Databricks chief Ali Ghodsi has become unstoppable after closing the biggest funding ever. He recently made a bold statement: people matter, but company culture matters more—those who do not fit should be shown the door.
“You shouldn’t have people in your company that don’t satisfy [the culture]. Maybe you made a mistake. Maybe you hired the wrong person. Maybe you promoted the wrong person. But now that you know that person is not the culture [of your company], you move them out,” said Ghodsi in a recent interview with Dan Primack at the Axios AI+ SF event.
Ghodsi, on several occasions, has been candid with his say. During the interview, he outlined his process of hiring and advocated seeking the truth right when interviewing employees for the job. He does that by asking questions like: “Tell me when you screwed up” or “Tell me something you really botched up.”
He also ensures that people who satisfy the company culture and the principles are the ones promoted. “That one is hard because you might have steller employees that have had a huge impact, but they’re actually not really the culture of the company,” he said.
However, Ghodsi said he prioritised a ‘cultural match’ over most things, even while doing what the company loves doing, which is acquiring other startups.
“Corporate development teams look at the EBITDA, revenue growth, finances, excel math, and they convince them [the buyers] that we’re a good idea—but we kind of do that last. We start upfront with the people and ask, ‘Are we going to have a cultural match?’” Ghodsi added.
Ghodsi was hitting it out straight. He revealed that it was dumb to go for an IPO in 2024 and that the media was responsible for nearly doubling their funding capital.
“The culture of your company is mostly the personality of the CEO,” said Ghodsi in a talk at Stanford earlier. “What I’m saying is that Bill Gates’ Microsoft is a very different culture from Steve Balmer’s Microsoft, which is a very different culture from Nadella’s Microsoft.”
“If you can get along with Satya now, and you operate the way he likes to operate, you’re going to be successful with him. And if you operate probably the way you would do with Steve Balmer, it’s not going to work,” he added. It is the reality that most companies at the forefront of AI are experiencing today.
He emphasised that technical prowess and sound capital aren’t enough to thrive if you’re building something as disruptive as AI. “There is no one way to go about it. You may want to halt progress or propel it. You may highly prioritise safety or lean towards a balanced approach.”
When companies like Google, Microsoft, and OpenAI take charge of the AI revolution, they will have to make two choices: one, on their goal for AI, and two, to understand who is aligning with that vision, said Ghodsi.
Pick Your Players and Teams Wisely
Over the past few months, we’ve seen companies and individuals prioritise their values and principles and break ties due to conflicts over them.
Take Ilya Sutskever, for instance. One of the major forces behind OpenAI, he ultimately decided to leave due to conflicting ethical considerations within the company.
Before that, OpenAI’s board, which Sutskever was a member of, briefly parted ways with the CEO, Sam Altman, because the board deemed his approach to running the company not transparent enough.
Daniel Kokotajo, another employee from OpenAI, resigned ‘after losing confidence that the company would behave responsibly in its attempt to build AGI’. He also said that he did not sign confidentiality agreements on his departure and forfeited his equity in the company.
Did Altman’s ‘culture’ at OpenAI change over time, or did he make the mistake Ghodsi was referring to? In a round table discussion with his employees, Dario Amodei, CEO at Anthropic, said that industries have only gotten outside of a risk aversion mentality in recent years.
More notably, Amodei quit OpenAI in 2021, along with Jack Clark and others, over a disagreement on the company’s commercial focus.
“If you have the vision of how to do it, you should go, and you should do that. It is incredibly unproductive to try and argue with someone else’s vision,” said Amodei in a podcast with Lex Fridman.
When Amodei and Co. proceeded to build a company with people aligned with their culture, they achieved monumental successes. Anthropic has repeatedly offered performance parity with its models compared to those of Google and OpenAI.
The company isn’t distracted by the competition and doesn’t focus on multimodal AI, text-to-video, or reasoning models. Instead, it stands out with features like computer use, model context protocol, and so on.
On the other hand, OpenAI’s mass exodus involved more names such as Mira Murati, John Schulman, and Logan Kilpatrick.
“It feels great going back to the AI roots,” said Kilpatrick, in an exclusive interaction with AIM, when sharing his experience of working at Google.
But OpenAI hasn’t faced a setback, at least not in terms of building what it set out to. In fact, its latest O3 model is superior to every other offering from the competition.
People who have left OpenAI in the last 12 months:
-Mira Murati: CTO
– Greg Brockman: President & Co-founder
– Ilya Sutskever: Co-founder & Chief Scientist
– John Schulman: Co-founder
– Bob McGrew: Chief Research Officer
– Barret Zoph: Vice President of Research
– Jan Leike:… https://t.co/bBFH7UKD8S— Jack Prescott (@JackPrescottX) December 19, 2024
Microsoft is another big name in this area. Last year, the company laid off a team that taught employees how to build AI tools responsibly. The team called the ‘ethics and society team’ was responsible for mitigating the risks involved with AI.
It also worked on identifying risks that were attached to Microsoft’s adoption of products built by OpenAI. In a similar incident in 2020, Google fired Timnit Gebru, an AI researcher, when she published a paper critical of AI and its risks.
On the flip side, there have been multiple scenarios where companies have hired leaders with a great cultural fit and are making the most of it. In March 2024, Microsoft announced Mustafa Suleyman, a former Google DeepMind researcher, the CEO of Microsoft AI.
Suleyamn heads all consumer AI products, and the company has achieved quite a success with its Copilot + AI PCs and features like Copilot Vision.
His appointment can be attributed to the vision and culture Satya Nadella set when he was appointed as the CEO in 2014. “I believe that over the next decade, computing will become even more ubiquitous, and intelligence will become ambient,” he said.
The post You Shouldn’t Have People Who Don’t Fit the Company Culture—‘Move Them Out’ appeared first on Analytics India Magazine.
