Movement in the USA defends absolute freedom for AI – 12/14/2023 – Tech

On a Monday night last month, a few hours after OpenAI held a developer event in downtown San Francisco, hundreds of artificial intelligence (AI) aficionados packed a three-story nightclub several blocks away to celebrate. a more relaxed and less corporate vision of the future of technology.

Under colorful lights and screens showing anime footage, the mostly young and male crowd danced to the tunes of Grimes, who is best known in technology circles as Elon Musk’s ex.

A large banner on the wall read “Accelerate or Die.” Another sign showed a diagram of an AI neural network emblazoned with the motto “Come and Take It.” An AI startup distributed promotional flyers that said “the messenger to the gods is here for you.”

The party was called “Keep AI Open” and was a launch party for Effective Accelerationism, one of the strangest and most interesting splinter groups to emerge after the AI ​​boom.

Effective Accelerationism (often abbreviated as “e/acc”, pronounced “e-ack”) is a loosely organized movement dedicated to the unbridled pursuit of technological progress.

The group believes that artificial intelligence and other emerging technologies should be allowed to advance as quickly as possible, without barriers or gatekeepers in the path of innovation.

The group formed on social media last year and has bonded in Twitter spaces and group chats through memes, late-night conversations and shared contempt for the people they call “decels” and “doomers” — the people who care. with the safety of AI or regulators wanting to slow it down. They also have offline activities, with parties and hackathons in the San Francisco Bay Area and beyond.

Effective Accelerationism began as an ironic response to an older, more established movement — Effective Altruism — that has become a major force in the AI ​​world.

Effective Altruism, known as EA, began by promoting data-driven philanthropic giving, but in recent years has become concerned about the safety of AI and promoted the idea that powerful AI could destroy humanity if not contained.

The battle between the e/accs and the Effective Altruists is one of many quasi-religious splits emerging in the San Francisco AI scene these days, as insiders argue about how quickly the technology is progressing and the what should be done about it.

e/accs prefer the “foot on the gas” approach. Its adherents prefer AI software to be open rather than controlled by large corporations, and unlike Effective Altruists, they do not see powerful AI as something to be feared or guarded.

They believe that the benefits of AI far outweigh its harms and that the right thing to do with such an important technology is to get out of the way and let it advance.

Some of the ideas espoused by e/accs, such as their opposition to regulation, are part of the standard techno-libertarian gospel.

Others resemble principles of older Silicon Valley subcultures, such as Transhumanists and Extropians, who also valued progress and resisted attempts to contain technology.

The movement also takes inspiration from the works of British philosopher Nick Land, who wrote years ago that the accelerating forces of capitalism and AI would eventually collide in a “technocapital singularity,” a point at which technology would outpace our ability to contain it. Land has recently fallen out of favor for endorsing far-right ideas about race and authoritarianism.

In a manifesto posted online last year, e/acc’s founders — all using internal pseudonyms like “Bayeslord” and “Based Beff Jezos” — described their goals in grandiose, bombastic terms, writing that their goal was to “inaugurate the next evolution of consciousness, creating unthinkable next-generation life forms.”

Most people, of course, want to maintain the life forms we already have, and critics of e/acc are uncomfortable with the idea that we should surrender and allow robots to overtake us.

Peter S. Park, an AI researcher at MIT and director of Stakeout.AI, an AI safety advocacy group, told me that he considers e/acc “a dangerous and irresponsible ideology inspired by the replacement of humanity with AI.”

I first heard about e/acc about a year ago. At the time, the movement seemed to consist mostly of bored tech workers who gathered late at night to have deep conversations about politics and philosophy, discuss the news, and complain about the emerging narrative that AI was an imminent threat to humanity.

“A lot of my personal friends work on powerful technologies and end up getting depressed because the whole system tells them they’re bad,” said Guillaume Verdon, a 31-year-old French-Canadian physicist who once worked at a Google experimental lab, in a Twitter Space earlier this year, which was transcribed by someone who participated. “For us, I was thinking, let’s create an ideology where engineers and builders are heroes.”

Initially, I dismissed the movement as a fringe novelty — a group of Twitter-addicted techies with persecution complexes turning Ayn Rand into edgy memes.

But a few months later, tech luminaries like Marc Andreessen, co-founder of the venture capital firm Andreessen Horowitz, began appearing on e/acc Twitter Spaces and proclaimed that they, too, believed in Effective Accelerationism.

Andreessen’s Platform X profile now includes “e/acc” and he listed Based Beff Jezos and Bayeslord as two of his “patron saints” in the “techno-optimist” manifesto he published in October.

Garry Tan, president of influential startup incubator Y Combinator, signaled his support for e/acc. Sam Altman, CEO of OpenAI, responded to a tweet from Based Beff Jezos and joked “you can’t speed me up.”

And the movement has gradually expanded beyond AI, with some leaders advocating cryptocurrencies or nuclear fusion.

Soon, the movement was gaining steam in Silicon Valley, and officials in Washington were warning about its growing influence. It was a sure sign to the e/acc crowd that they had provoked the right people.

Recently, Forbes revealed that Based Beff Jezos was actually Verdon, who now runs an AI hardware startup called Extropic.

Verdon, who had enough media exposure for a week, declined to be interviewed for this column. His unmasking took some of the mystery out of e/acc, but it didn’t seem to dampen followers’ enthusiasm.

I recently interviewed several e/acc supporters, from early adopters to more recent converts. They all praised the move as a refreshing antidote to the pessimism of the AI ​​security group.

Amjad Masad, CEO of AI coding startup Replit (and investor in Verdon’s startup), told me he liked e/acc “as a meme counterbalance to all the AI ​​doom and gloom.”

Startup investor Julie Fredrickson said ae/acc was “a fun way to shortcut a future that prioritizes progress and solutions.”

These are, of course, opinions about the feelings caused by e/acc, not about its ideas, some of which are still too extreme for many people to accept.

Critics have pointed out the fact that some of the e/acc leaders, including Verdon, seem to actually agree with the Effective Altruists that an uncontrolled AI could wipe out humanity, but are unfazed by the idea, since a superhuman AI could represent a logical next step in evolution.

And some noticed that the movement became more partisan and serious as it grew. “I liked it when it was a tongue-in-cheek movement rather than what appears to be turning into a sincere libertarian movement,” said Aidan Gomez, CEO of AI company Cohere.