Wildebeest crossing the Mara river in Serengeti national park, Tanzania. Great migration - stock photo
Photo: OlyaSolodenko/iStock /Getty Images Plus

Generative AI finds its FOMO moment

Protocol Pipeline

Hello, and welcome to Pipeline. I’m Biz Carson, and after spending 17 days on the road in the Midwest and New England, I’d argue that the Midwest’s fall is underrated and has some pretty great leaf-peeping too. Thanks to Tomio for filling in last week in my absence. Also happy birthday, Dad!

This week in the startup world: the weird place people listen to Clubhouse, why managing out is the new layoffs, and generative AI’s breakout moment.

The FOMO around generative AI

Generative AI can create more than just text and images: It’s clearly generated a hype cycle around AI companies and rabid investor interest in the space.

This week was generative AI’s breakout moment, at least if you’re judging by the parties, blog posts, funding deals, and new use cases that all debuted this week.

  • It kicked off Monday with Stability AI’s funding round and kick-off party at San Francisco’s Exploratorium. Sergey Brin, Naval Ravikant, and Ron Conway all attended what The New York Times described as “a coming-out party for generative A.I.” The makers of Stable Diffusion, one of the most popular image-generation models, had raised $101 million and become a unicorn.
  • The next day, AI copywriter Jasper announced its own $125 million round and new billion-dollar valuation as well. Sequoia partner Sonya Huang also released a market map of companies in the space, which included Jasper. The fact that two newly minted unicorns were on it (plus investor herd mentality) made it go viral .
  • On Thursday, fashion maker Cala announced that it was using DALL-E to generate new ideas for clothes and accessories. I’m guessing “powered by DALL-E” is going to be a tagline you’ll see on company's pitches even more often.
  • By Friday, the thought leadership around generative AI was in full force with people like Elad Gil , NFX’s James Currier , and several partners at Coatue all publishing their views on the space.

The hype may be high, but it’s “absolutely justified,” said Sequoia’s Huang. While models like GPT-3 and DALL-E have been on the market for a short time, Huang’s map was one of the first to really lay out what the applications of the technology could be and where people are already building on it.

  • Image generation, text, and code are the three main areas where there’s a lot of concentrated activity, as the funding rounds of Jasper and Stability AI prove. “People are moving on beyond one-shot image generation to a few different branches, whether it's ‘Let's make this really good for the process of interior design or product design’ or ‘Let's make it a really, really good image generator that you can continue innovating with the machine on,’” Huang said.
  • But there were empty boxes in areas like biology that were still “to come” on Huang’s market map, meaning green fields abound for entrepreneurs. “If I was a founder in [Y Combinator] right now, I would 100% be pointing my guns at one of these models and seeing what I can do,” she said.

The end goal isn’t more whimsical images, but giving humans “superpowers” by having a machine work alongside them.

  • When Huang started thinking about which areas could find applications, she turned to GPT-3 to help her brainstorm. She told me that when writing Sequoia’s original blog post on the subject, she started by using GPT-3 to help describe the differences between classical AI and generative AI, and from there, asked GPT-3 to fill in the blank on potential applications, generating even more ideas. “That human-machine iteration loop I hadn't experienced before,” she told me.
  • Learning how to work with these models is going to be a new skill set, one that could make people more creative or more productive. “We have to train how we work with the machines, but I think the result really is we are superpower humans as a result of being able to work with these machines,” Huang told me.

It will be a while before generative AI can generate its own market map, but given its ability to manufacture buzz, the future maybe isn’t too far away. Huang’s already working on an update to a landscape that’s changing fast. Let the FOMO begin.

For more on generative AI, read my Q&A with Sequoia’s Sonya Huang.

Overheard

Can you tell what’s AI or not? This week, YouTuber Ali Abdaal’s Twitter thread on productivity hacks went viral with over 1 million impressions. A day later, Abdaal revealed it was an experiment that was almost entirely written by GPT-3. The AI’s answers were good (and actually useful), which could threaten the creator economy that normally churns out these kinds of threads. “I was a bit scared initially, but this experiment makes me even more bullish on the idea of personality and personal brands, where an AI can’t (yet) replicate the vibe we feel with other humans (especially through video),” Abdaal tweeted.

“This is a giant office park. It is indistinguishable from Huntsville. How are you people not embarrassed? You’re supposed to be builders.” An hour visiting the heart of the tech industry left one person deeply disappointed by the low-slung, low-key office buildings of Silicon Valley. Just don’t tell them how it used to be beautiful orchards before it was turned into office buildings.

Quiet quitting is out. “Quiet layoffs” are in. That’s the new moniker for the term when companies just try to quietly manage out a bunch of people by giving bad performance reviews. For example, Stripe managers were reportedly told to start giving lower ratings to chunks of the workforce and were asked during performance evaluations whether they’d hire that employee again.

This week’s weirdest stat: 24% of Clubhouse users use the app in the shower.

A MESSAGE FROM GOALS HOUSE

It's becoming increasingly appreciated among the broader business and NGO community that the planet and people elements of sustainability are mutually dependent, and as such a focus on one at the exclusion of the other will be fruitless. But balancing profit and sustainability progress remains a more thorny debate.

Learn more

Inside track

How low can we go? The public market continues to compress, writes Redpoint’s Tomasz Tunguz . Revenue multiples are now below 2016 levels.

It feels early for 2023 predictions, but it’s time to break out the crystal ball. Contrary’s Eric Tarczynski predicts a bifurcation in the market where the best companies will find as good of terms as any and everyone else will remain in the depressed market of this past year.

Consumer social is saturated already, but innovation is still happening, and apps like Gas are still having their breakout moments. Index’s Rex Woodbury breaks down why consumer social isn’t totally dead yet and where there are still areas (like photo sharing) for startups to tackle.

Need to know

Venture deals declined in Q3 quite dramatically. “The magnitude of the decline from 3Q21 -> 3Q22 is the largest such decline ever observed in a 12-month period on AngelList,” the company wrote about the sharp pullback seen last quarter.

SF is getting its own tech week. This week already kind of felt like tech week between TechCrunch Disrupt and the Masters of Scale event in the Presidio, but not to be upstaged by Miami and LA, VCs are banding together to create an SF Tech Week in early November. Seems like a bit of an oxymoron if you ask me.

Instacart’s IPO is back on ice. The real surprise is that anyone ever thought an IPO this year was possible.

Ye acquired Parler, but perhaps the most important part is what he isn’t buying .

It’s time to build (more crypto companies). Andreessen Horowitz is going toe-to-toe with Y Combinator with the launch of its new accelerator, Crypto Startup School , as more firms build their own YC alternatives. Meanwhile, a16z partners have been heavily involved in political donations to crypto-friendly candidates, and a Forbes deep dive into their spending argues its paying off for the firm.

Moves: In more niche climate tech VC news, HubSpot co-founder Brian Halligan has a new $100 million fund focused on ocean-related climate tech and Chris Sacca has a new $250 million fusion-focused fund . Marc Lore and A lex Rodriguez (yes, A-Rod) are teaming up on a new sports ticketing startup that would resell seats when people leave midway through a game. Kanjun Qiu is starting an AI research company, Generally Intelligent , to create actually intelligent AI agents.

From Protocol: Peter Thiel's gamble against the “somewhat fake California thing,” and how the fate of the Republican party (and his role in it) hinges on the Senate races in Ohio and Arizona.

Also from Protocol: How Big Tech critic Meredith Whittaker decided to leave the FTC to run Signal.

Also, also from Protocol: Clubhouse and its clones continue to try, but audio will never capture the hearts of social media users.

Your weekend reading: He’s already a big name in tech, but now Craft Ventures’ David Sacks is making a name for himself in politics too. The New Republic took a look at the “quiet political rise of David Sacks, Silicon Valley’s prophet of urban doom” and how Sacks is following in the footsteps of buddies Elon Musk and Peter Thiel.

A MESSAGE FROM GOALS HOUSE

Currently, much of the ‘E’ in ESG is focussed on climate only, and it is essential that companies also focus on biodiversity, recognizing nature-climate linkages in order to optimize mitigation and build resilience. ESG will prepare us for the necessary paradigm shift, driven by increasing external pressures forced upon us as a result of short-term profits.

Learn more

Thanks for reading! If you like what you’re reading, sign up here to get it in your inbox. Send story tips and newsletter feedback to bcarson@protocol.com.

Recent Issues