NEW: I spoke to the leader of one of Facebook's biggest anti-AI data center groups As opposition to AI data centers grows across the country, Dally Holloway says legitimate local concerns risk being overshadowed by misinformation and conspiracy theories. www.groundlevel-ai.com/p/i-spoke-to...

NEW: I spent time in a tiny Texas town facing the prospect of a $10 billion AI data center. For my final Fortune story before launching Ground Level AI, I found out what happens when a megaproject lands next door—but your town has no authority to stop it www.groundlevel-ai.com/p/i-spent-ti...

Rising like fluffy angel food cake

The US wants Anthropic to secure its most powerful AI models. But one researcher says that is mathematically impossible open.substack.com/pub/sharongo...

My final story for Fortune as a staffer is the last of four deep-dives I reported on the ground about communities grappling with the AI data center boom: Two mayors, one $10 billion AI data center, and a growing divide in small-town Texas fortune.com/2026/06/16/a...

Yesterday was my last day as a staff reporter at Fortune. I'm walking through a new door. On June 16, you'll see where it leads! sharongoldman.substack.com/p/yesterday-...

Such a weird deal

Third in my series on the AI data center boom: In Saline Township, Michigan, residents voted down a giant OpenAI-Oracle data center planned for farmland — then watched construction begin anyway after a lawsuit and settlement. AI, power, zoning, and who gets a say. fortune.com/2026/05/06/a...

NEW: Tech has been killing my jobs for decades. Somehow I ended up covering the very thing (AI) that could kick us to the curb. open.substack.com/pub/sharongo...

My sister’s anatomy test reveals what AI image models still don’t understand OpenAI's Images-2 can generate beautiful visuals. But when it comes to structured knowledge like the human body, it still falls apart in ways only experts can see. sharongoldman.substack.com/p/my-sisters...

After the attack on Sam Altman, anti-AI groups took center stage. I took a closer look at two of them, Pause AI and Stop AI, this week for Fortune. (gift link: sharongoldman.substack.com/p/after-the-...) Some included details:

Pause AI and Stop AI: Meet the anti-AI groups facing questions after the attack on Sam Altman fortune.com/2026/04/15/p...

People will read 4000 words — or at least try — when it’s the kind of reporting they can’t get anywhere else. Books didn’t disappear, and neither will deeply-reported stories that have some real non-AI effort attached to them. open.substack.com/pub/sharongo...

How AI is affecting journalism was big news this week. My take: open.substack.com/pub/sharongo...

It is amazing that 2 or 3 years ago (maybe even 1?) no one would have written this story or been willing to speak on the record www.wired.com/story/tech-r...

NEW: Meta’s $27 billion AI data center is causing chaos in small town Louisiana I spent several days on the ground in Richland Parish last month, where a massive Meta AI data center is rising. The area has become a chaotic boomtown—with both winners and losers. fortune.com/2026/03/26/m...

Even with all the news dropping in AI this week, my Fortune editors were kind enough to let me go on an extended reporting trip for the series on the mega AI data center boom I’ve been working on. Here's a free preview: open.substack.com/pub/sharongo...

BREAKING: OpenAI robotics leader resigns over concerns about surveillance and autonomous weapons amid Pentagon contract fortune.com/2026/03/07/o...

BREAKING: Sam Altman told OpenAI employees at an all-hands meeting on Friday afternoon that a potential agreement is emerging with the Dept of War to use the startup’s AI models. The contract has not yet been signed. fortune.com/2026/02/27/o...

On Monday, AI influencer and CEO Matt Shumer reached out to say he was writing an article to explain to the average person “what’s going on in AI and how it’s going to affect them.” Here's what happened next. open.substack.com/pub/sharongo...

NEW: Just back from Richland Parish, Louisiana, where former cotton and soybean fields are now Meta’s Hyperion AI data center. It’s so big it has its own road: Far Far Away Lane. But this isn’t a galaxy far away — it’s right here in the U.S. More soon. sharongoldman.substack.com/p/reporting-...

Scoop: I learned Meta has quietly purchased roughly 1,400 acres—nearly 2X size of Manhattan’s Central Park—next to its 2,250-acre Hyperion AI data center site in Richland Parish, La., Fortune has learned. I also observed active work underway on the newly acquired land. fortune.com/2026/02/04/m...

Meta has quietly purchased roughly 1,400 acres—an area almost twice the size of Manhattan’s Central Park—adjacent to its already-mammoth 2,250-acre Hyperion AI data center site in Richland Parish, La., Fortune has learned. fortune.com/2026/02/04/m...

Citation is the foundation of academic promotion. It’s noisy, sure, but its integrity is worth fighting for. Hallucinated citations should be a desk reject.

Wow, thanks Toby!

In mid-2025 I reported fake references in a 2023 paper to NeurIPS Ethics and Grievances Committee. They said they had contacted the authors, who apologized & sent a corrected version. This story prompted me to check the NeurIPS proceedings, and it's still the original, uncorrected pdf.

I'm actually surprised it's not higher - it's not like reviewers and ACs are checking all the references.

What is wild to me is the defense, BY THE NEURIPS BOARD, that fabricated citations do not mean "the content of the papers themselves [is] necessarily invalidated" It does. It very much does. What do you think citing other work is for? What do you think writing a paper is for? What do you *think*?

NeurIPS reviewers typically review 6+ papers, each with 100 or so citations. The reviewers can't possibly check that every citation is real. But why don't we have an automated way to do this checking (and desk-rejecting if fake citations found) *before* the papers are sent to reviewers?

This community knows just how unreliable AI sourcing is, and yet. The cognitive bias toward trusting these things is so strong even the people who should really, really know better can't help themselves. Which means you (and I) probably can't either. Be careful out there, knowledge workers!

Accurate and thorough representation of prior and related work is one of the cornerstones of good research. It is shocking to me that so many published NeurIPS papers, even from top institutions, have fabricated references. I recommend reading the original report: gptzero.me/news/neurips/

The disparity between fields and rigor of peer-review is wild. I've been scrutinized for citing the wrong edition of a particular book, meanwhile some people are just having AI write the whole damn lit review apparently.

NEW: NeurIPS,one of the world’s top academic AI conferences, accepted research papers with 100+ AI-hallucinated citations, new report claims fortune.com/2026/01/21/n...

I’ve been covering AI as a daily beat since April 2022. From the start, the AI beat felt more fast-paced than anything I had covered in more than two decades as a journalist, writer, and editor. open.substack.com/pub/sharongo...

We talk about AI in abstractions — compute, scale, trillions. But on the ground, it’s about water, land, noise, jobs, and who gets a say. I spent time in the Arizona desert reporting on what the AI data center boom looks like up close. open.substack.com/pub/sharongo...

That tension's the real story behind the tech headlines.

NEW: A few weeks ago, I traveled to the desert outside Phoenix to report a story that’s a true microcosm of the AI data-center boom: Silicon Valley billions colliding with rural America, land, power, water, and local politics.

NEW: AI data centers have suddenly become one of the most potent political flashpoints of 2025, igniting fierce debates over power, water, land, and jobs. fortune.com/2025/10/22/a...

From Fortune reporter @sharongoldman.bsky.social's email announcing this story: "Reporting the story, what struck me most is how belief itself has become a form of capital — how people are literally investing in a worldview about where AI is heading." fortune.com/2025/10/08/l...

How a 23-year old former OpenAI researcher turned a viral AI prophecy into profit, with a $1.5 billion hedge fund and outsized influence from Silicon Valley to D.C. I spent several weeks reporting on the unlikely rise of Leopold Aschenbrenner 👇 fortune.com/2025/10/08/l...

Many assume C2PA fights deepfakes. But a new report from the World Privacy Forum says it doesn’t — instead it’s building a vast metadata system that raises thorny questions about privacy & who gets to count as ‘trusted. Thanks to @katekaye.bsky.social for chatting! fortune.com/2025/09/18/b...

It's a podcast world (I regularly think about being at the crazy AI Action Plan rollout a couple of months ago with the All-In bros where JD Vance sat on a couch with them and later the whole thing was rolled away to make way for Trump and his lectern and American flags )

"Person says: "So you mean, they're like dumb animals". Me: "HOW DARE YOU SIR ANIMALS EXHIBIT RICH MORAL SENSIBILITIES BACK AWAY FROM MY OTTERS AND TURTLES" (ok maybe I didn't quite say it that way)" Yaaaas

Most AI red teams break models. Anthropic’s Frontier Red Team also evangelizes the risks of AI itself — a rare hybrid of security + policy. I did a deep dive on this unusual group that sits in Anthropic's policy organization under co-founder Jack Clark. fortune.com/2025/09/04/a...

“Late last year, David Shor, one of the Democratic Party’s top data scientists, surveyed some 130,000 voters about whether they had a “favorable” or “unfavorable” opinion of Jewish people. Hardly anyone over the age of 70 said their view was unfavorable. More than a quarter of those under 25 did.”

Yes amen exactly this

NEW: China’s DeepSeek quietly released an open-source rival to GPT-5 — optimized for Chinese chips and priced to undercut OpenAI. Here's what that means and why that's a big deal. fortune.com/2025/08/21/c...

Superintelligence, shmooperintelligence - for Meta, it's still always going to be about ad revenue, which is, hello, like 98% of revenue 🤷‍♀️ fortune.com/2025/08/20/m...