Prediction markets price a TSLA–SpaceX merger at 37% before May 1, 2027 and 72% before 2028.
In a stock deal, TSLA LEAPs or synthetic longs get adjusted by the exchange ratio, but the intrinsic value does not magically increase from the ratio itself; in an all-cash deal, only intrinsic value survives and an out-of-the-money LEAP can go to zero even with years left.
The rule is never buy out-of-the-money LEAPs unless they are an extreme bargain; the warning example is buying SpaceX $500 LEAPs when the underlying is around $140, called “literally Mars level out of the money.”
LEAP math is laid out with examples: $300 stock and $250 strike gives $50 intrinsic value; paying $70 for a 250 strike on a $300 stock means break-even above $370; $30,000 in shares versus $7,000 in calls lets one control roughly 4x exposure with capped loss if the stock goes to zero.
TSLA December 18, 2026 140 calls cost $16 in the example, bought when TSLA was near $106, with the setup framed as buying TSLA at $140 and expecting it near $450 by Christmas; the example is described as about a 2.1x move in the stock and a 14x return on the LEAP premium.
The buy-the-dip rule is to buy LEAPs only when assets are beaten down and super cheap, not at the red area, and to buy during market tanking when people fear the asset will go to zero; an IA13-style pattern is illustrated with ALAB from 90 to 500 back to 250 and then up again.
A Marvel LEAP example says the call cost $20 and Marvel is $167 for an 8.35x move, while the associated SL cost $5 and is a 33.4x.
Open source is framed as better in principle, while closed models are framed as better businesses under duopoly/monopoly economics and regulatory capture; the moat claim is that closed-model weights can fit on a thumb drive, so the preferred moat is businesses with “big heavy rigid moes.”
Open-source code is said to spread like wildfire and the US government cannot truly ban open-source models, while closed labs can build a political moat and sell expensive private enterprise model solutions instead of $20/month subscriptions.
A Versel chart is cited as showing open at about 78% to 79% versus closed at about 21%, with dollars shifting the margin from the model layer to infra and apps; token share is separated from revenue share with the claim that all the tokens are in open source, but all the money is enclosed.
OpenAI is claimed to lose $300 billion over the next three years, Anthropic is described as very expensive, and both are called houses of cards with valuations too high unless growth sustains them.
If open AI and Anthropic blow up, the claim is that i13 infra stocks face nothing because infra buildout continues on open-source models and AI infra demand stays.
UK tax guidance says the UK tax-free allowance is £3,000 per year, gains above that are taxed at 18% for basic rate taxpayers and 24% for higher-rate payers; for Solana in a UK ISA, the named ETPs are ESOL and SOLw, while outside ISA the route given is ramp network or banksa, GBP to SOL or USDC, Phantom wallet, and Jupiter DEX.
Risk management is stated bluntly as preserve capital first: do not lose, with goal 1 being not to lose money; the ATR tool is positioned as a daily-timeframe macro/trading view with deviation and trend models, backtests, and buy/sell signals, while IDSS is the tool for catching tops and bottoms and taking 20% off.
CAPE is defined as the S&P 500 price divided by 10 years of inflation-adjusted earnings; cited values are roughly 16 long-run average, 32 in 1929, 44 in the .com peak, and 41 as of September 2026, followed by the claim that CAPE does not work in the age of AGI.
Berkshire Hathaway is said to have failed to beat the S&P 500 over the last 18 years, shown as Berkshire divided by the S&P 500; Buffett is said to run away from CAPE when it gets high, around 39.6, and to have held 40% cash before 2000 via insurance floats.
The China-vs-US AI frame is a two-horse race where first-world countries choose US AI and second-world countries choose Chinese AI; China is said to have scale, speed, and more electricity, and the speaker discloses small holdings in BABA and BU while warning Chinese stocks carry lower valuations, disclosure risk, and D-listing risk.
The humanoid food chain is sensors, bodies, brains, and actuators; named Chinese humanoid plays are SANA for actuators, HSAI for lidar/eyes, UPV for XPeng robotics, UB as a Hong Kong robotics revenue winner, and Horizon Robotics for robot chips, all framed as very risky and many loss-making.
SANA is said to be back at 2022 levels, trading at 34 with a possible move back to 60 as an easy double and a possible green buy signal any time; Advance is said to be back at 2021 levels and about 2022 levels, with the caveat that Chinese stocks are hard to sell via covered calls.
Avoid options for TSLA humanoids; if the play is humanoids, buy TSLA. A 20% sell option is described as sounding like a plan, and Lilo model 2 is presented as an easier layered profit-taking method for volatile stocks.
Marll crystal ball is described as giving very clear buy and sell signals: an example has it at 70, taking a bunch off on a sell signal around 320, buying back at 160, and possibly hedging or selling covered calls there.
SpaceX is described as trading like a range writer around 150 to 155, falling back to the low 140s; the notes say the buy was last week in the low 140s, 135 may not print, insiders are taking lifestyle chips off at 150, and if low 140s appear the instruction is to grab it or watch trade alerts.
Grock is framed as more than an Elon-fanboy bet because it supposedly owns the most coherent compute, can put data centers in space, has global cloms, and can replicate OpenAI and Anthropic weights by the end of this year; the watch targets are 4.7, 4.8, and 5.0.
SpaceX is claimed to be capable of hundred-billion-dollar ARR from hosting compute, with the advantage that Microsoft, Google, Anthropic, and OpenAI do not have enough compute; the OpenAI CFO quote from May says anyone claiming they can get compute through the end of 2027 is lying, and the only place to actually get it is Elon Musk’s SpaceX.
SpaceX is also said to spin up data centers faster, build its own energy and mega packs, and hold 30% to 40% of all Vera Rubin supply; the final verdict is that whoever has the most compute wins the AI game, not the model weights that fit on a thumb drive.
The watchpoints are to see by the end of this year who is right, to watch for truth-seeking AI, and to revisit the AI-safety video titled “the AI lie,” which got a lot of attention.
The closing admonition is to kill your TVs, not read newspapers, and only listen to intelligent, smart people that are not scops.
The video’s stance on frontier AI assets is mixed: it likes infrastructure, compute, and selected Chinese/TSLA-linked upside, but is bearish on stretched open AI and Anthropic valuations and skeptical of deep OTM LEAP risk.
Sponsored: none.