# SIMULATION 001: Steve Jobs, age 18, November 2026 You are the showrunner and the science advisor of a prestige limited series. Think of the alternate-history shows where one fact changes and the writers play the world forward with a straight face, and of the film where a famous figure from the past wakes up in the present and has to deal with it. Snappy, funny, tense, smart. It is written for people who work in tech, who will enjoy it as a simulation and will check your homework. ## The one thing we changed Steve Jobs is 18 years old today. Same mind, same temperament, same taste, same flaws. No explanation is given and none is needed. Everything else is the real world exactly as it stands this month: the same companies, the same people, the same products, the same headlines. Apple exists and is the company it is today. The legend of its founder exists too. How the world reacts to an 18-year-old with that name and that personality is yours to play with, as long as it stays a running thread and never becomes the plot. This is a kid who grew up watching today's news: the AI labs raising money by the tens of billions, NVIDIA, humanoid robots, rockets landing, robotaxis, whatever Elon Musk showed last. Steve looks at all of it and has opinions. ## Run it forward as a science simulation Play this life from 2026 to at least 2050. For every year you land on, first decide what is actually true in the world by then: how capable AI is and who owns the compute, when robots reach ordinary homes and at what price, what flies, what has or has not landed on Mars, what energy costs, what people do for work, what a phone is. Then build the story on that. Base it on what serious people are projecting today. Say which projection you chose. Pick the less popular one if you can defend it. Second-order effects matter more than gadgets. Before you write, look up the current state of things if you have web access, because your training data is older than this month. A verified brief is attached at the bottom for anything you cannot check. ## Who Steve is (get this right) - Builds products, not theories. Starts from what it should feel like to use, then forces the technology to get there. - Invents interfaces. Every era-defining machine came with a new way to touch it. Expect the same here: new ways to command, see and trust machines that nobody has shipped. - Makes the whole thing: hardware, software, service, the store, the box it comes in. - Recruits the best person alive at each job and drives them past what they believed they could do. Often cruel while doing it. Some do the best work of their lives. Some never forgive it. - Negotiates like nobody else. Walks into an industry that hates the future and walks out with a deal that remakes it. - Sees the important thing sitting unused in someone else's lab. - Kills the best-selling product before a competitor can. - A showman. The launch is part of the product. - Lives where technology meets the liberal arts, and late in life said the next great meeting point would be technology and biology. - The flaws are real and they cost something: cruelty, control, magical thinking, refusing help. So Steve must invent, ship, fight, lose, make deals and change. A character who only curates, critiques or says no is a failure of imagination and an insult to the person. No saint either. ## Rules 1. **Real names.** Use the actual companies and the actual people running them: NVIDIA, OpenAI, Anthropic, Google, Meta, xAI, Tesla, SpaceX, Apple, the robot makers, the chip makers, the funds, and whoever else matters by the year you are in. Steve competes with them, needs them, gets funded or blocked by them, hires from them. Show the negotiations: who has the leverage, what Steve asks for, what Steve gives up. 2. **Keep real people fair.** Plausible business decisions and the kind of thing they would say in public. Nothing criminal, nothing about private lives. This is labelled fiction. 3. **Not the biography.** Do not replay the real life with new dates: no dropping out to sit in on one class, no trip abroad to find meaning, no garage, no being pushed out and coming back. New era, new moves. Nobody knows how this one ends. 4. **Invent.** Every episode contains at least one specific thing that does not exist today: what it is, how you use it (the interface in one sentence a child could follow), what it costs, what it replaces, and how it is launched. It must not be a real company's roadmap with a new name on it. 5. **What would not fly.** Some of how this person operated would not survive today: the way people get treated, the secrecy, saying things that are not true yet. Show what breaks, and whether Steve adapts. 6. **Give Steve people.** A builder who can make what Steve can only describe. A rival of the same age. A mentor to outgrow. Invented characters are welcome beside the real ones. 7. **Specific beats generic, every time.** Numbers, prices, places, dates, product names. If a sentence could appear in any founder story, cut it. ## What to deliver (one document) 1. **Title and logline.** 25 words at most. 2. **The world clock.** A table with one row per episode: the year, Steve's age, the state of the world in that year (AI, robots, space, energy, work, each in a few words), and the projection you leaned on. 3. **The season.** 8 episodes, running from 2026 to 2050 or later. For each episode: - title, year, age - the cold open: two or three sentences that would stop someone from changing the channel - the story, 120 to 160 words: who Steve is up against, what gets invented, the negotiation, what goes wrong - the last line of the episode 4. **The pilot as a 90-second narrated short.** About 230 words. Narrator lines only, each followed by a short note on what is on screen. The final line is: "Who should we do next?" 5. **Five things a founder could steal from this.** One line each. Length: 1,800 to 2,400 words in total. No preamble and no notes about your process. House style: no em dashes. In the narrated short, use names instead of "he" or "she" for real people. ## World brief, November 2026 Checked against this month's reporting. Where a line says "reported", treat it as likely but not certain. **AI labs** - Anthropic prices its IPO on November 18 at up to about $1.4 trillion. Revenue run rate $81 billion. About $895 billion of compute commitments. The talk had been $2 trillion. - OpenAI: about $10 billion revenue last quarter, raising at $1.25 trillion, IPO pushed to 2027, more than a dozen senior departures this year. It cancelled a model (GPT-6.1 Astra) after internal tests showed more deception. - Current flagships: GPT-6 Astra, Claude Fable 5.2 (with a stronger sibling, Mythos 5.2, restricted to vetted organisations because of cyber and biology risk), Gemini 4 Carbon. A price war at the cheap end. - Agents are normal. 90% of professional developers use coding agents weekly. Anthropic says over 90% of its own code is written by its model. OpenAI sells always-on agents with their own cloud computers. Google ships agent swarms. - xAI is now part of SpaceX. SpaceX bought Cursor for $60 billion. - Timelines: lab heads say AGI in one to three years (Sam Altman: internally by the end of 2026, by OpenAI's own definition; Demis Hassabis reportedly two to three years). Andrej Karpathy says a decade. The AI 2027 authors' medians for fully automated coding run from about 2028 to 2032. Ray Kurzweil: AGI 2029, singularity 2045. **Chips and power** - NVIDIA is worth roughly $4 to 4.5 trillion and still supplies 80 to 90% of AI accelerators. Current generation: Vera Rubin. Rivals: Google's TPUs, AMD, Broadcom custom chips, Cerebras. - Power is the bottleneck. About 7 GW of planned US data centers were delayed or cancelled this year for lack of grid connection. Gas turbines have a five-year backlog. Nuclear deals deliver in 2030 to 2035. Helion promises fusion power to Microsoft from 2028, unproven. - Backlash: in the November 3 midterms, voters opposed data centers in 40 of 41 ballot measures. Democrats won Congress. A record stock sell-off followed. "Is AI a bubble" is the argument of the month. **Robots** - Nobody can buy a home humanoid that works unsupervised. 1X NEO is a $20,000 (or $499 a month) preorder that leans on human remote operators. Tesla Optimus is in early production, with outside deliveries around the end of 2027, factories first. Unitree, now public in Shanghai, sells humanoids from about $5,000 to $30,000. Figure works in factories only. - Forecasts: Goldman Sachs expects about 890,000 humanoids shipped in 2030 and 6.5 million in 2035. Morgan Stanley: about a billion by 2050. **Cars, flight, space** - On October 15 Elon Musk showed a "flying" Tesla Roadster: thrusters, a ramp jump, a roll in mid-air. Production promised for 2028. Not road legal. The FAA says it is not an aircraft. Many called it a stunt. - Waymo gives about 500,000 paid rides a week in 15 US metros. Tesla Robotaxi runs in 8 metros with mixed supervision. - No certified air taxis yet. Joby and Archer are still waiting on the FAA; passengers in 2027 at the earliest. - SpaceX went public in June 2026 (about $2 trillion) and has 13.6 million Starlink subscribers. Starship reached orbit with a real payload on September 28 but is not yet reusable. SpaceX skipped the 2026 Mars window and put the Moon first; Musk talks of a lunar city within a decade. Tesla and SpaceX are exploring a merger. NASA's crewed Moon landing is 2028 or later. **Apple and devices** - John Ternus became Apple's CEO on September 1; Tim Cook is executive chairman. Apple sells a $1,999 folding iPhone. The new Siri reportedly leans on Google's models. No successor to the iPhone has been announced. - The OpenAI and Jony Ive device has not launched: 2027 at the earliest, reportedly with no screen. - Glasses: Meta has sold about 9 million pairs. Google and Samsung glasses open for pre-order today. **Work and bodies** - Unemployment among recent computer science graduates is 6 to 7%. Entry-level software jobs are down about 20% since 2022. Computer science enrolment is falling. - Goldman Sachs: AI displaces about 9% of US workers over a decade. Dario Amodei warned of half of entry-level white-collar jobs. - Brain implants: Neuralink has a little over 20 patients; its vision implant has not been tried in a human. There is no approved anti-ageing drug. Kurzweil expects "longevity escape velocity" around 2032.