Home services

San Francisco’s ‘Cerebral Valley’ Booms in ChatGPT and Generative AI

  • Because of pandemic-era closings, San Francisco became somewhat of a ghost town for two years.
  • As people left and worked remotely, some wondered whether San Francisco was dead as a tech hub.
  • Now a race to succeed in the fledgling space of generative AI has founders flocking back.

Loading Something is loading.

Thanks for signing up!

Access your favorite topics in a personalized feed while you’re on the go. download the app

San Francisco’s tech scene is back.

After the pandemic effectively shut down the city for more than two years, sentiment is shifting away from proclamations of a once great city’s demise and toward the good old times, when it was the destination for people trying to reshape the world’s technological vision.

Across the city, founders are planting their flags, with dreams of riding the wave of a new technology that’s been said to be a step change akin to the iPhone: generative artificial intelligence. Amber Yang, an early-stage investor at Bloomberg Beta, recently tweeted that startups in that field were flocking to San Francisco’s Hayes Valley neighborhood, which founders have renamed “Cerebral Valley.” The tweet was made somewhat in jest, but Yang added that the nascent field of generative AI was advancing so quickly that teams felt being together in one hub was necessary to keep up.

Generative AI takes training data — for instance, a vast collection of written text — and teaches itself how to produce unique works. In the first five days of its release, more than 1 million people tried out ChatGPT, an AI chatbot that can respond to questions with humanlike answers, according to Sam Altman, the CEO of OpenAI, the company behind ChatGPT. Microsoft is reportedly investing $10 billion in OpenAI, with plans to incorporate ChatGPT into its Bing search engine and Azure cloud offerings.

To be sure, ChatGPT has limitations. It may know how to form humanlike sentences, but it can’t discern whether they’re accurate.

Still, the underlying technology of generative AI is impressive, and startup founders see much potential. Twenty-two percent of generative-AI companies are based in the greater San Francisco Bay Area, and 55% of capital invested in the space is landing there, James Currier, a partner at the early-stage investment firm NFX, said.

The ‘crazy hackers’ are here

Víctor Perez and Diego Rodriguez knew San Francisco was the place to be for generative AI when they moved there several months ago to build KREA. Their startup, in the city’s Hayes Valley neighborhood, creates models for high-quality image generation and asset management services.

Originally from Spain, the duo first landed in Miami last year, where they developed generative image models. There, they said they noticed that most of the “crazy hackers” they met came from San Francisco.

After they gave New York a try for several weeks, the generative-AI boom picked up. People began telling them to head West. Dave Fontenot, who founded a 12-week founder residency program in San Francisco called HF0, told them they’d be “crazy” and “irresponsible” not to work on generative AI in San Francisco, Perez said.

Picture of a bare apartment with a desktop and scattered items.

The home where KREA’s founders work.

Thomas Maxwell/Insider

Perez and Rodriguez initially planned to stay briefly, but when they started meeting people around San Francisco working on generative tech — including people working on artificial intelligence at Meta and OpenAI — they knew they had to stay. They said they felt the excitement and the motivation of developers to build something new.

Picture of a bare apartment in San Francisco.

Pre-pandemic, these types of bare homes for startup founders were common.

Thomas Maxwell/Insider

Perez said that the sense of urgency to get working on building better AI models came from how generative AI improved with more data. Models must be trained using real data created by humans — the more images of a fish that an AI model sees, the better it gets at producing its own image of a fish, for instance.

“We feel urgent,” Perez said. “But it’s not because some other people can create a better model than us today. It’s creating the best models tomorrow.”

Picture of a desktop computer displaying KREA's image generation software.

KREA is developing high-quality image models, like one that can generate images in the style of a Studio Ghibli film.

Thomas Maxwell/Insider

Another founder who recently landed in San Francisco, Nicholas Locascio, is working on Booth AI, which targets e-commerce with a tool that generates professional product shots without the user having to pay for an expensive photo shoot. Customers of the tech upload an image of a product — say, a coffee mug — and then Booth AI can place it in a lifestyle scene that makes it look appealing on e-commerce pages.

Along with his cofounders, Ian Baldwin and Mitra Morgan, Locascio was recently accepted into the vaunted Y Combinator accelerator program in Silicon Valley.

“This is a technology that a lot of traditional thinking about programming just doesn’t work for,” Locascio said. “Nobody knows the best way to do anything right now. It’s a complete gold rush.”

Picture of Booth.ai co-founders Nicholas Locascio and Ian Baldwin.

Ian Baldwin and Nicholas Locascio, two of Booth AI’s cofounders.

Nicholas Locascio

‘No skeptics yet’

While founders are confident that generative AI will change the world, they’re still trying to figure out exactly how it will play out.

“There are no skeptics yet, so it’s a unique time,” Currier, the NFX partner, said. “But entrepreneurs have to figure it out because the Big Tech companies aren’t sitting around.”

To Currier’s point, Google issued a “code red” in recent weeks to respond to the threat of generative-AI products to the company’s search and other key services, The New York Times reported in December.

That’s why it’s key for up-and-coming AI startup developers to work together to get ahead while they can and share ideas with each other — similar to how in the early days of the sharing economy, the Uber cofounder Travis Kalanick and Airbnb CEO Brian Chesky would have dinner together frequently and leave with ideas on how to improve their companies, according to the Amazon biography “The Everything Store.”

“The lunch and the parties happening in the Bay Area are going to have a substantial impact” on who figures out the strategies to win, Currier said, adding: “So being in the same place matters.”

Related Articles

Leave a Reply

Your email address will not be published. Required fields are marked *

Back to top button