OpenAI – Techin' Geek

How Shivon Zilis Functioned as Elon Musk’s OpenAI Insider

May 1, 2026sparta

As the first week of the Musk v. Altman trial wraps up, one individual has emerged as vital in managing communications and personalities during OpenAI’s early days: Shivon Zilis. A long-time associate of Musk and mother of four of his kids, Zilis became an adviser to OpenAI in 2016, later sitting on its nonprofit board from 2020 to 2023, and has also held executive roles at Musk’s other companies, Neuralink and Tesla.

In court, Musk referred to Zilis as a “chief of staff,” “close adviser,” and noted that “we live together, and she’s the mother of four of my children,” although Zilis mentioned in a deposition that Musk is more like a regular guest. Last September, she told OpenAI’s attorneys that she and Musk started their romantic relationship around 2016 after she began advising OpenAI. They welcomed their first two children in 2021.

OpenAI’s lawyers claim that Zilis has acted as a clandestine link between OpenAI and Musk, even after he departed the board in February 2018. In a text message to Musk on February 16, 2018, Zilis inquired whether she should “stay close and friendly to OpenAI to keep info flowing,” to which Musk responded, “Close and friendly, but we are going to actively try to move three or four people from OpenAI to Tesla.”

On the witness stand, Musk expressed his desire to stay informed of developments. During the same conversation, Musk voiced skepticism about OpenAI’s future if he concentrated on Tesla AI. Zilis concurred, citing concerns regarding Demis Hassabis of Google DeepMind. She emphasized the necessity of exerting influence over Hassabis due to the potential dangers of AI advancement.

Two months later, in an April 23, 2018 email, Zilis updated Musk on OpenAI’s fundraising and a new AI initiative. She indicated a shift of her focus more towards Neuralink and Tesla but was willing to reallocate her attention based on Musk’s desires.

Previously, in the summer of 2017, Musk attempted to assert control during discussions about OpenAI’s corporate framework. In an August 28, 2017 email, Zilis communicated about her meeting with OpenAI’s leadership concerning equity distribution. Musk replied, “This is very annoying. Please encourage them to go start a company. I’ve had enough.”

Certain Jurors in the Musk v. Altman Case Have an Aversion to Elon Musk

April 28, 2026sparta

A jury was selected on Monday as the trial of Musk v. Altman commenced in a federal court in Oakland, California. Some of the jurors indicated concerns regarding Musk and the AI technology at the heart of the case, yet assured their ability to put these aside for the duration of the trial. The trial’s commencement also triggered a series of events outside the courtroom.

Sam Altman and Greg Brockman from OpenAI were spotted in the courthouse security line, with Elon Musk notably absent. Journalists crowded into an overflow room to catch an audio feed of the proceedings.

The goal was to select nine unbiased and fair jurors, a daunting task given the prominence of the tech leaders involved. Although many jurors expressed unfavorable opinions about Musk, the majority were not disqualified, though one was excused due to strongly negative feelings toward Musk.

Judge Yvonne Gonzalez Rogers recognized that numerous individuals held negative perceptions of Musk but maintained that jurors with such views could still support the judicial process. The jury will determine whether Altman and others diverted OpenAI’s nonprofit mission from its original purpose, potentially violating the law. Their verdict will be advisory, with Gonzalez Rogers making the ultimate decision.

The selected jurors represent a varied group, including a painter, a former employee of Lockheed Martin, and a psychiatrist. While some held negative views on AI technology, they assured the court that these would not hinder their ability to ascertain the facts.

OpenAI attorney William Savitt expressed his satisfaction with the jury selection process. He conveyed that Altman, Brockman, and OpenAI are keen to present their case and are confident in their position, aiming to reveal the truth.

In the meantime, Musk is actively seeking public backing, utilizing his social media platform X to promote a New Yorker inquiry into Altman’s supposed business misconduct. This aligns with OpenAI’s newsroom account describing Musk’s lawsuit as an effort to derail their mission to ensure that AI benefits humanity. Demonstrators outside the court demanded a halt to AI development.

The trial proceeds on Tuesday with opening statements from attorneys and the first witness taking the stand.

OpenAI Secures Tech Talk Show ‘TBPN’—and Benefits from Favorable Coverage

April 2, 2026sparta

OpenAI announced on Thursday that it has acquired the online business talk show *TBPN* for an undisclosed sum. This move comes in response to the challenges OpenAI has recently encountered regarding its public perception.

Since its launch in 2024, *TBPN* has garnered attention in Silicon Valley by offering a daily livestream centered on the tech sector, viewed as more favorable towards technology than conventional media. Hosts John Coogan and Jordi Hays provide up-to-the-minute news analysis, review trending social media content, and interview executives from companies like Meta, Salesforce, Palantir, and OpenAI. It has gained particular popularity among OpenAI staff and other AI researchers, many of whom are active users of the social media platform X.

The acquisition prompts inquiries into how a media startup fits with OpenAI’s core functions of promoting ChatGPT, Codex, and a forthcoming super app. In March, OpenAI’s applications CEO, Fidji Simo, indicated that the organization should discontinue side projects and concentrate on its main activities.

In a staff memo regarding the acquisition, Simo claimed that ordinary communication tactics do not apply to OpenAI. “We’re not a typical company,” she mentioned in the memo, which was released as a blog post. “We are initiating a significant technological transformation. With the objective of delivering AGI to the world comes the obligation to cultivate an environment for genuine, constructive discussions about the advancements AI induces—centered around users and developers.”

*TBPN* is a relatively small operation in comparison to OpenAI. The media entity reported $5 million in advertising revenue last year and anticipates over $30 million by 2026, as per The Wall Street Journal. The show is said to draw around 70,000 viewers per episode across various platforms. A source familiar with OpenAI noted that the company does not expect *TBPN* to be a financial contributor, but it supports OpenAI’s communication approach.

OpenAI has been facing increasing public scrutiny recently. Following a February deal with the Department of Defense, Anthropic’s Claude experienced a spike in downloads, becoming the top free app on Apple. OpenAI’s leadership is also tackling the growing QuitGPT movement, comprised of individuals vowing never to use OpenAI’s offerings. OpenAI president Greg Brockman cited the challenges posed by AI’s popularity as a significant reason for his increased political spending.

This acquisition situates OpenAI among recent efforts in Silicon Valley to manage news enterprises. Over the past decades, tech giants have acquired media outlets, including Jeff Bezos’s purchase of The Washington Post, Marc Benioff’s acquisition of Time magazine, and Robinhood’s takeover of the newsletter company MarketSnacks. Such acquisitions often raise immediate concerns about the independence of the outlets. Simo reassured employees that *TBPN* would retain its editorial freedom.

“My favorite tech show is *TBPN*. We want them to keep excelling in what they do best,” OpenAI CEO Sam Altman posted on X. “I don’t expect them to be easier on us, [and I] will likely contribute to that with occasional whimsical choices.”

OpenAI affirmed that *TBPN* will maintain autonomy regarding its programming, guest selections, and editorial decisions, according to Simo’s memo. It also indicated that *TBPN* would report directly to OpenAI’s VP of global affairs, Chris Lehane. WIRED previously uncovered challenges faced by an economic research team under Lehane in addressing the adverse economic effects of AI.

OpenAI Unveils Major Growth of London Office

February 26, 2026sparta

OpenAI has unveiled intentions to convert its London location into its most extensive research center outside of the United States. Founded in 2023, the UK office seeks to broaden its research team by drawing talent from premier British universities, though exact hiring figures remain unspecified.

Mark Chen, OpenAI’s chief research officer, underscored the globally recognized talent and institutions in the UK as pivotal factors for this expansion, reinforcing AI safety and utility. This initiative positions OpenAI in direct rivalry for top research talent with Google DeepMind, which is steered by British researcher Demis Hassabis and operates from London. DeepMind actively partners with Oxford and Cambridge Universities through professorships, funded research endeavors, and collaborative projects.

Recent career fairs at Oxford University demonstrated an increasing appetite for AI positions, marked by a rise in both job opportunities and student interest. Jonathan Black, director of the university’s careers service, highlighted the beneficial effects of these trends.

The London team will play a role in products such as Codex and GPT-5.2, while also prioritizing model safety, reliability, and performance assessment. UK science and technology secretary Liz Kendall commended the initiative as evidence of the UK’s preeminent position in AI research.

This announcement coincides with the UK’s initiatives to upgrade data center and power infrastructure to accommodate the growing computing needs of AI firms, including OpenAI.