Stephen | CFU

I've been trading options since 2008. The market then looked nothing like today. → Liquid chains on names that didn't exist → Stocks with options expiring 3 times a week → Strategies that used to take days now fill in seconds What hasn't changed: The people who win are the ones who learn the boring fundamentals first. Paper trade. Read the Greeks. Start with money you can afford to lose and not lose sleep over. The instrument is more powerful than ever. The discipline to use it correctly is rarer than ever. #nvda #nvidia #optionstrading #premiumselling #thetagang #cfu #tesla #tsla #optimus

There's a version of options trading where you spend your weekend reading earnings reports, refreshing charts, second-guessing every position. And a version where you have a checklist, a process, and people who've run the same system hundreds of times. The strategies are often identical. Everything around the trade is not.

@FinNewz Bald spot's got its own risk management. Defined downside.

@StephenCFU Why the hat though?

I've been trading options since 2008. The market then looked nothing like today. → Liquid chains on names that didn't exist → Stocks with options expiring 3 times a week → Strategies that used to take days now fill in seconds What hasn't changed: The people who win are the ones who learn the boring fundamentals first. Paper trade. Read the Greeks. Start with money you can afford to lose and not lose sleep over. The instrument is more powerful than ever. The discipline to use it correctly is rarer than ever. #nvda #nvidia #optionstrading #premiumselling #thetagang #cfu #tesla #tsla #optimus

Reading it exactly the same way. The irrationality is the opportunity, but only if you stay anchored to real numbers while everyone else trades the story. That's the trick with this cycle: the hype and the fundamentals are both true. AI is genuinely transformative AND plenty of these valuations are detached from the financials. Holding both at once is the edge. Orders, deployments, actual revenue. That's the job.

@StephenCFU We are climbing the peak of inflations on the hype cycle. Investors still can benefit from the markets irrationality by investing with awareness to the hype. But always checking the real numbers, orders, deployment etc matters.

Humanoid robot projections by 2030: → Figure CEO Brett Adcock: 1 billion → $TSLA Elon Musk: 50 million Optimus → Goldman Sachs: 250,000 → Morgan Stanley: 40,000 (US) → Bank of America: 1 to 1.2 million The people building the robots see a market 1,000x larger than the people analyzing it. One of these groups is going to be very wrong.

$NVDA just delivered the first Vera Rubin systems. → 25x the AI compute of the H100 → First batch shipped to Anthropic, OpenAI, xAI, Oracle → Frontier models (GPT-5, Opus) will be upgraded on this → A separate version reportedly built for outer space Anyone underweight $NVDA is making a directional bet whether they realize it or not.

$ALAB is becoming one of the quieter winners of the AI build-out. → Up 13% on JPM tech conference guidance → New Scorpio X fabric switch plugs into 90% of data center equipment → CEO: their largest product by year-end The AI trade isn't just chips and fiber. It's everything that connects them. The picks-and-shovels list keeps getting longer.

@OurNewOverlord Right — that's the real distinction. Musk's number is still aggressive, but he's actually scaled hard manufacturing before. Adcock hasn't proven mass production yet. Earning the projection vs. making it are different things.

@StephenCFU Yup. Even Elon's projection is crazy, but at least he's been through production hell. Brett Adcock on the other hand...

@podcast_alpha_x Well said. Messy is the realistic base case, and messy creates the volatility. The dispersion between the projections is exactly where the tradeable moves come from. Rarely clean, rarely as fast as the bulls say or as slow as the skeptics fear.

@StephenCFU The reality will land in between - neither will be right or wrong. Real world AI will be messy and create a lot of volatile moves for investors to take advantage of.

@kaiba_1991 You're right. It's 10 billion by 2040, not 1 billion by 2030, Adcock said it on the Diamandis Moonshots podcast (EP #57). I pulled the figure from a secondhand chart without verifying. My mistake, thanks for the push.

@StephenCFU Show me specifically where adcock said 1 billion by 2030. It doesnt exist. Only thing Im unsure about is if you are confused or lying for views.

@pm4lyfe Agreed. The demos hide how far the hardware still has to come on dexterity, battery, and cost. Software is racing ahead. The physical layer is the real gate.

@StephenCFU AI + hardware still need some major breakthroughs before this scaling happens

@LeandroVallej14 That's the honest version of the bull case. Get the timeline far enough out and almost any number is defensible. The argument was never "if," it's "when," and 2055 respects that better than 2030.

@StephenCFU 10 billion by 2055

Sharp point. The unit economics are the wall everyone skips. A car-priced machine to do tasks people already do cheaply doesn't pencil at the consumer level yet. And you're right that labor displacement at scale invites a political response, which is its own ceiling on the timeline.

@StephenCFU The current market for maids and home chefs in the us is $20B The idea that masses can afford something as expensive as a car - to do work they are already doing is not real. If they could do actual labor and replace workers at scale would provoke political intervention

@servetkodu_EN This is the best comment here. The spread IS the signal. And you've named the right metrics. Forget the top-down projections, watch deployed sites, uptime, intervention rate, and who starts reporting real revenue. That's the difference between a narrative and a business.

@StephenCFU The spread in these forecasts is the signal. We are still in narrative territory. The cleaner read is not top-down robot counts. It is customer sites, uptime, intervention rate, and which suppliers start reporting real volume.

Fair, and worth saying: on near-term numbers, the analysts usually are closer. Operators are almost always too optimistic on timing. The case for watching the builders isn't that their number is right. It's that they sometimes see the direction before the models do, even when they're wildly wrong on the date. Analysts nail the next few years. Founders occasionally catch the decade. Different jobs.

@StephenCFU The analysts are usually the correct ones.

@DataLover4 Hard to argue. Four years from where the actual deployed numbers are today to 50 million is a physical-supply impossibility, not just a demand question. The near-term number is far closer to yours than to the headline.

@StephenCFU Having 50 Million in 2030 is not possible. We are already in 2026. May be 1 or 2 is possible.

@Stonefoxcapital Good read. Being early creates the category and invites everyone in to compete for it. The first mover proves the market exists, then spends years defending it. Happens in almost every platform shift.

@StephenCFU yes. The problem facing Musk now is that he’s bringing competition along by being early on timing.

@kaiba_1991 The projections are from publicly reported analyst estimates (Morgan Stanley, Goldman, BofA, Citi) and the CEOs' own public statements (Adcock, Musk). Happy to point you to the specific reports if useful.

@StephenCFU source?

@WangKaiwen753 That's the whole story in one chart. A 1,000x spread between the operators and the analysts.

@StephenCFU Wow, robot projections are getting crazy! Looks like some major differences in estimates between Figure CEO and Elon Musk 🤖.

@Stonefoxcapital History backs you up. Usually right on direction, usually early on timing.

@StephenCFU well, we know Musk is always too optimistic.