At OpenAI, two people now have the final say. One is the well-known Sam Altman, who is responsible for looking up at the stars and pursuing the poetry and distance of AGI. The other one is the newly airborne “application CEO” Fidji Simo. Her mission is much more down-to-earth: making money.

This "business queen" who helped Zuckerberg build the Facebook advertising empire and led Instacart to successfully go public is taking over everything about OpenAI except scientific research and computing power, especially how to make up for the huge losses of billions of dollars every year, and how to elegantly "embed" ads in ChatGPT that we use every day.


1. The ubiquitous “invisible” leader

Since taking the job in August, Simo has made limited appearances at OpenAI’s San Francisco offices. Her telecommuting is not because of "coldness", but because of illness.

Seamus suffers from postural orthostatic tachycardia syndrome (POTS), a condition that prevents her from standing for long periods of time and can even lead to fainting in severe cases. She developed the chronic condition after a difficult pregnancy, with symptoms worsened by a surgery in 2019. In her words: "Sometimes good and sometimes bad."

But this did not weaken Seamus' fighting ability. Instead, she transformed this physical limitation into ultimate efficiency.

"I am online from 8 a.m. to midnight every day, and I will respond within 5 minutes."

Seamus said in an interview. OpenAI is known for its Slack culture, which is so intense that even new employees can’t resist it, but for her, Slack is her battlefield. Employees also said that Seamus seemed to be everywhere, appearing on important channels at any time and asking tough questions at any time.

"Sam needs to focus on scientific research and computing power. My role is to free up his time to turn OpenAI from a R&D laboratory into a successful product company." She defined her role this way.

2. Commercial breakthrough: How to fill "trillion-level" black holes?


OpenAI currently faces a huge financial paradox: It is one of the most highly valued startups in the world, but it also burns through billions of dollars every year.

Seamus knew very well that the monthly subscription fee of $20 alone would not be able to support the expensive expenses of leading to AGI. She must dig deeper business value.

In terms of commercializing AI products, Simo faces fierce competition from technology giants such as Google and Meta, as well as startups founded by former OpenAI employees. “What keeps me up at night is that our models are much smarter than people can use them,” she admits. “I see my job as closing that gap.”

Simo believes, “We don’t need to compete for more areas, we just need to deepen and refine the current areas.”

The new product ideas brought by Simo are very clear: personalization and irreplaceability. She led the launch of Pulse, which is currently being tested among Pro subscribers. The product connects users' calendars, chat history and provides personalized briefings.

"In the past, only the rich had personal shopping assistants, travel or financial advisors. ChatGPT allows everyone to have such a team." Simo believes that when AI truly changes from a "chat tool" to a "life manager", users will naturally be willing to pay a higher premium.

In order to support this vision, OpenAI must make larger investments in computing power. Facing outside doubts about his trillion-level computing power transaction, Simo’s attitude is very tough:

“I know these transactions look risky to the outside world, but from the inside, if you don’t bet on computing power at this moment, that’s the biggest risk.”

3. Touching the "forbidden zone": What will ChatGPT's advertisements look like?

Before joining OpenAI, Simo worked at Meta for ten years and was well versed in advertising monetization. Everyone is concerned: Under her leadership, will ChatGPT become the next Facebook full of ads?

Seamus didn't shy away from the question, but her answer was strategic. She denied inserting advertisements into the conversation and instead proposed the concept of "Commerce Intent."

“When users come to us for shopping advice, there’s a huge commercial intent here,” explains Simo. Her logic is: What users hate is not the ads themselves, but the bad experience and misuse of data.

For example, she said, if a user asks "Recommend a tent suitable for camping", if AI can give accurate purchase links and comparisons, this is no longer an interruption, but a service.

But Simo also knows that privacy is the "Achilles heel" of OpenAI. "We haven't announced any advertising plans yet because if we're going to do it, we have to create a very different model than we have in the past."

“What I’ve learned from building ad platforms is that it’s often not the ads themselves that people dislike, but the use of data behind them.”

Apparently, Simo is waiting for the perfect moment to get ads into ChatGPT in a “painless” way.

4. Walking between business and ethics


In addition to making money, Simo must also deal with OpenAI's increasingly complex product ethics issues.

As a patient with a chronic illness, Seamus has an unusually high sensitivity to health and mental health. When she looked at the big picture, she immediately realized that “mental health is something we have to address, and employment is clearly going to see some disruption, and we have a responsibility to help minimize that disruption.”

· Mental health: One of her initiatives after taking office is to improve ChatGPT’s response to mental health crises. For example, when a user is extremely excited and hasn't slept for two days, AI needs to be able to identify the signs of "mania" and provide appropriate guidance.

· Job disruption: She also launched OpenAI certification and a job marketplace, hoping to certify 10 million workers as AI-ready and connect them to new job opportunities.

In the months since Seamus took over, OpenAI has reportedly significantly reduced the incidence of negative mental health reactions and rolled out parental controls.

5. Conclusion: OpenAI’s “Second Brain”

The joining of Simo marks OpenAI's complete bid farewell to the era of pure idealism.

If Ultraman in the San Francisco office represents looking up at the stars, then Simo, who works remotely from his home in Los Angeles and cannot even stand for a long time, represents being down to earth.

She is using an extremely calm business logic to tame one of the most powerful AI models. For users, this means that ChatGPT will become more user-friendly and user-friendly, but it also means that the era of pure dialogue without commercial considerations may be entering a countdown.

If ChatGPT can accurately recommend the products you need, would you mind seeing ads in the conversation?