25 Years After 9/11: New Health Records, an AI‑Driven Math Feud, and Rising Data‑Center Pollution Fears

It’s hard to believe that a quarter‑century has passed since the towers fell, yet the fallout is still being written, page by page. On Tuesday Mayor Zohran Mamdani stood before reporters and announced that more than 170,000 pages of internal city records are now online – air‑quality tests, memos about asbestos, and dozens of other files that show officials knowingly told the public the air around Ground Zero was safe, even when the data said otherwise.

The newly searchable portal (sept11documents.cityofnewyork.us) is a stark reminder that roughly 400,000 New Yorkers were exposed to a toxic cocktail of dust, smoke, and stress after the attacks, according to the CDC. The World Trade Center Health Program, which offers free monitoring and treatment, already serves about 140,000 people, but many who qualify have yet to enroll. If you lived or worked near Canal Street, DUMBO, Brooklyn Heights, or even farther afield on that day, you can still apply for help at cdc.gov/wtc/apply.html.

While the nation pauses to reflect on that tragedy, another kind of disruption is making headlines: artificial intelligence claiming a prize that mathematicians have chased for decades. Two researchers using Anthropic’s Fable model, cross‑checking results with OpenAI’s ChatGPT and Codex, announced a breakthrough on the Navier‑Stokes equations – the very problem that underlies our understanding of fluid flow. Almost simultaneously, an internal OpenAI system reported it had cracked the same problem within days, using a method strikingly similar to the one the Anthropic team had been feeding it.

Who, then, deserves credit? The math community is split. Some argue the OpenAI model merely recombined prompts supplied by the researchers; OpenAI counters that no user data was accessed. The debate has turned surprisingly heated, with a few senior mathematicians warning that credit‑allocation in an age of machine‑augmented discovery may never be clear again. Regardless of the politics, solving a Millennium Prize problem – a challenge worth a million dollars and half a century of effort – marks a seismic shift in how we think about mathematical insight.

And while the AI triumphs (or tussles) capture imaginations, a quieter, yet equally pressing issue is gathering steam: the environmental toll of the AI boom. Training massive models consumes staggering amounts of electricity, and the data centers that house them are increasingly blamed for higher carbon footprints. Recent studies point to a jump in power demand that rivals the output of small cities, prompting calls for greener cooling systems, renewable‑energy contracts, and stricter reporting standards. As AI spreads into more corners of society, the conversation about progress now has to include a hard look at the planet we’re powering.