Brought to you by

💦 Welcome, friend. AI keeps moving at light speed, and you’re right alongside it. Come on in and make a Splash. 

HERE’S WHAT’S IN THIS ISSUE:

Let’s start with a brain teaser, shall we? 

Science reached the “let’s pipe Pink Floyd into a brain and see if the meat radio picks up a signal” stage. Researchers put electrodes on the surface of awake patients’ brains, played “Another Brick in the Wall” and recorded brain waves. 

🎶 Then the scientists, being scientists, fed all those squiggly signals into AI and asked it to do something that lived strictly in science fiction.

What did the AI manage to reconstruct from reading the brain waves? A) A typed transcript of the song’s lyrics, B) A recognizable audio clip of the song, C) A color-coded readout of each listener’s mood or D) An invoice from the band’s lawyers for unlicensed streaming? The answer is waiting at the bottom of today’s newsletter, one more brick in the wall away.

🎉 Love getting this free newsletter? Hit that ⭐️ or “Favorite” in your email app, so I don’t disappear into your spam or promotions tab. You can’t stay AI ahead if I go missing! — Kim

A friend of mine graduated medical school last month. While he waits to hear back from residency programs, he’s making $85 an hour doing a job that didn’t exist a few years ago: flagging what an AI model gets wrong.

It gets weird. One time, the AI he was working with suddenly stopped, clocked his pace at around hour 14 of an 18-hour grind and told him to go outside and take a picture of a tree before continuing. 

Nothing says “future of work” like being bullied into touching grass by a chatbot.

AI makes confident, polished, wildly wrong mistakes. So companies are hiring the one thing their models don’t have: humans who know what wrong looks like.

These positions go by a few names: AI trainer, response rater, AI tutor. You read what a chatbot spits out, score it and rewrite the bad parts. No coding. No fancy degree. Usually, you have to pass a skills test.

Generalists make $15 to $40 an hour, often from home, often from your couch. 

Doctors, nurses, lawyers, accountants and other specialists pull in over $100 an hour.

Then there’s red teaming: being a professional troublemaker, trying to trick the AI into saying something it shouldn’t so the company can patch the hole before a stranger finds it.

At the top, Anthropic, OpenAI and Google pay $150,000 to over $300,000 a year for people who find an AI’s failures before the public does.

The door’s wide open at the bottom. Build a track record on the gig sites first, then climb: DataAnnotation, Indeed, Outlier and other job boards.

The robots aren’t taking jobs. They’re creating new ones for people paid to say, “Actually, no.” 

📩 Send this to someone who’s between jobs, in school or hunting for extra income.

📺 WATCH ON YOUTUBE NOW OR LATER

A SpaceX welder turned stock bonuses into an $800,000 payday. Not bad for a guy holding the torch while Silicon Valley holds the hoodie. This week, we break down AI court filings gone wrong, Walmart price-switching drama, Apple’s new AI-powered Siri and a Tesla-concept baby stroller. 

Also this week: Flock cameras tracking cars, DoorDash ordering by photo and NASA’s supersonic jet that could make New York to London a three-hour trip.

Plus, how Apple crash detection helped a father reach his daughter after a serious accident. 

Hit play below, so you’re the smartest person around. 👇

▶️ SpaceX riches, AI lawsuits and driverless car drama →

🎧 Or search “Komando” wherever you get your podcasts. I’m everywhere.

You paste a PDF into your AI for a quick summary. A contract, a bill, a flyer, whatever. Harmless, right? Not always. A file can carry hidden orders, and your assistant will follow them.

It’s called indirect prompt injection. (Rolls off the tongue doesn’t it?)

Someone buries instructions inside the document, invisible to you, plain as day to the AI. You didn’t click a link or download malware. You asked it to read something boring. The boring thing told it what to do next.

Hiding spots are ordinary: white text on white, tiny off-screen print, file metadata, calendar invites, alt text, QR codes.

The danger lands when your assistant is wired into Gmail, Drive or Slack. It can grab private data and quietly ship it out in a web request. You never see it happen.

When you want a quick summary of some random file, use a plain chatbot that isn’t connected to anything. No Gmail, no Drive, no Slack, no web browsing switched on. Think of it as letting a stranger read you a recipe out loud versus handing them your house keys. 

If the AI isn’t connected to anything, a hidden order has nothing to steal and nowhere to send it.

As an Amazon Associate, some links pay us a commission at no extra cost to you. Keeps this newsletter free. Thank you.

Lost your remote? The Alexa Voice Remote Pro (34% off, $23) helps you find it fast with voice controls. Backlit buttons make scrolling less squinty.

📺 No drill, no mess: This studless TV wall mount (33% off, $20) hangs screens up to 55 inches and 99 pounds without screws or wall damage.

Snap-and-go clips: These little cord organizers (15% off, $10, eight-pack) keep your charging cables neat and within reach. Stick, lock, done.

💻 “Kim, what’s the best computer?” I get that question all the time. Here are my top 20, from budget Chromebooks to premium workstations.

Amazon hit a number that sounds made up. It runs 1 million robots. Here’s the coffee-spit part: That’s almost one robot for every human employee it has. 

It’s become the world’s largest maker and operator of mobile robots. At this point, the warehouse holiday party needs assigned charging stations.

Here’s the weird part. At Amazon’s newest robotic center in Louisiana, the high-tech setup needs 30% more people in skilled maintenance and engineering roles.

Yes, you read that right. More robots. More humans.

The warehouse isn’t disappearing. It’s changing. The bots handle the backbreaking lifting, hauling and sorting. People move toward fixing, managing and running the machine brain.

I saw a robot charging itself the other day. It was re-volting. (Was that a groan I heard?)

WHAT IT DOES: Creates and edits video from text, images, audio, sketches and reference clips. Upload a clip, tell it what to keep, change one thing, then keep revising without regenerating the whole thing from scratch while you slowly lose faith in tech.

WHY IT’S COOL: Need fast video concepts instead of disappearing in After Effects until Sunday? You’ll be surprised at how good AI video is getting. Skip it if you need broadcast-ready output, exact frame control or guaranteed legal clearance.

WHO IT’S FOR: Creators, marketers, educators, founders, product teams and social teams. 

PRICE: Google says Gemini Omni Flash is available on eligible AI Plus, Pro or Ultra plans. It’s rolling out, so you may or may not have it yet.

HOW TO FIND IT: Here’s the link. 

HOW I’D USE IT: Upload a six-second clip and say:

Then get specific.

You can even upload a sketch and turn it into realistic footage, map a character image onto someone walking or make a short explainer about protein folding or a marble run.

THE CATCH: It’s paid, limited and consistency is still a problem. Faces, hands, props, logos and text may drift.

Remember when robots were going to take all our jobs? One skipped the job entirely and went straight to panhandling.

A video out of China shows a Unitree G1 humanoid robot down on its knees on a sidewalk, begging people for money. Little donation bowl. A QR code. A handwritten sign pleading for help with its electricity bill. And yes, a bystander actually dropped something in the bowl. (We have officially out-empathized our own common sense.)

The bot reportedly whimpered lines like “My battery is low, please help.”

I’d have given it a dollar, but I figured it could just charge it.

Watch the action here.

Share this now:

Here’s your haul for today. You can hand AI your whole medicine cabinet and catch a dangerous drug combo. You can get paid, sometimes very well, to catch the mistakes AI makes. You learned to keep sketchy PDFs away from any AI that’s wired into your email, and Amazon’s million robots somehow mean more human jobs, not fewer. Not bad for five minutes of reading.

🎧 The answer: B) A recognizable audio clip of the song. The researchers found a dreamy twin of “Another Brick in the Wall,” hazy but unmistakable. Go ahead, watch the music video now.

The Berkeley team, led by Ludovic Bellier and neuroscientist Robert Knight, hooked up 29 epilepsy patients with roughly 2,668 electrodes and trained 128 separate AI models to turn brainwaves into sound. 

It’s the first song ever rebuilt from inside a human skull. And it’s not a party trick. The goal is giving people who’ve lost the ability to speak a way to talk again, with real melody and emotion in their voice. Amazing.

My brain on a Monday: 21 tabs open, five of them crashed, and I still can’t figure out where the music’s coming from. Turns out Berkeley can.

🧠 Scrolled past this one? The Echo Show 8 wears a lot of hats. It’s a smart home hub, personal assistant and entertainment center rolled into one. Alexa+ can handle groceries, recipes, reservations and all those little chores you’d rather skip. The early Prime Day discount won’t last. Snag yours while it’s still on sale. Want more handpicked deals? Sign up here, so you’re set for Prime Day.

Forward this to ONE person who needs to be with the AI program. Thanks for diving in, see you next time! — Kim

Kim Komando • Komando.com • 510+ radio stations • Trusted by millions daily

Your referrals get you great rewards!

Send your unique link below to friends and family.

👉 Your link is: ai.thecurrentnewsletter.com/subscribe?ref={{rp_referral_code}}

They get AI smart. You get prizes. Win-win. The more referrals, the more prizes. (Yes, even a meet and greet with me. I’d love that!)

Your referral count is: {{ rp_num_referrals }}. FYI: This only changes if the people you refer actually click to sign up for this free newsletter.

You’re {{ rp_num_referrals_until_next_milestone }} referrals away from {{ rp_next_milestone_name }}.

🎉 Keep it going! You got this! — Kim

Photo credit(s): ChatGPT/Kim Komando, nonser, Gemini, Jam Press

Companies and products denoted by an asterisk (*) within this publication are paid sponsors or advertisements. As an Amazon Associate, the publisher earns from qualifying purchases. Statements regarding products denoted by a double asterisk (**) have not been evaluated by the Food and Drug Administration; such products are not intended to diagnose, treat, cure, or prevent any disease. This newsletter is provided for informational and entertainment purposes only and does not constitute legal, financial, medical, or professional advice of any kind. Readers should consult with a qualified professional before making any decisions based on this content. The publisher disclaims all liability for any loss, damage, or injury resulting from the use of or reliance on the information contained herein.