1/ In 2019 we made this video internally at ZUtA Labs. I've never shared it publicly. This week's reporting on OpenAI's first device made me dig it up. Watch it, read @markgurman's description of what's shipping - and let me tell you exactly what Sam Altman told me about this.

Did they actually just put all their chips on the Sixers?? 🧐

Netflix is crashing and slowly becoming irrelevant - and it’ll be eaten by the one thing it spearheaded. Netflix reports its Q3 earnings tonight, and I don't think Netflix its problem is competition. Last week Netflix cut about 1,000 jobs, the biggest layoff in its history. Its chief content officer is out. And in the first half of 2026, viewing on Netflix grew only 2%. A few weeks ago I wrote an article called "The Last Campfire", about a Hebrew concept called מדורת השבט, “the tribe's campfire”. It's a term to describe the shared story a society gathers around: the village fire. For many years, it used to be the culture, the tradition that connected us. Then in the industrial age, it became the morning paper, followed by radio and TV, to the point when in one night 106 million Americans watched the final episode of MAS*H together, at the same moment. Netflix is one of the companies that began to put that fire out. Netflix created the binge - watching shows on our own schedule, alone. But more than that, Netflix worked hard so that every person got their own homepage and their own recommendations, all driven by an algorithm built to keep you watching, as much as possible. It wasn’t about giving us something to share together - it was about maximizing engagement. Two people in the same house could watch Netflix every night and never see the same thing. Netflix wishes it could create a campfire - a show everyone talks about. But the tools and tech it built - drove the exact opposite - everyone having their own. No matter who you are and where you are in the world, you can find that one show that’s perfect for YOU. Doesn’t matter how different or obscure you are, from your own society that you live with - around you. And now, AI is of course taking it all the way. When a model can generate a show just for you - it can be shorter, or in your language, or even have the main character a redhead and don’t have the dog die - you don't need Netflix's library anymore. You don't need anyone's. The company that personalized the evening is about to be personalized out of it. Netflix seems to know it. The company that taught the world to binge-watch has gone back to releasing episodes weekly, and it's betting on live events. Those are exactly the things the old campfire had. Tonight's earnings will tell us how Netflix's business is doing. They won't bring back the evenings we used to share. And I doubt they’re even aware or care about the importance for a society to have a worthwhile campfire to gather around. Logging onto Netflix, browsing around and finding nothing to watch with my wife, has been my nightly routine for months. Maybe years. What's the last show you, your family and your closest friends all actually watched together?

Can Apple innovate? There’s a huge red flag that says they can't, and it’s not from its products. Some background. Back in July, Apple sued OpenAI, claiming former Apple engineers took confidential hardware files to help build OpenAI’s AI device. That’s the device Jony Ive is designing since OpenAI bought his startup io for $6.5B. Last week, OpenAI fired back. It claims an engineer who left Apple for OpenAI, then went back to Apple six months later, sent himself more than 100 files about OpenAI’s secret device on his last day, including device images and manufacturing data, and presented its product plans to Apple’s senior leadership. Two of the most valuable companies on earth are accusing each other of espionage over a device nobody has even seen yet. That tells you how big the prize is. I’ve written many times that killer software needs killer hardware, and whoever builds the device AI lives in takes the consumer crown Apple has held since the iPhone. Apple knows it, and so does OpenAI. But this is where the sign of a lack of innovation comes in. Jobs famously said that Apple has "always been shameless about stealing great ideas", quoting Picasso "Good artists copy, great artists steal". The Mac’s interface was born after a visit to Xerox PARC. Xerox, the behemoth, couldn’t find it within itself to turn it into a product and launch it. Apple, the up-and-coming underdog with a pirate flag flying over its building, moved fast and changed computers forever. Now the tables have turned. Apple is having a ridiculously hard time launching innovative products and moving fast, and OpenAI is breathing down its neck. Having nailed the software with ChatGPT over Siri, it “just” has to nail the hardware. And when an underdog moves fast and the market leader can’t seem to get its act together on innovation, the leader does what market leaders always do. It calls its lawyers. That’s the red flag. The company that grew up under a pirate flag is now the one guarding the gates, fighting in court over a category it hasn’t built yet. When Tim Cook handed Apple to John Ternus, I wrote that leading an underdog and leading a market leader are two completely different jobs. This month Ternus launched HomeView, a well-built device answering a question the market asked - ten years ago... And in the same month, Apple’s biggest AI headline came from a courtroom. There might be truth to the lawsuit, who knows. But history has taught us that courtrooms don’t win markets. Products do. So which way is Apple heading - toward groundbreaking, category-defining products, or toward gatekeeping and courtroom drama...?

You were surprised by @elonmusk's flying car? This is what he’s working on next. It's just as crazy. Back in 2017, SpaceX released an animation that looked like pure science fiction: a rocket taking passengers from London to New York in 29 minutes. Thinks about flying from Tel Aviv to anywhere in under an hour. Any city on Earth to any other city in under one hour. Most people watched it, smiled, and filed it under "Elon being Elon". It looked about as real as an AI video does today. Then in November 2024, someone posted that clip again, and Elon replied with four words: "This is now possible". After this weekend, when we all saw a Roadster drive off a ramp, flip and land on thrusters, exactly as Elon promised in 2019, when everyone laughed - I went back to these statements about suborbital point-to-point travel. The thing is, when I showed the Roadster video to people over the weekend, every single one said "that’s AI". That’s the world we live in now. AI has made the fake look so real that we’ve started doubting the real. But Elon keeps proving the opposite: what looks fake today is often just the real thing a few years early. The 2017 animation was a “fake video” too. It was a vision rendered on a computer, waiting for the engineering to catch up. Today, anyone can create a video like that with AI in minutes. Every crazy idea can look real on a screen. And maybe creating "fake videos" isn't only the danger of AI, but also its gift: we can finally see the impossible before we build it. Judging by Elon repeating the concept of a 30min flight from London to NYC different times over the years, he's probably working on it and well see this as reality very soon. Bending reality into animation with AI, or turning animations into reality with engineering - I wouldn't bet against Elon. Would you?xitter.cf/elonmusk/status/185421…

Sir, this is not the flex you think it is.

Can’t wait to hear the figures - if Apple ever releases them. I doubt they had a real significant amount of iPhone Duo even available.

"The new Roadster will actually do something like this" - @elonmusk Jan 9, 2019. How many people laughed and were skeptical back then? xitter.cf/elonmusk/status/108309…

I haven't felt this inspired and excited about technology for a very long time. I've already showed the video of the Roadster going off the ramp to a dozen people - and they all reacted with WHAT?? Is this real?? No way, it's AI! - every. single. one. I had that exact same feeling when I watched it the frist time. For a few seconds I felt something I’ve only felt a handful of times in my life - but why I love technology so much, and why I chose entrepreneurship as a way of life and building tech. It wasn’t about the specs, or whether a Roadster really “flies” if it only stays up for a few seconds. It was the feeling of watching something that was impossible a second ago become real in front of your eyes. Like magic - but real. Rarely do we get that feeling. But when it does, we remember it forever. It’s the feeling people had in 1969, watching a man step onto the moon. It’s the feeling people had when Steve Jobs showed the world a computer with icons and a mouse, and suddenly anyone could use one. It’s the feeling the iPhone reveal gave us in 2007, when Jobs scrolled a list with his finger with multitouch on the screen and the whole room gasped. It's the feeling I felt seeing the ZUtA Labs robot drive and move for the first time. It’s the feeling of typing your first question into ChatGPT and watching it answer like a person. And now a flying car. These moments are extremely rare. Most of technology moves in small steps, a little faster, a little cheaper, a little better. We get used to it so quickly that we forget to be amazed. Then every few years something breaks through, and for a moment we’re all kids again - witnessing this magic with "oh my God, it worked"! Do you know that feeling Im talking about? What was your reaction the first time you saw this car go off the ramp?

Showing the new Roadster's clip to people this morning - without telling them what they're about to see - and waiting for their reaction. Folks, this is the closest we'd get to getting a "moon landing" reaction in this generation.

Suddenly the conversation about merging Tesla and SpaceXai is making sense 😂

"We wanted flying cars, instead we got 140 characters" - Peter Thiel Now his close friend is bringing us both! @elonmusk 🤯

Amazing Elie!

Looks like @elonmusk just declared a new battle with @JeffBezos. It's all about agentic commerce. The past few weeks Muse, Grok Bot, Instinct and other personal assistants have taken the headlines, and one of the main use cases that keeps on coming up is letting it do your online shopping. What we call agentic commerce. And for some, it's driving much more sales - Mark Zuckerberg himself said that Muse has been ordering for him products from Instacart. A marketplace he's only used twice in the past and he'd never check on his own. However, on the other hand, there's one major company out there that isn't happy about agentic commerce. Amazon has been blocking those agents, and as of last night - it blocked Grok Bot too. That was expected. What wasn't, is Elon's reply about an hour ago when he posted: “Maybe we should start a competitor to Amazon 🤔”. Yes, Elon likes to post stuff and say things with a "maybe" and a wink. But he also knows how to choose his battles. If you haven't been following, Musk and Bezos have been at it for quite some time. In 2021 NASA picked SpaceX to build its moon lander. Bezos’s Blue Origin protested, then sued, and lost. Musk joked that Bezos had retired to become a full-time lawyer. Since then they’ve been fighting over who puts internet satellites in the sky, Starlink against Amazon’s Kuiper. But this battle isn't about "random bots". This might be mcuh bigger - and effect us all. A few days ago I wrote that AI is taking apart the internet, and that the website is becoming a data feed for someone else’s interface. Amazon is THE marketplace of the internet. Bezos is the one that taught the world of ecommerce how to work and perfect the customer cycle. And now, Amazon is the first giant to say out loud: we're not going to just be a data feed and work with your AI. I can understand why. Amazon’s storefront isn’t just where you buy. It’s where Amazon sells ads and learns about you. An AI agent walks in, grabs the product and walks out. An AI agent never sees a sponsored listing - this is a huge hit to Amazon's bottom line - ads. Amazon might be doing this because they're working on their own AI personal assistant - they were the first with the Alexa device - but the irony is that Amazon became Amazon by going around the bookstores. Twenty-five years later, Amazon is now the bookstore, standing at the door telling the new way of shopping it can’t come in. And history isn’t on the side of the store that locks the door.

Honestly had the same feeling. I love Apple's ads. Studied them for years. This one just rubbed me the wrong way so many times.

Apple just introduced its first AI-native device - it looks familiar. And that's the problem. Today Apple unveiled HomeView, its first new product category under John Ternus. As always with Apple, they're VERY late to the game. “IoT” and the smart home were the hottest terms in tech over a decade ago. Amazon put Alexa in our kitchens in 2014 and gave it a screen with the Echo Show in 2017. Google followed a year later. We're almost a decade later, and Apple finally have a device with a screen, that sits on your kitchen counter. But there’s one thing Apple did differently. In the very first sentence, Apple introduces this device as “Built for Siri AI”. It's clear what Apple are going for - for Apple, in a very Apple like way, this is not a speaker with a screen. It's not a smart home hub to control your AC. In a time everything is about AI and everyone is chasing the next AI device - Apple is trying to say that this device is built around AI. The thing is, as I wrote last week, with AI taking over HOW we interact with our devices, we’re about to enter a new generation of interfaces. The graphical interface with apps and icons that Jobs brought us in 1984 doesn’t fit a world of devices running on AI. Yet HomeView still greets each person with their own Home Screen with familiar apps: Safari, Music, Photos. Call and hang up buttons - you know, the familiar Apple interface feeling. And that's the "problem" here. This device was reportedly in the works for years, and was delayed again and again. When reports first started coming out about this device, it was well before ChatGPT, Claude and of course Muse took over our lives. AI was added to the device, the device was not built around it. It’ll be interesting to see how it plays out. Maybe the kitchen is the right place for Apple to learn quietly, while the iPhone stays the agent’s real home. It’s still TBD. A few weeks ago I wrote that Apple swapped an operator CEO for a hardware engineer at exactly the moment the world needs new hardware for AI. HomeView is Ternus’s first answer. It seems like a well-built and well-thought-out answer, but to a question the market asked ten years ago. Which brings back the one big question mark hovering over Apple for the past decade - can Apple really innovate again?

Love this! If you're building in hardware - join the group!

OpenAI is taking apart the internet and most people around you aren't even noticing. Let me explain. On Wednesday OpenAI rolled out GPT‑6 to 1.2 billion people, and almost every headline was about the model. But when a different company launches an improved model every few weeks, what caught my eye this time was a feature they called “Intelligent UI.” ChatGPT no longer just answers in text - it builds an entire interface itself. To understand how big of a change this is - think of it as moving from a calculator, or a watch, that just shows numbers, to a full computer display. I mentioned this in a post last week, how we're witnessing a change to how we interact with computers, as big as when Steve Jobs introduced the mouse and app icons back in 1983-4 when he showed the world the graphic user interface. Think about what OpenAI introduced means for every business online. Every website you know. For 30 years the website was the front door to your business. You digital storefront. That's where people would land - and where every single advertisement or social media post - would send the user to in order to turn them into leads. Now the customer asks ChatGPT - and ChatGPT doesnt answer them with text and a link anymore, rahter it actually builds the answer, in its own design, using the products it finds and recommends. As a business, your products show up in cards in ChatGPT, designed by ChatGPT, in a layout that does not care about your business, your brand or your storefront. That digital presence you spent so much in creating to express your product and brand - isn't relevant anymore. Or look at it differently, the website is becoming a data provider for someone else’s interface and design. There are so many ramifications to this change - but to point one out, I'll just emphasize a simple known fact: the internet is "free" because businesses pay for it, essentially through buying ads. Every ads' purpose is to drive more leads to the business' website where a lead would turn into a customer. We can assume the internet will continue to be free - but where will businesses see the ROI on the money they're spending to make it free? In other words, if the customer stops visiting the business' website, where does a brand still get their full attention, in its own design, with its own voice? Where can a business express their brand digitally?

“We dOn’t WaNT tO fIgHt IsRAel’S WaRs!” A. Israel fights its own wars B. Israel, time & again, is saving your troops you dumb idiots

Incredibly disturbing history is repeating. In the 1930s, Nazi propaganda machine used radio and sound film- new technologies - to turn old resentments into hating Jews. Tech so new, people weren't aware they were being manipulated. Tiktok is the anti-west propaganda machine.

במהלך חול המועד ישבנו במשפחה המורחבת וזרקנו הימורים על הפוליטיקה הישראלית, ,בחירות וכו׳. אמרתי - רוצים הימור ממש מטורף? הימור מארץ ההימורים? אם בצלאל סמוטריץ׳ ישפר את האנגלית לרמה נורמאלית (רמה של ישראלים כן? לא צריך ישראלי-אמריקאי כמו ביבי) - הוא יהיה ראש ממשלה, ומחליף ראוי ליום שאחרי @netanyahu. אני יודע ש-X זה לא המקום להגיד את זה, אבל הייתי צריך להוציא 😆