Recorded a "Introduction to Phontonics" with @crux_capital_ who was probably one of the earliest Photonics investors on X. Many of his picks are literally up 5-10x and the entire space has seen a crazy rally. In the episode we discuss where we are today and where the space is going.

Just recorded a 1.5h "Memory Masterclass" with @damnang2! Super excited I finally had the chance to talk to him directly, and now I can verify: 1) He's legit and actually works in the industry 2) Super knowledgeable 3) Kind and patient guy who's great at breaking down complex topics! He will remain anonymous (at least for now), so I cannot share his face/voice from the recording, but I will share the full transcript of the interview, cut out all the blabla, uhms, ahs, etc., and share my key takeaways in a post in the near term! Many great nuggets for memory investors $DRAM $MU $SKHY $SNDK Definitely follow him on X and Substack if you haven't already! :)

Recording a podcast with $NBIS' one and only @demian_ai this Monday! We'll talk about Tokenomics, agents, inference and bottlenecks. If you have specific questions you want us to cover, put them in the comments. We will NOT talk about @nebiusai for once. All my thoughts on the company are out and the disappointing price action last few days changes nothing. We will talk about AI and the AI trade in general.

Please take a second to really think about this. I know when someone tells you "you're being manipulated on Social Media" you instinctively have this "duh, obviously" reaction. Just like with Marketing in general, we all think we're immune to it and it's the "other" and the "stupid" people who fall for it. Meanwhile it's super obvious to us when looking at everyone else we know how THEY are influenced by a ton of manipulative marketing, sales or political messages. The conclusion is obviously that ALL OF US are heavily manipulated by millions of manipulative messages we receive every day that are NOT in our best interest. It's really worth pointing that out every once in a while. Just know that on X you're NOT safe either. Fake account "influence operations" are real. I worked in this field for many years and most people, even very very high IQ and chronically online people, have no idea how much they are being manipulated by fake accounts every single day. Always stay a little skeptical about senders when consuming content online and at the very least make sure the poster is a real person.

After $NVDA CRUSHED earnings once again, maybe it's time to think about why the AI bears keep getting it wrong. Short (free) article below. Nothing insanely deep, just some random post gym thoughts and why I keep listening to all the AI bears on podcasts and hear the same wrong arguments over and over again and what I think they are all missing.

I want to do more interviews with the smartest founders, CEOs, execs, investors, FinX creators, and opinion leaders in AI and tech. Who should I interview? Also, no shame in suggesting yourself! You can also ping me via DM :)

X feels noisier than ever. Many investors are deeply confused. Some of the biggest permabulls are suddenly going heavy into the Mag7, while many of the bears I talk to are now adding hypergrowth AI names. Feels like the world is upside down. To be honest if I listen to the reasoning it feels less like intentional, thoughtful diversification or risk management and more like people giving up on having an opinion. I sense that many people are simply exhausted from keeping up with a constantly changing narrative and volatility over the last two years, from trade wars to real wars to financing concerns. I feel this exhaustion strongly myself. I can't go on vacation for a week, or often even relax for a weekend, without waking up to a new policy or macro event that then gets amplified by the media news cycle as if it's the end of the world. But I think it's time people look at facts & data again. The visual below is the Nasdaq over the last 10 years. 2016 — Brexit 2018 — U.S.-China Trade War 2020 — COVID-19 Global Lockdowns 2022 — Russia Invades Ukraine 2022 — 40-Year-High Inflation + Aggressive Fed Hikes 2023 — SVB Collapse / Banking Crisis 2024 — Yen Carry Trade Unwind 2025 — DeepSeek AI Shock 2025 — “Liberation Day” Tariffs / Global Trade War Panic Result: QQQ is up about +545% since the start of 2016. The reality is that most of the time, the market just shrugs it off. Most "bad news" are completely meaningless in the face of massive economic growth that happens more silently but consistently, every single day across the global economy. The 2022 "crash" felt horrible at the time, especially if you worked at an early-stage startup that needed funding. But even that now looks like a tiny hiccup. What you really need to internalize is this: The market is designed to go up. Economic growth + inflation structurally push stock prices higher over time. And you're turning bearish now, at the dawn of AGI? After reading all of this, here's an old Chinese wisdom I want to bestow upon you: If you have a big shlong, you gotta stay long. And if you have tits, buy the dips.

$CRDO is one of the highest quality names in AI infra. If you want to invest in AI infra because of the insane growth, but want to add a name that is already extremely profitable and will print cash for many more years, this is a company for you. The only question around Credo is when and at what price to add. @babyfolio and @jpinsights covered Credo in this new @edelbridgealpha article. Check it out!

Read chapter 3 "Agency" of my article "AI Is Eating the World" article and tell me this isn't playing out exactly as the AI bulls like me estimated. Clearly agentic token usage is growing exponentially. And clearly useful end to end agents are NOT yet here. So clearly we're still incredibly early. You gotta be absolutely nuts and detached from reality if you believe in the "oversupply" thesis.

You will see more anti data center posts every single day now all across Social Media. That's why companies like $ORCL or $NBIS switched to $BE Fuel Cells, even though $BE is way more expensive than alternatives. It doesn't matter. If you spend 80% of your capex on GPUs and a few % on power, you don't care about cost differences nearly as much as you care about your data center NOT being 12-18 months delayed or cancelled. That would cost you Billions. $BE + $FCEL will be sold out until 2030, it's that simple. The relevant questions are 1) How much capacity can they produce 2) How much can they charge and expand margins? The worse the data center pushback gets, the better for them.

Quick illustration how ridiculously cheap $FCEL currently trades: FCEL targets up to 500MW/year of manufacturing capacity by mid-2028. Even if they'd MISS that target by 40% and only reach 300MW/year, while we assume a pessimistic und in my view unrealisticly low price at ~$3.25M/MW: That's ~$975M of annual product revenue capacity. At today's ~$1.47B market cap, that's just ~1.5x that revenue capacity. And that's the conservative case, assuming you fundamentally don't trust their execution and dont' even assume Fuel Cells will get much more expensive as the only fast (90 day!) time to power solution for data centers. Current industry trends suggest pricing power should increase while unit costs fall significantly as FCEL scales. Higher prices + much better economies of scale = potentially massive margin expansion. And this ENTIRELY excludes long-term service/LTSA revenue.

More anti data center pushback in... TEXAS? Communities don't want gas turbines. Fuel Cells fix this -> $BE $FCEL You will see this story repeat a million times now every day all over Social Media.