On a recent weekend afternoon in San Francisco, about 20 people sat in a loose circle in an empty office. There was a high school student, a few undergraduates and postdoctoral fellows, some startup founders, and a neuroscientist. In the city that is at the center of the artificial intelligence boom, they were not meeting for a hackathon or an investor pitch, but for a conversation about the challenges in collecting neuroscience data, and the need for shared research.
The gathering, which lasted over two hours, was an AI Salon. Founded in San Francisco in 2023 for conversations on the meaning and impact of AI, the nonprofit organization also has chapters in Bengaluru and Lagos, run by volunteers. The meetings, limited to roughly 20 people, have discussed topics including AI and climate, space, education, consciousness, art, and intimacy. The aim is to provide a safe space for different perspectives, Sharat Satyanarayana, head of the AI Salon in Bengaluru, told Rest of World.
“There’s so much going on in AI, it’s overwhelming, and everyone feels like they cannot keep up,” said Satyanarayana, a startup founder who organized the recent AI Salon session in San Francisco during a visit. “The idea is to bring together folks from the AI and non-AI worlds for more thoughtful perspectives of the broader discussions around AI.”
Addressing the knowledge gap
As AI adoption increases worldwide, concerns about the impact on jobs and other consequences are also rising. With governments slow to act, and a handful of American companies dominating the conversation, AI users like Satyanarayana are connecting with others for assurance and advice. Groups such as the AI Salon and the AI Collective, which began in Silicon Valley, are now global, drawing a growing number of participants.
The AI Collective, a nonprofit, was set up in Silicon Valley in 2023, shortly after the launch of ChatGPT. It** has more than 200,000 members and about 200 chapters in 50 countries. With a newsletter and in-person and virtual events, the aim is “to democratize, build trust, and coordinate how society navigates the rapid acceleration of technological progress,” according to its mission statement. Apart from a core team of about a dozen paid employees, the rest are volunteers, and while most members are in the U.S. now, the aim is to have a majority outside the U.S., chief operating officer Catherine McMillan told Rest of World.
“We want to encourage hyperlocality, building the most local communities around AI,” said McMillan, a former management consultant. “The goal is to remove anxiety around AI. There’s a lot of confusion and existential questions around its impact; we are not all going to agree, but we should have the hard conversations.”
At AI Collective’s overseas chapters, members also tackle the challenges of building systems without the resources of Silicon Valley. Kenya has among the lowest AI adoption rates globally, of less than 10%. So one of the goals of the 3,000-strong Nairobi chapter is to embrace AI and “not leave anybody behind,” organizer Francis Kangure told Rest of World. While there is great excitement about the technology’s potential benefits in education, agriculture, and healthcare, there are concerns about data privacy and job displacement, he said.
“There is a huge knowledge gap regarding AI and AI adoption in Africa,” he said. “We see the AI Collective as a key driver for community-led innovation, and a bridge between local talent and global opportunities.”
“A quiet worry”
At the Bengaluru chapter of the AI Collective, its more than 10,000 members are witnessing the changes sweeping through the country’s tech hub, Abhishek Kumar, the chapter head, told Rest of World.
“Most of the AI story in India right now is top-down: a national push, big corporates, large IT firms. What’s missing in the middle is the connective tissue for the individual builder,” said Kumar, an AI engineering adviser for a global tech services firm.
India is the second-biggest market for Anthropic’s Claude, and “there’s a quiet worry about whether India ends up only consuming and wrapping models built elsewhere, instead of building foundational technology of its own, and whether people who don’t speak English or don’t live in a metro get left behind,” Kumar said.
“A community [like ours] can’t fix the big structural questions … but can provide access — to the right tools, to people who have already done the thing, to a first audience for what you’ve built,” he said. “None of that solves job displacement, but it gives people agency.”
Collectives, which are generally community-led grassroots organizations, are seeing renewed interest as a form of resistance to big tech companies. In setting up the AI Salon in San Francisco shortly after the launch of ChatGPT, Ian Eisenberg sought “deep conversations” about human-centered AI.
Salons — made popular in France and other parts of Europe in the 17th and 18th centuries — were private social gatherings, where guests discussed literature, art, philosophy, science, and politics. Hosted by wealthy, well-connected patrons, salons helped spread influential ideas during the age of Enlightenment.
After the first AI Salon discussion in a friend’s apartment, Eisenberg has run about 40 salons, with the ground rule that everyone keeps an open mind, and speaks at least once, he told Rest of World.
“There are a lot of hackathons and startup meetups, but they are excluding a lot of people, and we want broader civil society participation,” said Eisenberg, a neuroscientist and AI safety expert. “There is low trust in governments, and people are having to figure this out themselves. Education and talking about it with others is not a panacea, but it’s a solution.”
“Reproducing the same power asymmetries”
Still, members of an AI Salon in Bengaluru or an AI Collective in Nairobi tend to work in technology, and are a part of the elite networks that have access to these spaces, Payal Arora, professor of inclusive AI cultures at Utrecht University, told Rest of World. So some of these gatherings are “reproducing the same power asymmetries they claim to address,” she said.
“We desperately need more public spaces where ordinary people can debate how AI is reshaping society, but I am skeptical of initiatives that originate in Silicon Valley or San Francisco,” said Arora. “It takes considerable economic, social, and cultural capital simply to inhabit those spaces. … If we truly want inclusive AI communities, they cannot simply globalize Silicon Valley’s conversations.”
Worldwide, another movement is also growing: those opposed to AI’s impact, who want better safeguards, a bigger share of the profits, or a halt to the so-called AI race. Groups such as San Francisco-based AI and Stop AI are expanding globally, organizing via group chats on Signal, staging street protests, and demonstrating outside the offices of Anthropic, OpenAI, and Google DeepMind.
The AI Salon in Bengaluru can seem cut off from those concerns, but it is still an essential part of the conversation around AI, said Satyanarayana.
“The gatherings are not representative of the entire population’s preferences and points of view,” he said. “I would hope that they can help people think differently about AI. But it’s a small subset of people.”