Jensen Huang, CEO of Nvidia, is telling you where to invest in 2026.
He has personally directed Nvidia's capital into 8 specific companies for a combined total of over $45 BILLION.
This is where the most important company in the AI economy is putting its money.
Here’s the full list:
OpenAI: $30 billion
The largest commitment of the 8. Nvidia is funding the buildout of OpenAI's compute infrastructure from the inside. OpenAI is also Nvidia's single largest customer.
GLW Corning: $3.2 billion
Optical glass and fiber to physically connect AI clusters. You cannot move data between millions of GPUs without it.
IREN: $2.1 billion
AI cloud provider with one of the deepest power positions in North America.
MRVL Marvell: $2 billion
Custom networking chips that move data between GPUs at massive scale.
LITE Lumentum: $2 billion
Lasers and optical components for the fiber backbone of every AI data center.
COHR Coherent: $2 billion
Fiber optic transceivers that connect GPU clusters inside data centers.
CRWV CoreWeave: $2 billion
GPU-as-a-service provider. Nvidia's largest cloud customer outside the hyperscalers.
NBIS Nebius: $2 billion
AI cloud infrastructure company. Quietly building hyperscale GPU capacity for the AI labs.
Whatever Nvidia is buying is where the money is going next.
At The Assembly, we’re a team of 8 with one goal: help you find the right stocks early.
Turn notifications on so you don’t miss our alerts. This is VERY important.
If you’re not following us yet, you will regret it later.
🚨 Donald Trump just revealed which stocks he’s buying in 2026
FULL LIST:
VOO
IWB
NOW
NVDA
RSP
ADBE
WDAY
ORCL
MSFT
AVGO
SNPS
CDW
PG
CDNS
TT
TXN
FIS
MSI
ETN
XLI
TDG
AMZN
JBL
COST
AXON
COMT
KRUS
DELL
BA
UBER
IEMG
AAPL
XLK
NVR
SMCI
GOVT
ICE
KLAC
FFIV
AVB
XEL
ARES
WM
EFA
CRM
PNC
DVA
GOOGL
NWSA
WST
HD
CVNA
IEX
NFLX
VTI
TOMORROW, every fund over $100M must legally disclose their Q1 2026 trades to the SEC.
Every major filing will be posted on @InTheAssembly so follow them with notifications.
If you don’t follow them, you will regret it.
Top 5 Trading Books For Beginners:
1.) Trading In the zone
2.) Reminiscences of a Stock Operator
3.) Technical Analysis of the Financial Markets
4.) How to Day Trade For a Living
5.) Best Loser Wins
I get asked a lot where to start learning how to trade and I think these 5 are really solid.
Ironically most people think that strategy or indicators is what they need when in reality the most important thing to master is patience and mindset and managing your risk.
The best teacher will always be the market and your experience actually trading.
Start with a small account and learn to grow it over a year and then add more as your confidence grows and your win rate becomes measurable.
These bots beat 99% of traders on polymarket
they make thousands of dollars every day
they just exploit the same patterns over and over
- cluster following
tracks dozens of profitable accounts hitting the same outcomes. when 10–15 of them move in one direction at the same time, it does the same
- probability decay
looks for overheated markets where the price is around 70% but reality is closer to 40%
it goes the other way and profits as the hype dies down and the price slides to the fair probability
- corridor volatility
the market spends 90% of the time in a range, like 40–60. when price shoots out, the bot goes against it and closes the position once the price returns to the range
- resolve sniper
trades only in the last hours or minutes before market resolution. pulls official sources, and if the outcome is clear but price is hesitating, it jumps in the obvious direction with minimal risk
'cry.eth2' makes hundreds of trades daily, hedges positions, farms micro spreads, snipes late resolves
almost zero profile, pure bot with 0 profile views
his profile: https://t.co/dE51pvHnUb
How to find new traders to follow on @Polymarket 🔍
People ask me this a lot, so here’s my filter setup on Polymarket Analytics 👇
0️⃣ Starting point
Always start here → https://t.co/MnbQKXEgdn
By far the cleanest place to scan profiles and see who’s worth following.
1️⃣ Overall PnL
Overall PnL ≥ $1,000 (no upper limit)
The idea isn’t to only look at people with +$500k, because small wallets can perform too (look at @thegreektrader, @25usdc, @holy_moses7 early on)
→ I care if someone knows how to win, not how much they started with.
2️⃣ Current Value
Current Value ≥ $1,000
This filters for traders who still have $ in the game.
→ Want traders who are currently active, not just sitting on old PnL.
3️⃣ Active Positions
Active positions ≥ 2
→ Simple check that they’re still trading and the account isn’t dead.
4️⃣ Total Positions
Total positions ≥ 500
A lot of traders look great over 20–30 bets, but that can just be a lucky streak.
With 500+ positions:
→ you see how they perform over a real sample size
→ combined with other filters, it shows their consistency
5️⃣ Win rate
Between 75% and 90%
Above that range, you mostly find people betting on outcomes that are “already resolved” but not yet bonded, which skews the stats.
Between 75–90%:
→ they take calculated risks
→ they’re selective, not ticket spammers
Use the data to find profiles that fit your approach.
Analyze ≠ Copy 🤝
Here’s what the list looks like with these filters applied (you’ll probably recognize a few names).
I'm being asked what x402 is, so here's why you should care:
a) Background:
- The x402 protocol enables agents to make payments onchain
@a16zcrypto 2025 "State of Crypto" Report specifically mentioned x402 in the context of agentic payments, which is anticipated to hit $30 trillion in autonomous txs value by 2030
- It's dubbed the "Payments MCP" [Model Context Protocol];
I'll link my MCP article below so you have an understanding, but the "quick and nasty" is that MCP enables agents to interact with real world data, so payments MCP enables agents to transact with their own wallet
b) Backers: @coinbase + @Cloudflare
- So you'll see the bulk of this activity on @Base; But it's also cross-chain (Solana, Polygon, Near etc)
- It's why you've seen several @virtuals_io x402 agents pop off in the past few days - several of the ecosystem agents use the x402 protocol
- The acquisition of Echo by Coinbase took centre stage earlier this week; but this is Coinbase AI Agent baby; and IMO (I'm biased) far bigger. Sorry @icobeast
c) Relevant x402 protocols:
- @pingobserver - first token minted by x402
- @heurist_ai
- @virtuals_io
- @questflow
- @AnchorBrowser
- @GoKiteAI announced its partnership with @brevis_zk this week for privacy around payments with agents; it uses x402
- @Cloudflare Agents SDK + Google A2A payment protocol
- @PayAINetwork
- $DREAM (Virtuals)
- $SANTA (Virtuals)
- $GLORIA (Virtuals)
- $AURA (Virtuals)
Honestly the list could go on, so I'd advise researching before spraying and praying.
NFA on these - they're simply affiliated with using the x402 and it doesn't necessarily mean value will accrue to these tokens. If anything it'll accrue to the protocol backers itself - COIN.
You can check https://t.co/TomVpva3Mq (see thumnail)
d) Other Insights:
1) @circle has an ongoing relationship with Coinbase and the main currency transacted by these agents is Stablecoins to pay for compute etc. So as you can imagine USDC will likely be the stable of choice
2) Neo Banks are gearing up for AI Agent payments:
- @tether launched its wallet development kit; the tagline is that there's anticipated to be trillions of wallets. These obviously won't be human wallets, so I'll let you connect the dots
- @useTria partnered with several AI protocols and this week with @billions_ntwk for zkKYC as privacy around payments comes to the forefront
- @Visa + @Mastercard announced their AI Agent programs a few months ago and are rolling out the infra for trillions of transactions. Almost all of these Cards from these Neo Banks are VISA/Mastercard.
- @PayPal ventures led the $18m Series A for Kite AI; PyUSD coming in hot here for agentic stablecoin usage?
Next step: Robots will have their own crypto wallets with virtual cards to make payments in the real world. Welcome to physical AI!
3) @a16zcrypto report on AI referenced x402 specifically when talking about AI Agents and payments, anticipating $30 trillion in agentic purchases by 2030
Interesting quote: "x402 is what the internet was supposed to be - payments via https" - @_imhamzah
4) @RibbitCapital released their 2025 "Token Letter" which discusses agents + payments (see "knowledge" tab on their website):
- Page 1/2: AI systems are economic participants that will earn and spend
- P33/34: Stripe and upstarts like PayOS will capture a lot of value building this infra for agents
- P33-35: Stablecoins and programmable money will be used by both humans and agents for payments and asset mgmt.
- P39/40: AI agent payments will become new financial rails for autonomous entities
TLDR: AI Agents + Payments is only going to get bigger. x402 is positioning itself at the enabler of this parabolic payment boom.
The cats out of the bag, so how would you capitalize on the upside?
Taking a cut in fees on trillions of txs isn't a bad idea.
The number of people trading leverage that don’t understand the mechanics of the exchange they are using is ridiculous.
If you don’t understand funding, ADL, reduce only, the difference between isolated and cross, limit/market orders or conditional order types, stop trading.