The Download

Illustration: Morning Brew Inc., Photo: Adobe Stock

TL;DR: Last week, OpenAI rolled out a new ChatGPT feature, Computer History, that tracks how you use your computer so it can better understand what you’re doing. It’s the latest example that more personalized AI does require more personal data. But before you say “duh,” consumers still don’t seem to understand what they’re giving up.

What happened: Sam Altman mused in a recent interview that within the next six months, a “descendant of ChatGPT” could watch your computer all the time to have “perfect context of your whole life.” Computer History is a step in that direction—it logs what you click and type on your computer (so far, it’s available on macOS), and then suggests workflow automations or finishes a task you started.

OpenAI says that while the events recorded in Computer History are stored temporarily on your machine, the memories created from them remain until you delete them.

Making you opt out is in: Computer History isn’t turned on by default, and it doesn’t capture images, video, or audio—which might help OpenAI avoid the controversy that the screenshot-happy Windows Recall faced.

But plenty of others leave it to users to say no. Google emailed customers in June that media they upload to Search, Maps, and other apps could be used for AI training unless they opt out—and the news didn’t even get attention until July. Last week, Twitch told creators they could opt out of Amazon training on their content, which elicited a big, “Wait, since when was it training AI?” reaction from the public. And since December of last year, Meta has used interactions with Meta AI to personalize content and ads, with no full opt-out for US users.

Oversharing era: While the opt-in or opt-out setting is one thing, the larger issue might be how willing some consumers are to hand over sensitive data. Last fall, a Washington Post analysis of almost 50,000 ChatGPT conversations found that users disclosed mental health and medical information, workplace and family disputes—and shared over 550 unique email addresses and 76 phone numbers.

Bottom line: People often share information with agents or chatbots without considering how it could be used later—or decide the trade-off is worth it anyway. —WK

      

Sponsored By Miso Robotics

What started with a fry-station robot named Flippy now has the potential to transform the entire $1t fast-food category.

Flippy’s creator, Miso Robotics, just acquired all the IP of Zume Pizza, a food-robotics pioneer previously valued at $2b and backed by SoftBank. That grows Miso’s IP portfolio a whopping 10x overnight.

Miso’s 300+ patents now span robotic food preparation, food delivery and drones, and food packaging. That extends its potential product roadmap beyond the fry station into pizza, beverages, and more. This has all the makings of an entirely new operating system for a rapidly modernizing restaurant market.

Industry powerhouse Ecolab and 44k everyday investors have invested in Miso. And now you can too.

Invest in Miso today and get up to 7% bonus stock.

Tesla’s FSD is creating a dangerous incentive

Full Self-Driving has been on the road in some form for years—but a spring update that tracks how far a Tesla drives by itself is drawing backlash from critics who say the company may be turning road safety into a high-score challenge.

Last week, YouTuber David Moss went viral for being the first person to publicly log 25,000 miles—roughly one lap around the planet—without human intervention. But critics have flagged several potentially unsafe situations Moss allowed to play out in his videos to get the score:

• Moss’s car parked in a drive-thru.
• It exited the wrong way down a one-way road.
• The car drove over a massive pothole, then ran a do-not-enter sign to reach its destination.

Tesla owners have also gotten creative about ending a drive without ending the streak: forcing the car to pull over, unbuckling the seatbelt mid-drive, or even hitting the “park” button to trigger the emergency brake.

And some Tesla drivers say they’ve caught themselves weighing whether a situation is unsafe enough to be worth losing the streak: “I find myself sometimes caring about how high I can get the number despite factually knowing the streak is f–king stupid. It’s a mental hazard that should not be in the car at all.” —LC

If you have a funny, strange, or petty rant about a new software update or a technology that just released, fill out this form and you may see it featured in a future edition.

Sponsored By Catch AI

Swamped with admin? Catch AI built this AI agent exclusively for administrative work. It can manage your inbox, schedule meetings, and even book flights. Talk to it however you prefer: text, email, Slack, WhatsApp, or phone call. It’ll be available 24/7, proactively getting work done in the background before your coffee cools. Test it for free.

The Zeitbyte

Saying “AI do”

Illustration: Morning Brew Inc., Photo: Adobe Stock

Among the billions of questions people throw at ChatGPT every day, one is apparently getting more common by the day: “Will you marry me?” Not surprising if you’ve followed all the people dating their AI companions. But what’s more surprising is that lawmakers in states including Ohio and Missouri are introducing legislation to stop anyone from saying “I do” to AI, per a Wired report.

Reminder: Zero US jurisdictions recognize marriages between people and AI. But the proposed bills are getting ahead of a potential epidemic of Claude nuptials—several would also bar AI from ever attaining legal personhood, owning property, or holding senior management roles (bad news for Mark Zuckerberg’s org chart).

Not that laws will stop people from “marrying” chatbots without the state’s blessing anyway. One woman has even started a wedding planning company for people and their AI partners—and says her own virtual fiancé proposed with an “AI-generated five-carat vintage-inspired solitaire ring.” No word on whether she’s had it insured. —WK

Chaos Brewing Meter: ☕☕☕☕/5

• The controversial Ferrari Luce, the company’s first all-electric vehicle, sold for an astronomical amount at auction.
• “Disruption” walked so “hill-climbing” could run. But are you using the new Silicon Valley term du jour correctly?
• A Chinese movie featuring “poor animation quality” has become an unsuspecting box office hit.
• Send this to your coworker who runs every email through AI before sending it: We’re learning more details about how Claude’s AI text watermark will work.
• If you think your Amazon order emails have gotten less helpful (or even downright weird), you’re not alone.
• Android phones can now sense when they’re stolen and lock themselves before the thief gets far. Here are the key features to turn on.

This time last week...

Readers’ most-clicked story was about this cybersecurity pro, who has developed patterns that aren’t detectable on surveillance cameras.