions surrounding AI don’t seem to be materializing.

Where do Americans stand when it comes to AI? Below are five charts on the state of generative AI models and how the public feels when it comes to AI regulation.

1. The interest is still here. Search results for “ChatGPT,” “Generative AI,” and “Artificial Intelligence” haven’t dropped very much since peaking in 2023. In fact, searches for ChatGPT have remained near its peak, suggesting that it’s a tool that people keep coming back to. Is AI here to stay?
2. Use cases are mixed. Among Americans that report using AI models, there aren’t any consensus use cases for AI. A plurality uses it to search for information, while a smaller minority uses it to complete tasks for work or to organize information and data. That said, it’s a piece of technology Americans are still using.
3. Real people are still more trustworthy. Companies are still searching for a way to build products around generative AI (to sometimes mixed results). These models have seen significant improvement since they were first introduced to the world, but the trust might not be there yet. For example, people are far more likely to trust product recommendations from friends, family, experts, and salespeople over personalized AI recommendations.
4. AI isn’t a major job disruptor yet (at least on paper). Mass job replacement has been perhaps the biggest driver behind all the AI-related anxieties. So far, that hasn’t materialized. That said, the true number of AI-related layoffs is unclear. It’s very possible that companies are afraid to attribute job cuts directly to AI or may have conducted layoffs to redirect resources to invest in AI.
5. Americans distrust the government and elite. Their opinions on AI reflect that. Most Americans aren’t optimistic it will benefit people like them. They also don’t seem to trust the government to effectively oversee the development of AI. More than anything, this may be a reflection of the high levels of distrust with the system and elites.

So far, opinions on the impact of AI and how AI should be regulated seem to reflect Americans’ opinions on wealth inequality and general distrust of the government. And it also might not be long before sentiment around AI begins reflecting our deep partisan divides.

Even so, the future of AI is extremely unclear. AI companies are already running out of energy and data. We could be one major innovation away from a real “AI revolution.” Or we could be standing at the peak of the latest tech hype cycle at the precipice of an AI winter.