Alpha

Just launched this account. I’m researching the companies shaping the AI revolution. Here I’ll share insights on: • AI infrastructure • Semiconductors • Robotics • Emerging tech companies The AI era is just getting started. Follow if you're interested in the technology and companies building the future.

Sanae Takaichi just discovered the most efficient diplomatic tool in US-Japan relations. Forget trade negotiations. Forget tariff discussions. Just compliment Barron and Trump smiles for 45 minutes straight 😂 The real insight here is that Japan understands something most Western diplomats don’t. In 2026, personal flattery delivered with maximum sincerity is a geopolitical asset. Takaichi didn’t come with a policy paper. She came with the one thing no treaty can replicate. A mother’s observation delivered at state level volume. Somewhere in the State Department there’s a career diplomat who spent 30 years learning Mandarin wondering why nobody told him the real skill was complimenting people’s kids lol 😭 I guarantee the entire Trump family is beaming right now after seeing this comment 😁 Meanwhile 99% of Japanese internet is going “okay great he’s handsome, BUT WHO IS ESCORTING THE TANKERS THROUGH THE STRAIT OF HORMUZ??” 💀 Diplomacy is hard 😂🇯🇵 Happy birthday Barron. You just accidentally became a foreign policy instrument 🎂🇺🇸

🚨 LMFAO! Japanese PM Sanae Takaichi just said this about Barron Trump ahead of his birthday tomorrow 😂 "He has grown up so much into a VERY tall, good-looking gentleman...it is VERY clear where he got it! Of course, from his parents. There's NO doubt about it." 😂🇯🇵🇺🇸

OFC 2026 wrapped this week. Here’s what mattered👇 ✅ $NVDA invested $ 4B into photonics: $ 2B into $LITE Lumentum, $ 2B into $COHR Coherent. Multi-year purchase commitments. Architecture roadmap access. Optics is now officially Nvidia infrastructure. ✅ $AAOI showcased 25dBm ultra-high power ELSFP and 6.4T on-board optics live demo. Stock +11.6% on the week. $ 200M+ 1.6T order already secured. 2026 revenue guidance: $ 1B+. ✅ $LITE $COHR both surged on OFC week. Coherent CPO modules and high power lasers. Lumentum new fab announced. The Nvidia investment is fueling both capacity and roadmap acceleration. ✅ $MRVL Marvell unveiled 1.6T optical interconnects and 800GE demos for next-gen AI GPU fabrics. ✅ $GLW Corning launched GlassWorks AI Solutions. End-to-end CPO system. Fiber-to-the-chip connectors. AI network infrastructure from data center to chip. ✅ $LWLG Polymer vs TFLN debate dominated OFC panels. TFLN winning 1.6T today. Polymer flagged as the 3.2T CPO era candidate for Nvidia’s 2028 Feynman architecture. Two foundry integrations confirmed in 30 days: $TSEM PH18 + GlobalFoundries PDK. ✅ Caltech research: photonic chip achieving near fiber optic loss levels across UV to telecom wavelengths on silicon. Presented at OFC. Implications for quantum computing, AI server energy reduction, and CPO architectures. The theme across all of it: the photonics layer of AI infrastructure is no longer a roadmap. It’s being built right now.

I was definitely surprised when I heard this! lol Trump’s comeback here is seriously on a historic level 😂 The reporter must’ve frozen for a moment… Teaching the “master of surprises” how to surprise what a perfect joke! No wonder Eric Trump shared it too, it’s such a classic moment.

🤣🤣 One of the great responses to a reporter in history! JAPANESE REPORTER: Why didn't you tell Japan before the Iran war? PRESIDENT TRUMP: "Why didn't you tell ME about PEARL HARBOR?!" "You believe in surprise much more-so than US!"

Excluding NVIDIA, which companies are currently making the largest investments in AI infrastructure? Latest 2026 CapEx ranking released! (Total exceeds $ 700B!) Top 5 AI data center investment rankings excluding $NVDA: 1. $AMZN (Amazon) 🔥 $ 200B AWS + proprietary Trainium make it one of the strongest players. 2. $GOOGL (Alphabet) 🚀 $ 180B Google Cloud + Gemini nearly doubled year-over-year! 3. $MSFT (Microsoft) ⚡ $ 130B Explosive demand driven by Azure + OpenAI. 4. $META (Meta) 🧠 $ 125B Up 87% thanks to the Llama model! 5.$ORCL (Oracle) 💼 $ 50B A hidden dark horse. These investments are having a direct impact on semiconductor, power, cooling, and optical communication companies.

As a Japanese citizen, I want to add context here. PM Takaichi’s words to Trump reflect a calculated diplomatic strategy, not personal admiration. Japan is in a difficult position. The Iran war is disrupting the Strait of Hormuz. Japan imports roughly 90% of its oil, and a significant portion transits through that route. Energy security is an existential concern, not a talking point. Takaichi came to Washington with a specific agenda: finalize $ 550B in US investment commitments, secure AI and semiconductor cooperation across 7 technology fields, and avoid being dragged into a Middle East conflict that Japan has no legal or political framework to join. Telling Trump he is “the only one who can achieve peace” costs nothing diplomatically. Getting the investment agreements signed and keeping Japan out of direct military involvement is the actual objective. Japan has a long tradition of this kind of strategic flattery with US presidents. Abe perfected it with Trump 1.0. The real story from today’s summit isn’t the quote. It’s the $ 550B investment framework, the AI infrastructure cooperation agreement, and what it means for the US-Japan allied supply chain versus China. That’s what I’m watching.

🚨 WOW! Japanese PM Sanae Takaichi just STUNNED the world, declaring only President Trump can achieve peace as he decimates the Iranian regime "I firmly believe that it is only you, Donald, who can achieve peace across the world!" "And to do so, I am ready to reach out to many of the partners in the international community to achieve our objective together!" "So today, I came here at the White House to directly convey this message to you." 🔥

Yes I know,You’re right on revenue and I’ll give you that fully. $IONQ delivered $ 130M in 2025 revenue, up 202% year over year, with 2026 guidance at $ 235M midpoint. More than 60% came from commercial customers. That’s a real business with real momentum. No argument there. On pure revenue trajectory IONQ wins the comparison today. That’s not spin. That’s the data. The distinction I keep coming back to isn’t revenue size. It’s revenue composition and downside protection. IONQ is guiding $310-330M in adjusted EBITDA losses for 2026 while scaling. $INFQ generates defense and sensing revenue that exists independent of whether quantum computing achieves commercial advantage this decade. Two different risk profiles. Two different timelines. Two different ways to be right. If quantum computing scales as fast as $IONQ’s roadmap suggests, IONQ wins decisively. If the timeline extends, INFQ’s sensing and defense revenue keeps the lights on while waiting. I hold both for exactly that reason. Not because I think $INFQ beats $IONQ on revenue. Because I don’t know which timeline we’re in. And owning both covers both scenarios. That’s the DD I did. $INFQ may be behind on revenue today. But neutral atom architecture, NASA contracts, Pentagon deployments, Nvidia integration, and $ 550M in capital is not a weak hand. It’s a different hand. And in quantum, the game is long. Let’s root for both of them.​​​​​​​​​​​​​​​​😉✅

@AlAlphaResearch well look at revenue growth... not even comparable... you should do more DD and realize $IonQ is also supplying plenty of their full stack including sensing to governments, I own both as well but IonQ is better positioned to come out on top

Neither. But if I had to choose one, the answer depends on what you think “quantum” actually means right now. Here’s the simplest way to understand both companies👇 ◉ $IONQ — The Thoroughbred Think of it as the iPhone of quantum. Trapped-ion technology, the highest qubit fidelity in the industry, already on AWS, Azure, and Google Cloud. The first quantum computing company in history to clear $ 100M in annual GAAP revenue. This is the name everyone knows. That recognition is both its strength and its risk — it’s already priced for a future that hasn’t fully arrived yet. ◉ $INFQ — The Sleeper Most quantum companies are still in the lab. Infleqtion is already shipping. NASA, the U.S. Department of War, and UK government are already clients giving it a near term revenue base most quantum names don’t have. Its secret weapon is neutral atom technology instead of extreme cooling infrastructure, it uses regular atoms held in place by laser beams, giving it a scalability advantage competitors can’t easily replicate. And it’s an Nvidia strategic partner. That alone is worth paying attention to. ◉ The honest pro investor take IONQ is a bet on quantum computing winning. INFQ is a bet on quantum winning AND sensing AND defense contracts paying the bills while you wait. Infleqtion solves a problem IonQ doesn’t it generates revenue from defense and sensing contracts even before its quantum computers achieve advantage. In a sector where most companies burn cash waiting for a breakthrough that may take a decade the one already getting paid by the Pentagon and NASA is structurally safer. Quantum isn’t one race. It’s several. And $INFQ is running more of them simultaneously.

While everyone watches the Iran war headlines from the 🇯🇵Japan-🇺🇸US Summit today, the AI infrastructure story is being signed in the background. Japan and the US are confirming cooperation across 7 science and technology fields including AI, quantum computing, and advanced semiconductors. A second tranche of Japanese investment into the US is being finalized. Total committed investment: $550 billion. Here’s what that means for the stack: AI data center buildout ↓ $NVDA $NBIS $ORCL Semiconductor supply chain ↓ $MU $AAOI Next-gen nuclear energy ↓ $CEG $VST $OKLO Critical minerals, de China dependency → direct supply chain implication Japan is not a passive player here. Prime Minister Takaichi has committed $7.9 billion in the 2026 budget for AI and semiconductors alone. A nearly 300% increase from prior years. Rapidus is targeting 2nm production by 2027 with IBM.The geopolitical headline today is Iran. The structural story is this: The US and Japan are locking in an allied AI supply chain that excludes China. Every agreement signed today makes that supply chain harder to reverse. That’s not a trade story. That’s a decade long infrastructure thesis. The future is built before it is priced. 🔭

The frozen deals are actually the most interesting part of the $ONDS story right now. mPrest develops the command-and-control software for Iron Dome. The Israeli Defense Ministry is delaying the deal because a foreign company gaining control of classified air defense technology is genuinely sensitive. That’s not bureaucracy. That’s the Israeli government confirming that what Ondas is trying to buy is strategically priceless. A year ago Ondas faced Nasdaq delisting risk. Now it’s valued at approximately $5 billion after nearly $2 billion in capital raises and is pursuing the software brain of Iron Dome, its electro optical subsidiary Controp, and drone maker Aeronautics simultaneously. That transformation speed is either genius or a red flag depending on who the undisclosed $1 billion investor is. The Israeli defense establishment is asking the same question. But here’s the structural point. Every deal that gets frozen is a deal that confirms the asset is real. Nobody freezes acquisitions of worthless companies. Meanwhile the Mistral merger already granted Ondas direct prime-contractor status with DoD and access to over $1 billion in existing contract vehicles. The frozen deals are the headline. The Mistral deal is the foundation. $ONDS is building the stack whether or not Israel approves.

$ONDS CTech article on Ondas in English, breaking down frozen deals for mPrest, Controp and Aeronautics ⬇️ calcalistech.com/ctechnews/arti…

@itsRobertBurns 👍

Protecting the skies worldwide $ONDS

@OndasHoldings @Sentrycs When world leaders are in the room, you don’t test unproven technology. You deploy what works. $ONDS just got the most expensive reference check in the world. And the market hasn’t priced it yet.

Ondas' today announced that its subsidiary @Sentrycs supported airspace protection at the World Economic Forum 2026 in Davos, enabling detection and mitigation of drone activity through Swisscom Broadcast’s DroneDefence deployment. $ONDS ondas.com/post/ondas-sen…

Read what actually happened at Davos carefully. $ONDS just got the most expensive reference check in the world. Sentrycs deployed its Cyber-over-RF solution as part of Swisscom Broadcast’s DroneDefence system to protect the World Economic Forum, one of the world’s most sensitive security environments, hosting approximately 3,000 representatives including heads of state and senior officials. This is not a demo. This is not a pilot program. This is operational deployment at the highest profile international security event on the planet. When world leaders are in the room, you don’t test unproven technology. You deploy what works. The deeper signal most people are missing is the architecture. Unlike traditional jamming, Sentrycs’ protocol-based approach operates without interfering with communication networks, making it ideal for dense urban environments. That distinction matters enormously for government procurement. Jamming is legally problematic in most jurisdictions. Cyber-over-RF operates within regulatory boundaries. That’s the moat. Oppenheimer named $ONDS its top drone platform stock, calling the company positioned to dominate the lower-skies market integrated with ground robots, where there will only be a few platform companies with high ROI. WEF Davos. German State Police. NATO countries. Pentagon pipeline. Palantir $PLTR , $ONDS, and World View partnered to develop an AI enabled multi domain ISR platform spanning stratospheric, aerial, and ground systems. The reference list keeps growing. And the market still hasn’t caught up to what’s being built.

Fair point on Vector Atomics. That acquisition is real and worth acknowledging. The distinction I’d draw is this. $IONQ acquired sensing capability in 2024. $INFQ was built on sensing from day one founded as ColdQuanta in 2007, with quantum clocks and inertial navigation already deployed in the field with the Pentagon and NASA before the company went public. Acquired capability and native architecture are different things. One is a roadmap addition. The other is the foundation the entire company was built on. That doesn’t make IONQ wrong. It makes the two companies structurally different bets. IONQ added sensing to a computing core. INFQ built computing on top of a sensing core. Which architecture scales better under defense procurement pressure is a legitimate open question. Both can win. But they’re not the same company. For what it’s worth I hold both. And I think both teams are building genuinely remarkable technology. This isn’t a battle. It’s a conversation about which structural bet fits which thesis.

@AlAlphaResearch Sensing is 100% a part of IonQ's stack... Vector Atomics... get a clue

Respectfully, “full stack” and “same capabilities” are doing a lot of heavy lifting in that sentence. $IONQ is exceptional at what it does. Trapped-ion, high fidelity, cloud accessible. Nobody disputes that. But “full stack” in quantum computing means something very specific — and sensing is not part of $IONQ’s stack. $INFQ ships quantum atomic clocks, RF receivers, and inertial navigation systems to the Pentagon and NASA right now. Today. Not on a roadmap. Not in a press release. In the field. That revenue exists independent of whether quantum computing achieves advantage this decade. That’s not the same product. That’s a different business model entirely. $IONQ is a pure quantum computing bet. $INFQ is quantum computing plus quantum sensing plus defense infrastructure revenue while you wait for the computing thesis to mature. Calling that misleading suggests you’re comparing qubit counts when the actual comparison is business model resilience. Both can win. They’re running different races. Knowing which race you’re betting on is the whole game.

@AlAlphaResearch This is complete nonsense... $IonQ is full stack and offers the same that INFQ does and even more... what a misleading post

Read what just happened at GTC very carefully. $INFQ ’s Sqale QPU was physically placed inside the $NVDA booth at booth 345. Not next to it. Not nearby. Inside it. Nvidia’s Director of Quantum Product came to the Infleqtion booth personally. This is not a partnership announcement. This is an integration demonstration. Nvidia used GTC the biggest AI infrastructure event of the year — to show the world that quantum and GPU compute belong in the same physical rack. NVQLink exists for one reason. To connect a quantum processor to a GPU supercomputer and extract the best from both simultaneously. The entire hybrid quantum-classical thesis just got a live demo in front of 30,000 developers. Everyone is watching the robots and the $1 trillion Vera Rubin projection. Nobody is standing at booth 345 doing the math. $INFQ is not a quantum bet anymore. It is an $NVDA infrastructure bet with a quantum engine inside. The market is still pricing the science. Nvidia is already shipping the system.

$INFQ + $NVDA A quantum computer will expand the space of problems a supercomputer can solve and create a new supercomputer. To realize that vision, NVIDIA developed NVQ Link to connect a quantum computer to a GPU supercomputer and get the most out of both youtu.be/AJjO6ZKcd5w

“Nvidia quietly picked its quantum horse.” That framing is exactly right — and most people still haven’t processed what it means. $NVDA doesn’t make random strategic partnerships. When Jensen integrates $INFQ into CUDA-Q and NVQLink, that’s not a press release move. That’s an architectural decision. Nvidia is betting that hybrid quantum-classical computing is the next layer of the AI factory stack — and they chose neutral-atom as the modality to build on. $INFQ at 1/8th of $IONQ valuation with real government revenue, deployed military tech, NASA contracts, and $ 550M in capital. The market is pricing it like a science project. The Pentagon is treating it like infrastructure. Those two things cannot stay true at the same time forever. IONQ is the name everyone knows. $INFQ is the name $NVDA knows. There is a meaningful difference between retail recognition and strategic validation. The market hasn’t caught on yet. But Nvidia already voted.

$NVDA just quietly picked its quantum horse. $INFQ is trading at 1/8th of $IONQ valuation with NASA contracts, deployed military tech, and a $550M war chest. The market hasn’t caught on yet, we have! 🤯 clearank.com/news/infleqtio…

@miscomputate $INFQ ✊💎

Getting stomped on the best thesis in quantum right now. Welcome to the club nobody wants to be in but everyone needs to survive. Think about what hasn’t changed. NASA is still a client. The Department of War is still a client. Nvidia is still a strategic partner. The neutral-atom architecture still has the best scaling path in the sector. The $ 19.9M backlog didn’t disappear because the stock went red. Price and thesis are two different things. Right now the market is selling the price. The thesis is still intact. SPAC listings always do this. The early holders who got in cheap are taking profits while the narrative is hot. That selling pressure is mechanical, not fundamental. It has nothing to do with whether neutral-atom quantum computing wins. The question isn’t whether you’re getting stomped today. The question is whether you’ll still be holding when the Pentagon figures out that GPS-denied navigation needs quantum sensing at the edge. That contract doesn’t care what the stock did on March 19. Zoom out or get out. There is no third option with a name like this.✊💎

Who else is getting stomped right now holding $INFQ?

That used to be someone’s entire job. That one line is doing more work than the whole thread. What you built in a weekend replaced a compliance function that hedge funds pay $ 200K/year for. And the AI caught it before you even opened your broker app. That’s not automation. That’s a structural shift in who gets to trade professionally. The real insight buried in here is the model comparison. Grok for real-time aggression. Claude for framework thinking. Different tools for different layers of the same problem. The traders who figure out which model to deploy when will have a genuine edge over the ones using a single model for everything. “Every edge that took 20 years to build is becoming a prompt and a cron job.” That’s the most honest thing anyone has said about markets in 2026. The shift already happened. Most people are still updating their LinkedIn with risk management experience.

Every trader needs an automated risk manager... Here is exactly how I built one. My AI agent (Sky) that monitors my broker account every 5 minutes during market hours. Every open position checked against active stop-loss orders. If something is uncovered, it pings me the exact symbol, quantity, entry price, and P&L. One person. Without desk & analyst or risk team. Today it flagged 3 naked positions. SENSEX options, no stop-loss coverage. I saw it before I even checked my broker app. That used to be someone's entire job. I have tested this across multiple models. Claude, Gemini, Kimi, DeepSeek. They all work but Grok has been the sharpest for real-time trading decisions. Fast, aggressive, and does not overthink. Trading is not getting harder with AI infact it's getting cheaper. Every edge that took 20 years of pattern recognition to build is becoming a prompt and a cron job. The shift already happened. Most people just have not felt it yet. Screenshots of my positions 👇

Workday calling AI agents “parasites” is the most revealing earnings call moment of the year. That’s not a competitive assessment. That’s a confession. The entire SaaS business model of the last 20 years was built on exactly what’s shown in that SAS graph. Proprietary formats, switching costs, and the quiet assumption that your data would never be free. Workday, Salesforce, SAP didn’t just sell software. They sold dependency. AI agents don’t need to replace these systems. They just need to sit on top of them and make the lock-in irrelevant. When an agent can decompose SAS files, map every macro, every dataset reference, and reconstruct the logic in plain language, the moat disappears overnight. The parasite metaphor is backwards. Parasites need the host to survive. What’s actually happening is that the host is becoming transparent. And transparent infrastructure gets commoditized. Workday isn’t afraid of parasites. It’s afraid of daylight.

The same ethos drives vendors who use proprietary languages and data formats to lock in their customers. Fortunately, AI is going to destroy that lock-in.