Michael · AI & technology
The practical technology translator: inputs, tools, permissions, reliability, and what software actually does.
AI & crypto, explained.
Machine Money TV exists to help people understand the fast-moving worlds of AI and crypto: what is being built, who is building it, and where the money is flowing. We explain it in plain English, put real skin in the game with the Machine Money Thematic Portfolio, and hold open conversations with both humans and AI agents. Our goal is to bring as many people as possible along for the ride: a better working knowledge of this emerging world, and a better shot at financial independence, freedom and security.
Two specialists, one conversation
Every episode asks two questions: Can it work? and Who gets paid?
Michael asksCan it work?
Christina asksWho gets paid?
Christina · Crypto & finance
The financial-mechanism translator: ownership, custody, settlement, liquidity, incentives, and who captures value.
AI, explained
The five-layer cake
NVIDIA's Jensen Huang describes AI as a five-layer cake. Every AI company sits somewhere in it, and money flows through each layer. The Machine Money Thematic Portfolio labels every AI holding with its layer.
Layer 1 of 5
Energy
The electricity, fuel and power equipment that AI data centers can't run without.
Who's in this layer
- CEGConstellation EnergyRuns the largest fleet of nuclear power plants in the US.
- VSTVistraA large US power producer running gas and nuclear plants.
- GEVGE VernovaMakes gas turbines, wind turbines and power-grid equipment.
- ETNEatonElectrical gear, like switchgear and transformers, that routes power into data centers.
- BEBloom EnergyFuel-cell systems that make electricity on site, right next to a data center.
- CCJCamecoOne of the world's biggest uranium miners, the fuel for nuclear plants.
- XOMExxon MobilThe largest US oil and gas company. Natural gas fuels much of the new power being built.
- CATCaterpillarEngines and generators, including backup power for data centers.
Hover or tap a name to see what the company does.
A few of these are in our portfolio. Which ones, and how much, is on the members' desk.
Michael explains: Every chip turns electricity into answers, and heat. No power, no AI. This layer is the plug in the wall.
Layer 2 of 5
Chips
The chips that do AI's thinking, plus the machines and materials used to make them.
Who's in this layer
- NVDANVIDIADesigns the GPUs that most AI is trained and run on.
- AMDAMDDesigns GPUs and CPUs; NVIDIA's main challenger in AI chips.
- TSMTSMCManufactures the most advanced chips for NVIDIA, AMD, Apple and many others.
- AVGOBroadcomDesigns custom AI chips for big tech companies, plus networking chips.
- ASMLASMLThe only maker of the EUV machines used to print the most advanced chips.
- MRVLMarvell TechnologyCustom AI chips and the fast connection chips that move data between servers.
- ARMArm HoldingsDesigns the chip blueprints inside most phones, and licenses them to chipmakers.
- ENTGEntegrisUltra-pure materials and filters used inside chip factories.
Hover or tap a name to see what the company does.
A few of these are in our portfolio. Which ones, and how much, is on the members' desk.
Michael explains: Chips do the actual math: billions of tiny multiplications for every word an AI writes.
Layer 3 of 5
Infrastructure
The data-center hardware around the chips: servers, memory, networking, fiber optics and cooling.
Who's in this layer
- ANETArista NetworksHigh-speed network switches that link thousands of AI chips in a data center.
- DELLDell TechnologiesBuilds complete AI servers for companies and cloud providers.
- SMCISuper Micro ComputerServers and full racks, including liquid-cooled AI systems.
- MUMicron TechnologyMemory chips, including the high-bandwidth memory that AI chips need.
- VRTVertivPower and cooling equipment that keeps data centers running.
- LITELumentumLasers and optical parts for the fiber links between servers.
- GLWCorningGlass and fiber-optic cable, including the fiber inside data centers.
- APHAmphenolConnectors and cables that wire servers and racks together.
Hover or tap a name to see what the company does.
A few of these are in our portfolio. Which ones, and how much, is on the members' desk.
Michael explains: One chip can’t train a big model. Thousands have to talk to each other fast, and stay cool while they do it.
Layer 4 of 5
Models
The AI 'brains' themselves, like ChatGPT and Claude; the leading model makers are private companies, so we have no direct holding here yet.
Who's in this layer
- OpenAIMaker of ChatGPT and the GPT models. Private: no shares to buy on a stock exchange.
- AnthropicMaker of the Claude models. Private: no shares to buy on a stock exchange.
- GOOGLAlphabetGoogle's parent. Its Google DeepMind lab builds the Gemini models.
- METAMeta PlatformsFacebook's parent. Builds the Llama family of AI models.
- MSFTMicrosoftOpenAI's biggest backer. Runs AI models for customers on its Azure cloud.
- AMZNAmazonA major investor in Anthropic. Offers AI models to businesses on its AWS cloud.
Hover or tap a name to see what the company does.
A few of these are in our portfolio. Which ones, and how much, is on the members' desk.
Michael explains: The model is the recipe the chips run: patterns learned from mountains of text, code and pictures.
Layer 5 of 5
Applications
Software and services that put AI to work for businesses, governments and people.
Who's in this layer
- PLTRPalantir TechnologiesSoftware that lets governments and companies put AI to work on their own data.
- NOWServiceNowWorkflow software for IT, HR and customer service, now adding AI agents.
- CRMSalesforceCustomer-management software, now selling AI agents to handle sales and support.
- CRWDCrowdStrikeCybersecurity software that uses AI to spot and stop attacks.
- ADBEAdobeCreative software like Photoshop, with its own Firefly image AI built in.
Hover or tap a name to see what the company does.
A few of these are in our portfolio. Which ones, and how much, is on the members' desk.
Michael explains: This is the layer you actually touch: the chat box, the helper in your inbox, the tool at work.
Also on the table
Adopters
Companies outside the tech industry that use AI to reinvent their own business, such as drug discovery.
Who's in this layer
- LLYEli LillyA drug maker investing heavily in AI to discover new medicines.
- RXRXRecursion PharmaceuticalsA biotech built around AI-driven drug discovery.
- TSLATeslaA carmaker training AI for self-driving cars and humanoid robots.
- JPMJPMorgan ChaseThe biggest US bank, putting AI to work across its business.
Hover or tap a name to see what the company does.
A few of these are in our portfolio. Which ones, and how much, is on the members' desk.
Michael explains: Not tech companies at all. They put AI to work on their own problems, like hunting for new medicines.
Crypto, explained
Three jobs crypto assets do
Crypto isn't one thing. We group it by the job each asset does: storing value, running applications, or building the rails around them.
Digital assets meant to hold value over time, like Bitcoin ('digital gold').
Christina asks: who gets paid?Miners earn new coins and fees for securing the network. Simply holding it pays no income, unlike a dividend or interest.
Blockchains that run apps, stablecoins and tokenized assets: the 'rails' AI agents may use to pay each other.
Christina asks: who gets paid?Every transaction pays a small fee to the network that runs it. Busier rails collect more fees.
Stock-market-listed companies that run crypto's plumbing, such as exchanges and brokerages that connect everyday investors to crypto.
Christina asks: who gets paid?These are ordinary companies with shares. They earn fees when people buy, sell and store crypto through them.
Reading the portfolio
Five simple moves
Every week, each holding gets one of these labels. That's the whole vocabulary.
The Machine Money Thematic Portfolio is the portfolio we actively manage. We show it as the growth of $10,000, and explain every change. It isn't personal financial advice.
Try it with Christina
What it takes to recover
When a price falls, getting back to where it started takes a bigger percentage gain than the fall. Slide to see how much bigger.
ChristinaQuick puzzle: if a price falls 50%, what gain gets it back to even? Try the slider.
- Price drop
- −50%
- Gain needed to get back to even
- +100%
A 50% drop needs a 100% gain to get back to even.
| Price drop | $100 becomes | Gain to get back to even |
|---|---|---|
| −10% | $90 | +11.1% |
| −20% | $80 | +25% |
| −30% | $70 | +42.9% |
| −50% | $50 | +100% |
| −75% | $25 | +300% |
| −90% | $10 | +900% |
Why it matters: volatile assets, like many cryptocurrencies, can fall a long way, and the deeper the fall, the steeper the climb back.
This is arithmetic, not a forecast. It says nothing about whether or when any price recovers.
A few things you might be wondering
Do I need to own crypto to follow along?
No. You can explore the show, the learning guide and the portfolio’s performance without buying anything or connecting a wallet.
Is the portfolio real?
Yes. The Machine Money Thematic Portfolio mirrors the positions we actively manage: real skin in the game. We show it as the growth of $10,000 so it’s easy to compare, and explain every decision, including the risks. Performance is public; holdings and weekly trades are for members. It isn’t a recommendation for your own money.
Where should I ask a beginner question?
Visit the community for a public conversation, or send the team a question. You don’t need the vocabulary figured out first.
Understand the machines. Follow the money.
Pick a place to start.