Meta's AI Just Turned Brain Waves Into Real Sentences. Here's How.

On July 16th at 1PM ET, beehiiv is going live with a look at the future of publishing, audience growth, and digital business.

What started as a newsletter platform has evolved into something much bigger: a place where creators and brands can grow, monetize, and own their audiences without stitching together half the internet to make it work.

The next chapter starts live at the Summer Release Event. 

Join us to see what’s coming next.

RSVP now.

Two things that happened while you were busy

1. OpenClaw and Cursor both go mobile.

OpenClaw, the platform that popularized personal AI assistants back in January, now has native apps on iOS and Android, so you can run your assistant straight from your phone. Cursor, the AI coding tool, shipped its own iOS app too, letting you kick off always-on cloud agents from anywhere. Its release video on how that works has already passed 3M views.

2. Samsung, SK Hynix, and Micron get hit with a price-fixing lawsuit.

A new California lawsuit accuses the three memory giants, which together control roughly 90 percent of the global DRAM market, of coordinating to restrict supply and inflate prices. The suit lands as memory prices have climbed sharply over the past year, pushing up the cost of everything from Apple Macs to Microsoft's Xbox.

Reading thoughts into text is no longer science fiction. It is just early.

The upgrade. Meta released Brain2Qwerty v2, which reads raw brain signals from a noninvasive scanner and decodes them into sentences in real time at 61 percent accuracy. That makes it the company's best brain-to-text system so far, and it is trained on 10 times more data than last year's version.

How it learned. The model studied roughly 22,000 sentences gathered from nine volunteers, each recorded for 10 hours in front of the scanner. The point is not faster typing for everyone. It is to give a voice back to people who can no longer speak after a stroke, an accident, or a brain disorder. You can read the technical write-up here.

The bigger picture. What makes this notable is the word noninvasive. No surgery, no implant, just a scanner reading signals from outside the skull. Accuracy is still well short of perfect, closer to a rough draft than a finished sentence, but the trajectory is the real story. Each upgrade like this is why machines reading human thought feels less like a someday and more like a soon.

What people are actually watching and sharing

Better alternatives. It is frustrating when an AI does exactly what you said, even when there is a smarter path. One Redditor fixed that by adding a rule that tells the model to always suggest a better alternative, and the full prompt has 2.5K upvotes.

Five future roles. The creator of Claude Code put out a widely shared post on where tech jobs are heading, predicting five role archetypes that replace today's domain-specific titles. The thread has 2M views.

LLM tells. People are getting sharper at spotting AI writing, and the newest giveaway is the "fake profound" one-liner. If you are borrowing from a model, rephrase that structure before you publish.

Cartoon to live action. AI video tools now make it easy to turn 2D animation into 3D live action. A viral Tom and Jerry clip shows the effect, and it has pulled in 3.5M views.

Privacy guardrail. The National Design Studio released Rampart, an on-device tool that keeps your personal information walled off from AI chatbots. It is a small but useful step if you worry about what you paste into these tools.

Write a four-message follow-up sequence that closes without nagging

Chasing a prospect who went quiet is awkward, and most follow-ups either sound desperate or get ignored. Drop this into ChatGPT and it writes a four-message sequence that checks in, handles the unspoken objection, adds gentle urgency, and bows out gracefully, all under 80 words each. It is built to keep you on a prospect's radar without becoming the person they dread hearing from.