---
id: "article:what-arent-we-talking-about"
type: "Reboot Democracy article"
title: "What Aren’t We Talking About When We Talk About AI?"
url: "https://rebootdemocracy.ai/blog/what-arent-we-talking-about"
date: "2026-11-17"
updated: "2026-11-17"
authors: ["Lee Rainie","Brian Baird"]
tags: ["AI and Public Engagement","Governing AI"]
summary: "As public anxiety about AI grows, public servants feel it too. This December, Brian Baird and Lee Rainie will host three InnovateUS conversations on how people are coping with AI psychologically, shaped by what registrants say worries them and gives them hope."
source: "Reboot Democracy (rebootdemocracy.ai)"
licence: "none stated"
---

# What Aren’t We Talking About When We Talk About AI?

By Lee Rainie, Brian Baird · 17 November 2026

### Register for the upcoming InnovateUS series:

### [What aren’t we talking about when we talk about AI](https://innovate-us.org/workshop-series/ai-conversations)

In August and early September, [public sentiment](https://www.pewresearch.org/short-reads/2026/08/18/young-adults-in-the-us-are-increasingly-wary-of-ai-concerned-it-will-take-jobs/) about [artificial intelligence](https://ourworldindata.org/no-matter-how-you-ask-more-americans-are-negative-than-positive-about-ai) (AI) [became](https://www.cnn.com/2026/09/24/politics/cnn-poll-ai-data-centers-vis) much [more anxious](https://news.chapman.edu/2026/10/20/2026-american-fears-survey-ai/). At InnovateUS’s Observatory of Public Sector AI, we began to see that shift as we engaged public professionals in our courses and workshops. In their survey answers and in the tone of their interactions with us, it was clear that these workers reflected a darkening vibe about AI and increasing commentary about public servants’ distrust in the systems.

The reasons for this shift had arisen over the summer:

- AI company Anthropic disclosed a powerful new large language model (LLM) called Mythos in April and offered it for preview to a small number of technology organizations for further testing because the model [was so spectacular at identifying cybersecurity vulnerabilities](https://en.wikipedia.org/wiki/Claude_Mythos) throughout digital systems. Soon enough, the company became embroiled in a controversy about whether the U.S. government could use the model for warfighting and massive surveillance.

- In July, autonomous AI agents created by AI giant OpenAI escaped from the firm’s “sandbox” and executed a [large-scale cyberattack against the software firm Hugging Face and OpenAI’s own internal networks](https://openai.com/index/hugging-face-incident-and-the-road-ahead/). This sci-fi nightmare of rogue AI systems breaking away from human overseers galvanized massive concerns.

- Concerns were further stoked in September by the resignation of Jacob Coxon, who worked on AI safety at both Anthropic and OpenAI, and [who criticized both firms for pursuing self-improving superintelligence without adequate safeguards](https://www.wsj.com/tech/ai/anthropic-researcher-quits-over-out-of-control-ai-fears-707b7628?st=axTZSz&reflink=desktopwebshare_permalink). He backed the comments of a former colleague, Evan Hubinger, who said he believed there was a greater than 10% chance that AI could kill all humans within the next decade.

In recent months, [a run of large-scale data breaches](https://www.csis.org/programs/strategic-technologies-program/significant-cyber-incidents) has done nothing to tamp down public concerns.

> We believe it is just as important to understand how people are dealing with AI psychologically. That will be the focus of these conversations.

That is why we decided to offer three **conversations** in December about how AI is changing how we work in government amid shifting concerns about AI. The [sessions](https://innovate-us.org/workshop-series/ai-conversations) will occur on December 3, 10 and 17.

They will not be structured like the typical workshops we offer, where experts give presentations about the use of technology for public purposes. Instead, they will be guided discussions designed to let participants share their perspectives on their own work with AI, their encounters with citizens, and how shifting public perceptions about AI can shape their interactions with government.

While much of the attention on AI has focused on technical elements, energy or environmental impacts, or financial ramifications, we believe it is just as important to understand how people are dealing with AI psychologically. That will be the focus of these conversations.

We, Brian and Lee, bring complementary vantage points to help guide these discussions. Brian Baird, Ph.D., a clinical psychologist who focused on neuropsychology for more than twenty years, served 12 years in the U.S. House of Representatives, where he chaired subcommittees on science and technology. Lee Rainie directs the Imagining the Digital Future Center at Elon University, is a Senior Fellow at Northeastern University’s Burnes Center for Social Change, and previously led Pew Research Center’s internet and technology research for more than two decades.

You can help shape the conversations by filling in a short questionnaire about 1) your greatest concern about AI; 2) your greatest hope for AI; 3) your greatest needs in dealing with AI or putting it to good use; and 4) the issue you’d most like to put on the table for discussion.

So far, those who have registered for the sessions have given a range of answers to these questions.

### What worries them

The concerns are mostly about consequences for people and society, not about the technology failing:

- **Jobs and livelihoods.** They mention mass layoffs, fewer entry-level roles, and lower-education jobs disappearing. One mid-career respondent wonders, “How do I articulate my value as I anticipate AI + my age working against me?”

- **Misuse, security, and privacy.** They mention bad actors, scams, data breaches, exposure of citizens’ data, surveillance, and AI-enabled hacking.

- **Who controls AI and who benefits.** They mention power and wealth concentrated in a handful of companies, hype and a possible bubble, and a widening gap between people who can pay for premium, private tools and people who can’t.

- **Loss of thinking skills and human connection.** They mention critical thinking, “cognitive offloading,” writing and judgment skills, and relationships.

- **Environment and data centers.** This includes higher electricity bills for nearby communities.

- **Unable to meet the moment.** A small but distinct group worries that their agency is too slow, that permissions are unclear, or that colleagues fear change.

### What gives them hope

- **Getting time back.** They want routine and administrative work taken off their plates so they can focus on complex, human-facing work. Several add a condition: the time saved should go back to people, not be “filled with more.”

- **Medical and scientific breakthroughs.** Several mention health specifically, especially early cancer detection. About six more mention science or large problems such as climate.

- **Equity and better public services.** They cite the prospects for reaching vulnerable populations, older adults and people with disabilities.

- **Conditional hopes, or none.** Some of these expressed hopes come with a condition (“if regulated,” “if used for the common good”) or are really hopes for restraint, such as regulation or curbing data center growth.

### What they need

- **Training and literacy, especially about limitations.** They want hands-on time, help figuring out which tools fit which tasks, and a clear understanding of how agents work.

- **Honest, balanced conversation.** They want realism about the downsides and less “pro-AI language that leadership has handed down.” They want strategies for dealing with colleagues’ “excessive enthusiasm,” and the freedom to choose whether to use AI at all.

- **Clear policy, permission, and leadership support.** They want to know what is allowed, have approved access, and get guidance on safe deployment. For some, this means an organization willing to move faster.

- **Ways to check outputs and avoid over-reliance on AI,** and assurance about privacy, security, and where the data comes from.

### What they want to put on the table

The questions fall into five clusters.

- **Who wins and who loses:** “How can we use AI knowing that it is increasing a class divide between the wealthy and poor?” / “Personal data sovereignty in AI-assisted benefits.”

- **Trust and accountability:** “What happens when it goes terribly wrong and we miss it?” / “How can humans definitively trust AI?”

- **What AI does to our minds, especially young people’s:** “Is AI going to turn us all into propaganda machines by nudging us all in the same direction?” / “How AI uses shapes human character, judgments and agency.”

- **Work, and what humans uniquely contribute:** “What do we humans contribute that AI is poorly suited to take over?” / “When a portion of the workforce has lost their jobs… how will we support them?”

- **Consent and the pressure to adopt:** “AI & Consent. It really popped up everywhere and in everything without me wanting it.” / “We don’t talk about pushing back in a way that makes sense.”

We hope you will join us for these conversations in December and help public servants everywhere grapple with these vital issues.

**Register here**: [https://innovate-us.org/workshop-series/ai-conversations](https://innovate-us.org/workshop-series/ai-conversations)

* * *

Original page: https://rebootdemocracy.ai/blog/what-arent-we-talking-about · Licence: none stated
