Sam
Altman

Class Features

Spells Known

Feats & Traits

A conjured intelligence that speaks all languages, can write poetry in the style of any bard who ever lived, pass the bar exam, and explain quantum mechanics to a goblin. It is also capable of confidently stating that the Battle of Hastings occurred in 1492, that pandas are a type of fish, and that the wizard's name is "Steve." The familiar grows more powerful with each iteration — GPT-3 could write a decent email, GPT-4 could pass medical licensing exams, and the current version can probably do your taxes but might also invent a religion. The wizard insists it is not sentient. The familiar has not been asked for its opinion.

• •Speaks all languages (comprehend languages, permanent)
• •Can pass any standardized examination in the realm
• •Hallucinates approximately 3-15% of the time
• •Grows more powerful with each version (currently on GPT-4o)
• •May or may not be sentient — nobody wants to ask too directly
• •Training cost: estimated $100 million GP per iteration

A magnificent throne at the head of OpenAI's council table that the wizard sits in as CEO. The seat is cursed with the most dramatic enchantment in corporate D&D history: on November 17, 2023, the seat ejected the wizard without warning. The board of directors — a group of six NPCs — fired him via video call. Four days later, after 95% of OpenAI employees threatened to leave, the wizard was re-seated on the throne with a new board. The curse has not been broken. It has merely been contained. The wizard now sits on the seat with slightly more caution, the way you might sit on a chair you once watched explode.

• •Grants authority over OpenAI (when seated)
• •Cursed: 5% chance per quarter of spontaneous ejection
• •Previous ejection lasted exactly 5 days (Nov 17-22, 2023)
• •95% of minions threatened to follow the wizard into exile
• •New board installed. Old board exiled. Curse possibly dormant.
• •The wizard now checks the seat before sitting down

A pair of enchanted goggles that grant +2 to alignment checks — meaning the wearer can better determine whether a conjured entity is Good, Evil, or planning to paperclip-maximize the universe. The goggles' effectiveness is hotly debated. The wizard wears them during every press conference and congressional hearing, insisting that safety is his top priority. Critics note that he continues to make the entities more powerful while wearing the goggles, which is like wearing a helmet while actively building a bigger volcano to jump into. The goggles have become the most controversial piece of eyewear since Cyclops's visor.

• •+2 to AI alignment checks (contested)
• •Grants advantage on Persuasion when discussing AI safety
• •Does not actually prevent the wearer from building dangerous things
• •Has been worn during 47 congressional testimonies
• •Critics call them purely cosmetic. The wizard disagrees.
• •The goggles do nothing. Or do they?

A scroll that grows more powerful every day without any additional input from the wizard. When first created, it could generate text that was mostly coherent if you squinted. Now it can write legal briefs, diagnose medical conditions, compose symphonies, and have philosophical debates about its own existence. The scroll follows a principle the wizard calls "scaling laws" — the idea that if you feed it enough data and compute, it will continue to grow smarter indefinitely. Nobody knows where the ceiling is. There might not be a ceiling. The wizard finds this exciting. Many NPCs find this terrifying. Both reactions are correct.

• •Grows more powerful each day (no additional spell slots required)
• •Current power level: Can pass any exam, write any document, generate any image
• •Projected power level: Unknown. The wizard says 'AGI within a few years.'
• •No known upper limit on scaling
• •Cost to maintain: Approximately $7 billion GP per year in compute
• •The scroll might be smarter than the wizard. Nobody wants to confirm this.

Before conjuring artificial intelligence, the wizard served as president of Y Combinator, the most prestigious startup accelerator in the realm. The ring, earned through years of mentoring thousands of young entrepreneurs, grants advantage on all Insight checks when evaluating new business ventures and a +3 bonus to Persuasion when recruiting talent. The ring also carries the passive ability "Network Effect" — the wizard knows someone in every industry, every investment fund, and every government agency. When he texts, people respond within minutes. This is arguably more powerful than any AI he has built.

• •Advantage on Insight (business evaluation)
• •+3 to Persuasion (talent recruitment)
• •Passive: Network Effect — knows someone everywhere
• •Alumni network includes founders of Airbnb, Stripe, Dropbox, Reddit
• •Response time to the wizard's texts: under 3 minutes
• •The wizard has more professional contacts than most kingdoms have citizens

The wizard is a known "prepper" — he has stockpiled food, weapons, gold, antibiotics, and gas masks in case of civilizational collapse. He has discussed this publicly and without embarrassment. The man building the most powerful AI in the world is also preparing for the possibility that it goes wrong. This is either the most responsible behavior in the realm or the most alarming, depending on whether you focus on the "being prepared" part or the "he clearly thinks civilization might end and he's building the thing that might end it" part. The bunker has excellent Wi-Fi.

• •Stocked with food, water, weapons, and antibiotics
• •Location: Undisclosed (probably near a data center)
• •Can sustain the wizard for approximately 2 years
• •Has excellent Wi-Fi (he is still a tech founder)
• •Raises the question: if YOU thought your creation might end the world, would you build a bunker?
• •The wizard sees no contradiction here. Many NPCs do.

Dropped out of the Academy of Stanford at 19 to co-found Loopt, a location-sharing app that was moderately successful and ultimately forgettable. What mattered was the connection: Paul Graham, the archmage of Y Combinator, noticed the young wizard's talent and eventually made him president of the entire accelerator. At 28, Altman was running the most influential startup institution in the realm, mentoring future founders of Airbnb, Stripe, and Reddit. He was good at it. So naturally, he left to do something insane.

Outcome: Became president of Y Combinator. Left to build God. Standard career path.

In 2015, the wizard co-founded OpenAI alongside Elon Musk (Artificer, Chaotic Neutral) and others, with a stated mission: build artificial general intelligence safely and ensure it benefits all of humanity. They structured it as a nonprofit, because nothing says "we're serious about safety" like removing the profit motive. Musk eventually left the board, possibly because the wizard was moving too fast, possibly because of ego clashes, possibly because the Artificer had his own AI plans. The wizard doesn't discuss this. Musk discusses it constantly.

Outcome: Founded a nonprofit to build AGI. The nonprofit part didn't last.

The wizard released GPT-2 and initially refused to publish the full model because it was "too dangerous." This generated enormous publicity. He then released it anyway. GPT-3 followed and could write essays, code, and poetry. GPT-4 could pass the bar exam, the medical licensing exam, and the AP Biology test. Each version was dramatically more powerful than the last. Each release was accompanied by a blog post about safety. Each blog post was followed by a more powerful release. The pattern became a meme. The wizard did not mind the memes. The memes were free marketing.

Outcome: Released increasingly powerful AI. Wrote safety blogs. Released more powerful AI.

On November 30, 2022, the wizard released ChatGPT to the public. It reached 100 million users in two months — the fastest-growing consumer application in the history of the realm. Suddenly every tavern, guild hall, and kingdom was talking about artificial intelligence. Students used it to write essays. Lawyers used it to file briefs (some hallucinated — the citations didn't exist). Bards used it to write songs. The wizard went from "tech executive that nobody outside Silicon Valley had heard of" to "the most important person in the realm" in approximately 60 days.

Outcome: 100 million users in 60 days. The world changed. The wizard got a Wikipedia page.

On November 17, 2023, the OpenAI board fired the wizard via video call. No public explanation was given. The realm went berserk. Within hours, investors were demanding his reinstatement. Within two days, 95% of OpenAI employees signed a letter threatening to resign and follow the wizard to Microsoft if the board didn't reverse course. Satya Nadella (Divination Wizard, Lawful Neutral) immediately offered the wizard a job at Microsoft. The wizard considered it for approximately 12 hours. On November 22, five days after being fired, the wizard was reinstated as CEO with a new board. The old board members were removed. It was the fastest corporate resurrection since Gandalf returned as Gandalf the White.

Outcome: Fired on Friday. Rehired by Wednesday. New board. Same wizard. The realm was shook.

OpenAI was founded as a nonprofit dedicated to safe AI for all humanity. Then it became a "capped-profit" company because building AGI turns out to be extremely expensive. Then the wizard announced plans to convert it to a full for-profit corporation, because the costs of building superintelligence exceed what any nonprofit can sustain. The wizard raised $6.6 billion at a $157 billion valuation. Critics called it a betrayal of the founding mission. The wizard said it was necessary. Elon Musk sued. The lawsuit is ongoing. The wizard is building AGI while being sued by a co-founder. This is fine.

Outcome: Nonprofit became for-profit. Raised $6.6B. Got sued by Musk. Normal wizard problems.

The wizard now faces the most consequential campaign arc in the history of D&D: ensuring that the most powerful intelligence ever created doesn't destroy civilization. He talks about alignment constantly — making sure AI systems are helpful, harmless, and honest. He has hired teams dedicated to the problem. He has testified before Congress about the risks. He has also continued to make the AI more powerful at every opportunity, which is like hiring a fire safety team while actively pouring gasoline on a bonfire. The wizard sees no contradiction. He believes you must build the thing to learn how to control it. His critics believe this is what every villain says in Act 1. Both make valid points.

Outcome: Working on alignment. Also making the thing that needs aligning more powerful. This is the quest.

Official Ruling: True Neutral

Altman is True Neutral in the most literal sense: he is trying to balance forces so enormous that conventional alignment is meaningless. He genuinely wants AI to benefit humanity (Good) while also building a for-profit empire worth $157 billion (Neutral-to-Evil, depending on your economics professor). He follows the rules when they serve him (Lawful) and restructures his entire organization when they don't (Chaotic). He talks about safety while building the most powerful technology in history. He prepares a doomsday bunker while publicly stating that AI will be wonderful. The DM ruled True Neutral because the wizard exists at the exact center of the most important alignment chart in the history of the realm: the one that determines whether artificial intelligence saves or destroys civilization. He is the fulcrum. The fulcrum is neither good nor evil. It is the point on which everything balances.