There’s a sickening moment in the life of a high-flying startup when you can feel gravity reasserting its inevitable grip. One moment you’re accelerating into the future, the next you’re suspended, at the apex of an unexpected plunge back to earth. For years you’ve kept your rocketship thrusting up and to the right: higher revenues, bigger financing rounds, more new hires, loftier valuations. But then comes a reckoning – a cloudy sales projection, a round of unusually querulous investor calls, a restive board meeting. Something’s changed, and fixing it feels out of your control. Your rocketship has become a falling knife.
I’ve experienced that dreadful moment at least three times in my career. The first was at Wired, when our high-flying magazine attempted to go public and failed to produce an order book that let us ring the bell. The second was at The Industry Standard, when advertising sales fell off a cliff just as the dot com boom went bust. And the third was at Federated Media, as Facebook consolidated its grip on the burgeoning social web, and our partners began to move their budgets from blogs to Zuckerberg’s magical Newsfeed.
Right before, each of my startups had hit milestones that seemed to guarantee never ending optimism. Wired won three National Magazine awards, launched the first commercial website, and became a fixture of stories about the rise of the “digital revolution.” The Industry Standard hit nine figures in revenue and turned profitable, and I became a finalist for E&Y’s Entrepreneur of the Year. Federated closed a $50 million Series C, eschewing a billion-dollar acquisition offer along the way. Each time, my highest high came right before gravity kicked in. I learned to be wary of success. It dazzles, but it also blinds.
Running a startup is a confidence game. As a founder or CEO, your job is to see around corners, invent the future, and tell a story about how your impossible ambitions are not only reasonable, but inevitable. The people who gather around you are all willing conspirators in that game – your co-founders, your early investors, your partners and your public all want to believe your story, because your success is their success. On the way up, it’s a cheerleading camp. On the way down, it becomes a horror film.
The pain is far worse if your startup is the poster child for a larger movement. When things are great and all the press is laudatory, it’s easy to ignore the legions of folks who quietly resent your success. But when things go south, everyone looks for a culprit. If you’re in charge on the way up, you’re even more responsible on the way down. And schadenfreude’s a bitch.
Once the world’s highest-flying startup, OpenAI now finds itself in a nauseating moment of narrative weightlessness. Valued at $14 billion a year before ChatGPT’s launch in 2022, OpenAI added a staggering two orders of magnitude to its market cap in less than five years, peaking at an estimated $1.5 trillion when it filed confidentially for an IPO in June of this year. That filing likely marked its apogee – the company delayed its IPO early this Fall and attempted to raise $30 billion at a $1.4 trillion valuation. It failed to hit either of its targets, raising just $11 billion on a valuation of $1.25 trillion last week. Ouch.
When a company’s narrative shifts this quickly, the first casualties are the employees. What was once a culture of confidence, even arrogance, quickly becomes one of fear, uncertainty, and doubt. Internal message boards turn toxic, and even your most loyal staff begin to map their exits. As CEO, your job is to remain unruffled, to project faith and poise. But your closest advisors can see the panic you’re fighting to contain. Board meetings get ugly. Investors who for years loudly bragged about their stake in your company start to talk up other parts of their portfolio. And the press! The press turns from a fawning glee club to a tragic Greek chorus.
This is the moment you are most tested as a leader, when your words and deeds are scrutinized in ways you’ve never imagined. Every conversation with an employee instantly echoes throughout the organization. Every public utterance is examined for proof of dread or a glimmer of hope. Your every move is magnified. This is where, to put it bluntly, you can truly f*ck it all up.
As OpenAI began to look like the world’s largest falling knife, its leadership fought to regain control of the narrative. The results have been mixed.
As the company was delaying its IPO, OpenAI announced Dots, its entry into the red hot personal agent marketplace. But the launch came weeks after Meta debuted Muse, which earned heaping praise from pundits and press alike, then raced to the top of the app store charts. Around the same time, OpenAI published a commendable transparency report detailing how it contained half a dozen new instances of rogue behavior from its frontier models. But any positive PR was muted two weeks later, when reports surfaced that the company’s reputed revenues of $70 billion were overstated by more than $20 billion. That story accelerated market-wide skepticism around “the AI trade,” with OpenAI its unfortunate central character.
Along the way, OpenAI CEO Sam Altman did himself and his company no favors by appearing on the Cheeky Pint podcast and demonizing localized, distributed approaches to AI. His argument – that scarce resources like memory chips should be allocated to centralized services like OpenAI, and away from regular folks’ personal computers – went over like a lead zeppelin. This was not the first example of Altman’s tone deaf pronouncements. Earlier in the year he brushed off concerns about energy use at data centers by equating their resource consumption to humans, which, he claimed, “take a lot of energy to train.” His recent attempts to rebrand politically toxic data centers as “AI compute campuses” is just as likely to fall flat. No one is in the mood to be charitable with OpenAI’s brand at the moment. It’s time for a new playbook.
You can see the outlines of just that with a strategic leak from OpenAI’s consumer lead Adam Fry. Speaking to Big Technology’s Alex Kantrowitz earlier this week, Fry announced that ChatGPT has added 200 million users in the past two months. Perhaps seeking to draft off tech’s largest and most profitable consumer service, Fry also claimed that ChatGPT now serves the equivalent of 20 percent of Google’s search volume. That’s undeniably impressive.
But perhaps the best news for OpenAI came three days ago with the announcement that Jill Hazelbaker, president of Uber and formerly of Snap, Google, and various political campaigns, will join the company in the newly created role of Chief Corporate Affairs Officer. Hazelbaker will oversee “the artificial intelligence company’s global communications and corporate positioning efforts,” according to Bloomberg, including “company communications, reputation, and work around the corporate brand, social media and other partnerships.” In other words, it’s now Hazelbaker’s job to change OpenAI’s faltering narrative.
It’s a career-defining role for the veteran Hazelbaker, and I’d imagine she wouldn’t have taken it if she wasn’t guaranteed the authority to make sweeping changes not only to how OpenAI communicates its story, but what the story will be in the first place. As OpenAI’s rival Anthropic begins to wobble under the weight of Wall Street’s expectations, Hazelbaker has a once-in-a-lifetime opportunity to reset how the world understands her new company. Stepping in to catch what may be history’s biggest falling knife is both brave and possibly foolhardy. But you have to admire the guts it takes to try.
—