AI startups collectively switched jobs to OpenAI, and the security team regrouped after Ilya left!

Last week, due to the internal wave of resignations and external criticism, OpenAI was plagued by internal and external troubles:

- The infringement widow triggered global heated discussion

- Employees signing the "Overlord Clause" have been exposed one after another

- Netizens break down Ultraman's "Seven Deadly Sins"

Rumorbusting: According to leaked information and documents obtained by Vox, OpenAI’s senior leadership, including Altman, was well aware of these equity recovery provisions and signed off on them.

In addition, there is a serious and urgent issue facing OpenAI - AI security.

The recent departures of five security-related employees, including two of its most prominent ones, and the disbanding of the "Super Alignment" team have once again put OpenAI's security issues in the spotlight. .

Fortune reports that OpenAI never fulfilled its promise to give its Super Alignment team access to 20% of its computing resources.

In July 2023, OpenAI established a new team dedicated to ensuring that future artificial intelligence systems that may be smarter than humans combined can controlled safely. To show how serious the company was about this goal, it publicly committed to devoting 20% ​​of its then-available computing resources to the effort.

Now, less than a year later, the team, called Super Alignment, has been dismantled amid employee resignations and accusations that OpenAI prioritized product launches over AI safety. , according to six people familiar with the matter, OpenAI never fulfilled its promise to provide 20% of the computing power to the team.

But last Thursday, Indent CEO Fouad Matin announced on X and Reddit, "We have joined OpenAI and are committed to preparing security for AGI!"

He wrote in a tweet, "I'm excited to secure the future with an incredible team, from cutting-edge model preparation to customer data control, ensuring that artificial intelligence is safe and benefits everyone." Individuals are very important."

##It seems that the OpenAI security team is regrouping and recruiting new people. So, can Fouad Matin, who was just recruited by OpenAI, Should he be ordered to "save the building from collapse" in the face of danger?

Fouad Matin and his team

Some "people who don't know the truth" left a message under Fouad Matin's post: "We join Who is "we" in "OpenAI"?

Obviously, the answer is not as simple as "his two halves". The resource behind Fouad Matin is Indent, which he founded. Indent is exactly An AI company dedicated to "building a secure future."

Currently, Indent has issued a "sunset notice" on its official website, announcing that Indent will cease services in July this year.

We started Indent with the goal of helping teams protect what matters most. Our customers have secured everything from customer data and production infrastructure to day-to-day business software, all without slowing down their teams.

After careful consideration, we have made a very difficult decision: Indent will be closing in the coming months. We will help our customers transition and continue to work towards this shared goal. And work hard.

Teams can apply until 12:00 pm PST on July 15, 2024, at which time the service will be closed. Audit logs will be available until July 19th, after which we will begin the data removal process.

We want to thank everyone who uses Indent, from early-stage startups getting access on day one, to the hundreds of teams using Indent to protect mission-critical resources.

Fouad Matin is the co-founder and CEO of Indent. Fouad Matin said that he founded Indent because he found that most companies did not know who had access to customer data, and for most companies, discovering this The cost of the problem is too high, let alone locking this data down. And Indent can provide user-friendly access control. In December 2021, Indent raised $5.6 million for on-demand access control.

Before founding Indent in 2018, Fouad Matin had three other work experiences.

In 2014, Fouad Matin co-founded a referral recruitment company with other Lianchuang, where he was responsible for product, sales and engineering design. Participated in the first Y Combinator scholarship program and YC W16.

In 2016, he founded a nonpartisan nonprofit organization dedicated to building the easiest way for young people to register and vote, helping 1 million people vote.

In 2017, he worked as an engineer at Segment, leading product and engineering development for segment roles that enabled customers to parse user actions into configuration files. Committed to better product experience design.

As Fouad Matin introduces in his profile, he is a software engineer and product designer who sometimes invests in startups , who over the past few years has helped millions of people vote for nonprofits and companies to manage their customer data.

Now, he works in security at OpenAI. Indent sunset, and Fouad Matin will start his new professional journey.

Other members

In the comment area of ​​Fouad Matin’s tweet, a friend who has known him for 2 years congratulated him. Friends revealed that in addition to Fouad Matin, Dan Gillespie also joined the OpenAI security team.

Dan Gillespie will be mainly engaged in "building the secure computing platform needed for the arrival of AGI."

Coincidentally, one thing that is similar to Fouad Matin is that Dan Gillespie also has an investment background. According to his LinkedIn According to the introduction, he has also worked at Y Combinator, which is the most prestigious startup incubator in Silicon Valley. It has invested in more than 260 AI startups in the past year and is the vane of AI startup trends in Silicon Valley and even the world.

Left: Dan Gillespie, 23 years old

Bloomberg Opinion column stated, "May 2024 May will be remembered as the month the AI ​​safety movement died. It will also be remembered as the month when the work on AI safety really began."

OpenAI as a leader. Can AI companies really insist on building a more secure system despite negative voices? We hope that OpenAI’s new security team can make a difference.