AI Adoption Grows While Public Sentiment Sours, Startup CEO Confirms Doubts

A startup building a new language model says its creators are unsure they love AI, even as consumer adoption climbs to a billion monthly users. Springboards CEO told the reporter, "We often say that we’re a self‑loathing AI company. We don’t know if we really like what we’re doing."

Key points

• ChatGPT reached 1 billion monthly users in May.
• Gemini logged 950 million users by July.
• 71 % of U.S. adults oppose a new AI data center, higher than opposition to nuclear plants.
• Half of U.S. adults now use a chatbot, up from 2023.
• 2,100 AI‑related bills across 50 states represent a tenfold rise in three years.

What happened

The conversation began with Springboards’ CEO describing his company’s ambivalence. The startup is developing an LLM that aims to generate a broader range of responses than mainstream rivals. The CEO’s remarks came amid a wave of public skepticism.

Pew Research reports that more U.S. adults expect AI to harm them personally and society than to help. The report notes that the pessimism is strongest among younger cohorts. Stanford University’s study echoes this, finding that over half of people worldwide feel nervous around AI products.

Gallup’s May poll found that 71 % of U.S. adults would oppose building an AI data center in their area, surpassing the 53 % who opposed a new nuclear plant. In March, an NBC poll placed AI below the popularity of ICE. Despite this souring mood, usage metrics remain strong: ChatGPT hit 1 billion monthly users in May, and Gemini trailed closely with 950 million users in July.

Pew also reports that half of U.S. adults now use a chatbot, more than twice the figure from 2023, and that one in four people use them daily. Across 38 OECD countries, a third of adults have used generative AI in the past three months.

The backstory

The tension between adoption and sentiment is not new. Social media platforms like Facebook and Twitter saw explosive growth while facing techlash. The difference now is that AI offers a broader range of applications, and the regulatory environment is tightening.

All 50 U.S. states have passed or proposed AI laws, creating a patchwork of over 2,100 bills—a tenfold increase in three years. Meanwhile, open‑source alternatives to Google, OpenAI, and Anthropic are emerging, offering potential consumer choice and market pressure.

Causes of the paradox

The paradox stems from the way companies market AI as a transformative force. Users experience convenience and productivity boosts, yet they also see the same narrative warning of societal upheaval. The high visibility of AI in headlines fuels anxiety, while the tangible benefits keep users engaged.

The correlation between high adoption and negative sentiment is strongest in the Global North, where usage is highest. In the Global South, lower adoption levels coincide with more optimism, suggesting that exposure may amplify fears.

Market implications and regulatory leverage

Regulation could become a lever for smaller players. With 2,100 bills in play, companies that comply early may gain market credibility. Open‑source LLMs can offer a counterbalance to the dominance of large incumbents, potentially giving consumers a choice that mitigates techlash.

However, the cost of compliance can be steep for startups, and the patchwork nature of state laws creates uncertainty. Companies that can demonstrate transparency and responsible deployment may differentiate themselves.

Consumer sentiment vs. adoption

The data suggest that adoption is not confined to a single demographic group. The same users who embrace AI for productivity also report discomfort. This duality indicates that the market is not saturated with a single persona; instead, users are navigating a trade‑off between utility and perceived risk.

If the narrative continues to portray AI as a looming threat, everyday users may feel compelled to push back. Companies that can shift the conversation toward practical benefits and clear safeguards may retain users while reducing backlash.

What builders and businesses should do

• Conduct transparent impact assessments before launching new AI features.
• Engage with regulators early to shape compliant frameworks.
• Offer clear opt‑in/opt‑out controls for data collection.
• Provide educational resources that demystify AI capabilities.
• Monitor public sentiment through regular surveys and adjust messaging accordingly.

The bottom line

AI adoption will keep accelerating, but public trust hinges on responsible deployment and clear communication. The next wave of regulation and open‑source competition will likely decide whether the market coexists with techlash or dissolves it.