| Over 300 organizations rely on The Information for exclusive insights into public and private companies, including in-depth analysis of the ways in which tech moves markets. Click here to contact our corporate and enterprise team to learn more.
Welcome back! It's the startup equivalent of misplacing a winning lottery ticket. Former employees of Scale AI are being excluded from a planned share sale at a $25 billion valuation. "I just heard back from the stock admin. Former employees won't be participating in the tender offer. rip," one ex-employee wrote last month in a private Slack channel of former Scale workers. "So the only way is to re-join?" another person asked. I got outreach from other former Scale AI employees annoyed about this. The company has been around for nine years, has seen its valuation more than triple in the last two years, and doesn't appear all that close to an initial public offering. (In fact, it missed its revenue and profit forecasts last year, I reported last week, even as it grew rapidly.) "Even Palantir went public after 17 years, so just another 8 years to go," one ex-employee wrote in the Slack channel. The worries reflect some of the personal anxieties across Silicon Valley about how long this artificial intelligence boom will last—and who is getting rich off it. The potential for riches has been particularly prominent at Scale AI, whose CEO, Alexandr Wang, has been touted as the world's youngest self-made billionaire. (I profiled his rise last year.) It's also worth noting there's another group of people critical to Scale AI's success—the contractors who fine-tune and label AI models—that just get an hourly wage rather than equity. There are no rules, written or unwritten, that say whether companies need to allow former employees to sell stock in private company tender offers. OpenAI last year initially allowed only current employees to sell in a share sale. But it backtracked following negative media reports. Overall, though, more private companies appear to be allowing former employees to sell as the market has improved over the last few years. Databricks, for example, recently held sizable share sales, at eye-popping prices, for both current and former employees. Of the roughly 80 tender offers run through the Nasdaq Private Market last year, 52% allowed ex-employees to sell some percentage of their shares, up from 42% in 2022, according to NPM. There's a lot of emotion wrapped up in this debate. CEOs might think: Why reward people who left? Ex-employees think: I worked at a below-market cash salary because I expected stock rewards to come within a reasonable time frame. Companies add to the stress by staying private for longer than expected. But it's also perhaps a corporate finance decision. Larger tender offers cost more to run. Most importantly, companies don't want to flood the illiquid market with too many shares, possibly threatening their lofty valuations. How many potential buyers are there of Scale AI stock at a $25 billion valuation? Well, maybe only enough to satisfy some existing shareholders. Keith Block benefited from the rise of cloud software over the last two decades as much as anyone. He spent seven years as a top executive at Salesforce, including 18 months as co-CEO, as the company's stock price rose fivefold. Now he's looking at the landscape as a venture capital investor. His Smith Point Capital Management announced Thursday it had finished raising $360 million. He is skeptical of the current crop of pricey AI startups. "I think it's much harder to find a gen AI company that will really be a platform that is mission critical to the enterprise," he told me. "The valuations are not that interesting to us." That kind of analysis is running through the minds of investors as more hot AI startups like Glean, which builds chatbots for companies, raise at increasingly high valuations. We broke the news yesterday that asset manager Wellington is leading a financing of Glean at a $7 billion valuation. That would be over 70 times the annual recurring revenue Glean told investors it had in January. To several investors I spoke to, that seems frothy. Even the most innovative startups face the risk that an enterprise software giant like Microsoft introduces something similar and can take advantage of its pricing power, Block notes. Customers are "just going to say, 'You know what? It's easier for us to stay with Microsoft.'" (As if on cue, Microsoft shares soared on Thursday following a strong earnings beat.) Block instead talked up edge computing, another emerging technology that draws a lot fewer headlines (and fewer incumbent tech giants as competitors). Edge computing refers to bringing compute capacity physically close to where data is generated—an important advancement for companies in fields like robotics or oil drilling that are running operations away from a central hub. "We think that potentially can be just as powerful as the AI story," he said. Now over to Sri… Some new AI startups have recently raised supersized first rounds. But many early-stage firms are sticking to small checks. One is South Park Commons, founded in 2016 by early Facebook engineer Ruchi Sanghvi. The VC firm raised $275 million to back more pre-seed and seed startups, we're first to report. While that's double the size of its 2021 fund, it will keep checks small, in the $1 million to $10 million range. "I don't think you need billions of dollars to build a generational company," said Aditya Agarwal, a managing partner at South Park Commons and former CTO of Dropbox. South Park Commons operates like a startup accelerator, with a focus on highly technical founders at the "ideation" stage. It expects more than 20,000 applicants to vie for roughly 200 spots this year to join as members. They can work out of the firm's locations in San Francisco, New York and Bangalore, India and attend workshops and speaker events, such as with Meta Platforms CEO Mark Zuckerberg and OpenAI CEO Sam Altman. Founders can also apply to a fellowship program that runs twice a year. Of those selected, the firm invests $400,000 in exchange for 7% of equity and an additional $600,000 at the terms of the next round. The VC firm made over 60 new and 40 follow-on investments last year. Small has worked so far. Some of the firm's early investments include AI inference startup Baseten, valued at $825 million valuation, according to CNBC, and Replit, a coding copilot in talks to raise money at a $3 billion, according to Bloomberg. The firm's first, $55 million fund is in the top 5%, measured by multiple on invested capital and total value to paid-in capital, of funds started in 2018, the firm said. | 
0 comentários:
Postar um comentário