Daniel Bogda

Wow this is a MASSIVE upgrade to the Black Hawk. In a hugely ambitious feat @skyryse has completely reworked flight controls for this military aviation workhorse.

This morning at VERTICON, we provided a firsthand look at how SkyOS™, the world’s first operating system for flight, is maturing its capabilities on the Black Hawk platform. In a session led by former commander of the 160th Special Operations Aviation Regiment (aka the “Night Stalkers”) and our Director of Strategic Pursuits, Mike Hertzendorf, we shared how our team was able to rapidly expand our flight envelope for Black Hawk with SkyOS, to perform precision maneuvers and patterns with one hand. Just weeks after achieving the first-ever automated pickup, hover, and setdown of a Black Hawk with SkyOS, we’re continuing to witness our system’s maturity on a new platform. SkyOS unlocks safety and performance without compromise – making flying every aircraft more reliable and effective. Watch as this battle-proven aircraft enters a new era.

My second interview with @soren_ma, Co-Founder and CEO of @neros_tech. Neros just acquired a new 250k square foot factory which will allow them to ramp production to 1M drones a year. 0:50 New 250k square foot factory 5:13 How the Russia Ukraine war has evolved 8:51 Learning from the Starlink production ramp 12:37 Building a military spec drone 15:25 Prototypes vs production product 17:27 Creating a fully China-free drone 20:27 Scaling production capacity 25:53 Production hell vs supply chain hell 29:11 Timelines and urgency 34:08 Flyoff mode and PBAS 37:40 How Soren operates during a sprint 40:54 Shifting from building drones to company building 42:37 Leadership hiring 49:05 Focusing on execution 50:30 Defeating drone jammers 57:55 Oh sh*t moments 1:00:09 Refocusing on health and taking care of yourself 1:05:21 Measuring effectiveness by battle field results 1:13:04 Evolution of their testing process 1:16:05 Archer Fiber 1:17:55 Lifting up the entire drone industrial base 1:21:12 Building region-specific supply chains 1:23:49 How America can win over the coming decades 1:26:55 Are we going fast enough? 1:29:12 Choosing to work on lethal systems 1:32:35 Being pro-America 1:35:37 Future of warfare and autonomous systems

Most people argue vertical integration is the answer (Vertically Integrate or Die), but the truth is that is most often a function of bad suppliers You have to do it yourself if you can’t rely on others There’s a growing contingent of ADCU ‘supplier startups’ that will help change that

heavy endorse (obviously)

🚨 NEW: Small manufacturers will soon be eligible for our new 90% Made in America Loan Guarantee. It's another way we’re driving capital to reshore U.S. supply chains – to help manufacturers expand facilities, hire workers, increase production, and rebuild American industrial dominance. sba.gov/article/2026/0…

THIS GUY FOUND WHY YOUR CLAUDE CODE COSTS ARE 10-20X HIGHER THAN THEY SHOULD BE he reverse engineered the entire Claude Code binary. found two hidden bugs that silently break your prompt cache. BUG 1: if your conversation mentions billing, tokens, or Claude Code internals, the cache breaks on every single request. there's a hidden string replacement baked into the binary that targets billing attribution codes. if that string appears anywhere in your conversation history, it replaces the wrong one and your entire cache rebuilds from scratch. you get charged full price instead of cached price. fix: run Claude Code via npx instead of the standalone binary. BUG 2: every time you use --resume, your entire conversation cache rebuilds from scratch. one resume on a large conversation costs $0.15 that should cost near zero. the system injects tool attachments in a different position on resume vs fresh sessions. this changes the cache prefix and forces a full rebuild. every single resume costs you a one-time hit on your entire context. fix: no clean fix yet, as it was introduced in v2.1.69. downgrading to v2.1.30 works but you lose months of features. on a 500K token conversation these two bugs combined can cost you $0.20+ per request. if your usage has been burning way faster than expected, this is probably why.

@om_patel5 the irony of a tool that costs you more money every time you talk about how much money it costs you is genuinely poetic anthropic accidentally invented the first software that punishes self-awareness

My friend makes over $1M/year selling pokemon cards He buys packs for $10, opens them, and sells the rare cards inside for $20, making $10 profit per pack He does this every minute for 8 hours a day $10 × 480 min = $4,800/day $4,800 × 365 = $1,752,000/year What's your excuse?

We’ve reached a point where 9 out of 10 people I speak with think it’s all over and we’re going much lower I’ve never seen market fears this high before Just based on this alone it makes me feels like we’re closer to a bottom than most people think I’m going to bid some good coins/companies this week and keep cash in case we break lower Let's see how this plays out (nfa)

When you realize that paying 30% income tax means you basically work 100 days a year for free

FILMING OUTSIDE A PRISON… AND IT GETS TENSE FAST A man recording outside a correctional facility gets confronted by officers… He says he’s not doing anything wrong - no signs, no warnings, no trespassing Claims he was just checking out the new electric patrol cars and the area Officers clearly weren’t feeling it… Now it turns into a back-and-forth over rights vs security Do you think you should be allowed to film outside a correctional facility… or is that crossing the line? ⬇️ 🇺🇸

build a home lab. not for aesthetics. not for content. for capability. → a table becomes a workstation → a room becomes a systems environment → your house becomes a prototype factory start minimal: → power supply → multimeter → soldering iron → breadboards + wires → a microcontroller (arduino / esp32) this is enough to: → read signals → control hardware → break things and understand why then layer complexity: → sensors (imu, gps, cameras) → motors (dc, servo, stepper) → embedded compute (raspberry pi) → fabrication (3d printer, cnc) now you are not learning. you are building systems. home lab changes the feedback loop: idea → build → fail → debug → rebuild no permission. no waiting. no abstraction. software alone lies to you. hardware forces truth. you will: → miswire circuits → burn components → chase noise in signals → debug at 3am good. this is where intuition forms. over time: → your lab becomes organized chaos → parts turn into modules → modules turn into systems → systems turn into products this is how engineers are made. not courses. not certificates. iterations in a room full of tools.

In case you haven’t heard, we put up $3.5M+ to source optimistic visions of the future. We don’t care who you are or where you are. Tell me… what future do you want for your children?

My conversation with Sergey Levine (@svlevine). Sergey is the co-founder of @physical_int -- a company building foundation models that can control any robot to do any task in any environment. The company's thesis is that generality is more scalable than specialization, meaning that a model trained across many different robots and tasks will ultimately outperform any system built to do one thing well (eg, just wash dishes). Sergey is a researcher by background, but I think you will appreciate how practical and commercially grounded this conversation is. We discuss: - Why changing a diaper will be the last task a robot masters - The simulation v. real-world data debate - How multimodal LLMs give robots common sense - Moravec's Paradox + Robot Olympics - Why robots can do long-horizon tasks now - A realistic timeline for robots in our homes I should note that I am an investor in Physical Intelligence -- I made the investment because I believe it is one of the most important companies tackling the problem of robotics. Enjoy! Timestamps: 0:00 Intro 2:39 Defining Physical Intelligence 5:19 The Challenge of Building General Models 6:34 The Stakes and Future of General Purpose Robotics 8:15 Pros and Cons of Humanoid Robots 10:12 Historical Milestones in Robotics Research 15:31 Combining Generative AI and Deep RL 21:24 Moravec's Paradox 25:33 Kitchen Robots 29:30 Simulation vs. Real-World Data 30:48 The Robot Olympics 36:31 The Physiological Reality of Embodiment 38:56 Controversies in the Robotics Community 44:18 What Makes a Great Researcher 48:27 How Businesses Should Prepare for Robotics 54:09 Tracking Progress Through Research Papers 57:02 The Next Step: Mid-Level Reasoning 1:02:00 The Kindest Thing

Today, we're announcing Heaviside, our foundation model for electromagnetism. Trained on tens of millions of designs and over 20 years of proprietary simulation data, Heaviside predicts electromagnetic behavior from geometry in 13ms, which is 800,000x faster than a commercial solver. Heaviside is not a language model, and it’s not a surrogate model. Heaviside marks a new class of foundation model for physics which understands the fundamental relationships between materials, the geometries and the electromagnetic fields they generate. We’re releasing a research preview of Heaviside in Atlas RF Studio, an interactive agentic sandbox where you describe the EM behavior you want and the model generates the physical structure that produces it. @arenaphysica , we believe the implications of this class of model extend well beyond RF, as the frontier of exquisite hardware is electromagnetically-governed: wireless communication, radar, power delivery, high-speed computing, and the interconnects inside every chip on earth. In the months ahead, we’re excited to scale up Heaviside to broader frequency ranges, design spaces, and to support silicon-level designs, and deploy it with our closest partners and collaborators in service of their biggest design challenges. If you’ve read our thesis, this is just Step 2 in our pursuit of electromagnetic superintelligence. Read the full announcement and try Atlas RF Studio…tell us what you think: arenaphysica.com/publications/r…

I just got @Fried_rice from the X Handle Marketplace! Get your own at handles.x.com

vibe coded a fuzzing ai agent last month and let it run for a week using my $200 claude max. it then found 21 high/critical vulnerabilities in Chrome.