Breaking

It’s official: The first foldable iPhone was unveiled today at Apple’s Surprise and Shine event. Unfortunately, the company did not call it the Fapple or the iPhold, but the Apple Duo. It will have an asymmetric shape and be the largest iPhone display Apple has ever made.

The Download

Illustration: Morning Brew Inc., Photo: Adobe Stock

TL;DR: Yesterday, Google DeepMind released a massive, searchable AI-powered database of every possible one-letter change in human DNA (all 9 billion of them). It also predicts how each change could affect you. The database could help speed up rare disease and cancer research, but some experts are cautioning against using predictions as definitive answers.

What happened: The AlphaGenome Atlas database, which is available for free to academic researchers, has 1 petabyte’s worth of data. (That’s equivalent to 500 hard drives with 2TB of storage each…so, uh, no need to clear space on your laptop). It’s also more than 30 times larger than the team’s database of protein structures, AlphaFold—the work that won former DeepMind chief Demis Hassabis a share of the 2024 chemistry Nobel Prize.

Why this matters: A single-letter change in DNA can be the cause of diseases like sickle cell anemia and cystic fibrosis. More than half of rare disease patients go undiagnosed, partly because clinicians still struggle to identify the underlying DNA changes. Some mutations have also been tied to certain cancers—finding them normally involves sifting through billions of data points, which the Atlas could now help with.

DeepMind also released the AlphaGenome Variant Impact (AVI) score, which rates how much a genetic mutation is likely to matter. It blends AlphaGenome’s predictions with an earlier model that predicts whether small changes in proteins are harmful. The AVI score could help researchers sort DNA changes by importance and investigate what each one could do.

But wait: For now, Atlas is just a starting point for researchers. Some experts are questioning how accurate the predictions are and how researchers interpret the results.

While the AVI score’s tidiness is useful, it could be “easily misinterpreted,” says University of British Columbia genomicist Carl de Boer. Others say Atlas won’t replace experiments or account for individual cases in diagnoses—and warn that it shouldn’t be driving clinical decisions on its own.

DeepMind’s genomic lead acknowledged that the Atlas predictions are not as accurate as AlphaFold, and can miss some changes to regions like enhancers (DNA that makes other genes more active).

Bottom line: For the first time, researchers can look up any single-letter change in the human genome for free and get a quick read on whether it matters. What they do with that information will determine its usefulness. —LC

      

From The Crew

Most startup content tells you what worked…after everything already went right. Founder Brew goes deeper.

Every issue covers the decisions that defined a company’s trajectory: the pivots that paid off, the bets that didn’t, and the frameworks founders wish they’d had earlier. It’s the newsletter for people who want more than a highlight reel—investors tracking what’s next, operators building in real time, and obsessives who can’t stop studying the game.

If you want the behind-the-scenes reality of building great companies, you’ve found your place. Subscribe to Founder Brew today.

• 🪳 Your next first responder could be a cyborg cockroach. A team of researchers created “paraborg” (paramedic cyborg) roaches to help find survivors of disasters and provide first aid in places too dangerous or difficult for humans to reach. The roaches, which are capable of carrying 1.5 times their body weight, were equipped with either a camera or an automated injector designed to carry a liquid drug (such as adrenaline)—then remotely controlled through electrodes embedded in their antennae and abdomen. Medications injected at close range (within 150 millimeters of the target) succeeded about 95% of the time, while the success rate for the complete run from roach deployment to injection came in at 72%. (And if you’re a roach-lover, fear not: They all apparently lived normal lives after the equipment was removed.) The researchers hope to deploy a roach rescue team to disaster sites within five to 10 years with more field testing. —LC

The Zeitbyte

OpenAI’s million-dollar math beef

Illustration: Morning Brew Inc., Photos: Adobe Stock, Unsplash

Imagine spending a year trying to prove something that could get your name etched into the math hall of fame—and then a giant AI lab swoops in and takes all the glory. That’s the accusation being lobbed at OpenAI after it declared this week that its models had answered a longstanding question about a 200-year-old equation. The issue? Per NYU mathematician Tristan Buckmaster, the company’s solution (suspiciously) uses the same mathematical strategy he and his partner Levent Alpöge had been pursuing as a side project with the aid of several AI tools, including Codex. Oh, and the partner in question happens to work at Anthropic.

Buckmaster says he reached out to OpenAI researcher Sébastien Bubeck with his concerns. And the latter, he claims, offered ideas for who’d get credit for what—including one option where Alpöge specifically would go uncredited in a paper about the proof. (Counterpoint: Bubeck says he never suggested that.)

OpenAI, for its part, swears it didn’t see any of the pair’s work in advance—though it “cannot rule out” that their de-identified data helped train its AI. But the lab does admit that it went into full proof-maxxing mode only after hearing Anthropic might be close to a breakthrough—a rumor that turned out to be about Alpöge—and threw 10,000 AI agents and millions of dollars in compute at the problem. However this shakes out, here’s a sentence we never thought we’d say: Math beef has us on the edge of our seats. —WK

Chaos Brewing Meter: ☕☕☕☕/5

Sponsored By ActiveCampaign

It’s like flipping a switch. Outgrown your email platform? Switching doesn’t have to feel like an expensive, risky rebuild. ActiveCampaign helps teams migrate in minutes. Migration Agents help move your contacts, data, automations, and history over automatically—no exports or manual imports required. See what else they offer.

• Meta just released a personal AI agent called Muse, and it comes with a free tier—which gives you 100 million tokens per week.
• Another day, another AI researcher quitting their job over fears the tech is getting out of hand.
• Some cities have stopped using their Flock cameras after all the backlash—only to replace them with a different brand of license plate reader.
• A startup hopes to put an end to sky-high memory prices using a new chip design.
• ChatGPT’s Sketch feature can turn a stick-figure drawing of your family into a portrait worth adding to your fridge.
• Here’s something to do during your lunch break today: Explore the largest map of the universe ever created.
• This interactive tool lets you see a random life from the past—showing you a vivid portrait of what living looked like back then.

This time last week...

Readers’ most-clicked story was about this $40,000 solar-powered car that looks like a sci-fi dune buggy.