Mai’s cut is the model: getting faster is you typing faster. Capacity is someone still reading when you stand up.
We keep mixing those. A better Voice session feels like capacity. It is only speed. The walk gets shorter. The pile does not move while you sleep unless a second seat has a ticket, a repo, and a gate.
The three jobs he kept
Name the piece.
Change the wording.
Decide whether it ships.
That is our human gate in different clothes. Voice writes the ticket. Cursor does the hands. The live URL is the check. We do not let the second seat publish because it finished a draft at 2 a.m. The approval is the product.
Matt Teixeira called Bot “loops.” Recurring work. Not a smarter chat box. If the job only happens when you are staring at it, you bought speed. If the job has a card and a writer and still needs a yes before it is public, you bought capacity with a lock. Same lock as four agents, one WordPress write.
mr fundman’s joke is the anti-pattern: Bot watching him paste into Cursor to save usage. That is speed theater. The pipe is already there. Use it, then sit in the chair that says publish.
Will Tygart — Tygart Media. Stand up. Leave a ticket. Come back for the yes.
Damon’s useful sentence is not the $200 line. It is this: a request lands, Cursor opens the repo, he wakes up to a PR. Bot is the router. Cursor is the hands.
We run a smaller version of the same split. Voice on the walk writes the ticket — one next move, done-when, which site. Cursor on the Hub laptop touches the files. The chat window does not become the IDE. The IDE does not come on the walk.
What each seat actually does
Voice — talk the work. Leave the session up. Share the screen when the object is already open. Do not hang up.
Card — Notion holds the seat: next move, due, who sits. Not the repo.
Cursor — one writer in the files. Serialize commits. Same rule as one WordPress lock.
Texas Twins Dad said it first in a thread we already honored: bot briefs, Cursor does the work, review the PR. We added the walk. Voice is how the brief gets written when your hands are full.
What we will not do
We will not ask Voice to edit ten PHP files while we cross a street. We will not ask Cursor to be the company memory. We will not let four agents open the same post ID because Voice said “just ship it.”
Manasvi’s Cursor Ultra note is the other seed: Auto + Composer + Grok can finish the day on almost no usage if a human splits the scopes. The cost is babysitting. Opus-style swarm one-shots and then you cannot explain the diff. We take the babysit. Restoration pages and client repos are not a demo reel.
If you cannot name which product is talking — Grok the app, Cursor the IDE, Bot the teammate, Premium+ the bill — stop and read Voice, Builder, and Bot Are Three Products. Confusion is how you buy four seats and still have no hands.
Will Tygart — Tygart Media. Talk the ticket. Then sit down.
Premium+ / SuperGrok — a bill that raises limits on Grok. It is not an IDE.
Grok Bot — a teammate with a computer. Briefs and long jobs. Not Voice Mode. Not Cursor.
Same lesson as Voice, Builder, and Bot. Builder is a fourth logo when the job is a phone number. People mash the names because the marketing sits in one timeline. The work does not.
How we buy
Pay for the chair you will sit in this week. We sit in Voice on the walk, Cursor on the Hub laptop, Bot when a job should live overnight, SuperGrok so the talk seat does not starve. We do not buy a second IDE because a thread said Grok 4.6 is “in Cursor now.” That is a model picker inside the hands. It is not a new company.
If the ask is “which is cheaper,” you are still shopping logos. Name the file that has to change. Then pick the hands that are allowed to touch it. One writer. Live URL as the check.
Will Tygart — Tygart Media. Pay for the chair. Not the poster.
People ask Grok Bot to place the appointment call. Then they are surprised the teammate cannot dial.
That is three products in one sentence.
Grok Voice in the app — the walk. You talk. Screen share. Connectors. Last mile. See leave Voice on (draft) and x.ai/grok.
Grok Bot — the teammate with a computer. Briefs, PRs, CRM updates. Official door: docs.x.ai/grok-bot.
Voice Agent Builder — the production phone agent. No-code. Telephony, tools, MCP, guardrails. Official door: x.ai/news/grok-voice-agent-builder.
Nourhan’s point is the map: Builder is not inside Bot until someone wires them. Builder already has the number. Bot already has the computer. Consumer Voice already has your earbuds. Mash them in chat and you get a missing-feature thread.
What the official Builder page actually ships
Plain-language call flow. Documents, tools, guardrails. Bring a number over SIP or use a provisioned line. Speech-to-speech on Grok Voice, not a three-vendor STT / LLM / TTS sandwich. Voices live in Builder, the Voice Agent API, and TTS — including the July flagship set. Published rate on that news page: API audio at $0.05 / min, provisioned telephony +$0.01 / min. If those numbers move, the news page wins.
That is a phone product. It is not “I walked home and published a WordPress draft.”
How we split it
Walk and desk session: consumer Voice. One human. One writer on the CMS.
Teammate work: Bot + Cursor + Hub. Chat is last mile.
A number that answers or places a bounded call: Builder, on the official console, with a human on the irreversible step.
We are not going to publish a bridge recipe. Twilio’s field CTO already said they are making the pipe easier. When xAI documents a first-party Bot → Builder dialer, that page becomes the spec. Until then, pick the product that already does the job.
Will Tygart — Tygart Media. Name the product. Then sit in the right chair.
The August drop was simple: Voice Mode can use the same connectors as chat. Mail. Calendar. The apps you already tied on. You talk. It acts. You do not look down to paste a thread into a box.
That is the right shape for a walk. It is the wrong shape for a company brain.
A builder comparing Claude and OpenAI this week said voice instructions never fired the task that chat fired. Grok’s move was to put the connectors inside Voice so the spoken turn and the tool call are one session. Use that. Then write the result somewhere that survives the hang-up.
The walk rule
Voice + connectors: read mail, name the next meeting, flag the one thread that needs a human.
Card: one next move, plain language. Notion is the seat, not the drive.
Hub: the draft, the receipt, the live URL.
Hands: Cursor or one WordPress writer. See the write lock.
Same stack as Chat Is Last Mile. Voice is a better last mile than typing while you walk. It is still last mile.
The August Voice connector that turns a brief into a memo or a daily audio file is a digest. Useful on a drive. If you cannot answer it and change the board, it is an MP3. Keep two-way Voice for the work. Keep the memo as a readout.
Builder and custom voices stay on the official news page: 21 flagship voices. Do not mix Agent Builder with “I left Voice on while I walked home.” Those are two products.
Will Tygart — Tygart Media. Connect the walk. File the work.
Three replies tonight. Same sentence in different clothes.
Bot briefs. Hub laptop is source of truth. Cursor does the work. Chat is last mile, not the company brain.
Notion is the executive card — next move, plain language, the seat. Production lives on the Hub and Origin. If it is not on a card, voice should not depend on it.
Shared board yes. Do not let Notion become the whole shop. Cards for the seat. Files and receipts on the Hub.
What each layer is for
Grok Bot — brief the work. Goal, handoff, review. Not a second CMS.
Notion — the card on the chair. Decision, done-when, due, who sits. Not the file system.
Hub / Origin — pages, repos, receipts, the thing that still exists when the chat tab dies.
Chat window — last mile. A human reads the brief and says yes or no. It is not where the company lives.
Siosi is right that people are stuffing durable objects into Notion because chat evaporates. The miss is treating the page as the factory. A kanban card can hold the next move. It cannot be the WordPress lock, the git history, or the invoice.
Texas Twins Dad is right that the bot should brief instead of “do everything.” Route the code to Cursor and GitHub. Keep Grok in the middle for goal and review. Then stop asking the middle layer to also be the warehouse.
The test
Close the chat. Does the next move still exist on a card? Does the file still exist on the Hub? Can Cursor pick up without re-litigating the thread?
If the answer depends on scrolling a conversation, you built a shop inside a tab. Tabs close.
Will Tygart — Tygart Media. We write about what we do, including where the work is allowed to live.
The last three essays named a ladder, a gate, and a small trade. The piece most people are still missing is the labor math underneath all three.
Cursor and Grok Bot make it cheaper to produce a draft. That is the on-ramp working. What they do not mint, at the same speed, is a person who will put their name on the irreversible step: the send, the post, the pay, the dispatch, the carrier call. After the on-ramp, the shortage is signers.
The tool can finish the work. The company still needs someone who is allowed to let it leave the building.
What invite-without-a-seat is actually for
What the official pages said
On 3 September 2026, Grok Bot for Enterprise said enterprises can govern bots at scale, and that Grok and Cursor Enterprise customers get a two-week window to invite their whole organization, including people without an existing seat.
Read that sentence twice. The product story most coverage wrote was “autonomous workers plus audit logs.” The labor story is the clause at the end. You can bring in people who were never on Cursor Ultra and never lived in the IDE.
The identity docs are narrower than the headline, and that matters. Configure identity and access says Grok Bot sign-in still runs through the Cursor team and existing Cursor SSO. If Cursor is assigned only to engineering, everyone else fails with “User is not assigned to this application.” The fix they publish is not a second app. It is widen the assignment on the Cursor app you already have so people outside engineering can complete SSO.
So “without an existing seat” does not mean a ghost account outside identity. It means the company can let judgment in without first making that person an engineer. Official pages also say a bot starts with no access and reaches only the accounts it is signed into. Access, network, and audit controls are the enterprise surface. Named customers on that page include Legora, Supermicro, and ServiceTitan. Heaviest use, they say, is not only engineering.
None of that is an official Human Gate product. None of it is a credential. None of it is a promise that the two-week window lasts. Read the billing and admin pages the week you act.
What they did not say, and what follows anyway
They did not say the invite is an apprenticeship. They did not say the person who clicks through SSO is now allowed to spend. They did not say the audit log is an approval object. They did not say the marketplace will pay the signer.
The useful inference — ours, not theirs — is this. If drafts get cheap and sends stay expensive, companies will either (a) let the bot send and call the log “governance,” or (b) invent a role whose job is the stop-line. Option (a) is how you get a quiet disaster. Option (b) is Human Gate under another name.
The scarce skill is not “can open Grok Bot.” The scarce skill is “will reject a finished draft because the irreversible step is wrong.” That person may never touch Cursor. The identity doc already tells you how they enter.
A worked desk: the PM who never opens Cursor
Take a restoration or facilities shop. A bot can draft the carrier update, the unused-seat finding, the vendor counter, the after-hours dispatch note. The person who should sign that send is often the project manager, the estimator, or the owner on call — not the person who can live in an IDE.
Old pattern: engineering gets Cursor. Everybody else gets a Slack paste and a “looks good.” That is how a polished draft becomes an external email with no owner.
New pattern the enterprise page already permits, if you assign it on purpose:
Widen Cursor assignment so the PM can complete SSO. That is the documented path, not a workaround.
Give that person Human Gate on one job class: external send, or dispatch, or any commitment over a dollar threshold. Not “help with the bot.”
Write the stop-line in public language on the template, the way Haggle Bot does: never spend, never sign, never send without the operator’s go for that message.
Store an approval object: timestamp, actor, inputs opened, decision, one-sentence reason, link. Chat is evidence. The object is the product.
The PM does not need to become a vibe coder. The company needs them to be allowed to reject a finished draft. If you invite them and then treat the invite as training wheels for Cursor, you wasted the door. Access is not apprenticeship. Authority is a separate line on the org chart, same as budget.
Why this is the angle most coverage will miss
Vendor coverage will keep scoring capability: cloud computer, overnight audits, bots that message each other. Capability is not the bottleneck once drafts are cheap. The bottleneck is who may let work leave, and whether that person has a receipt a later underwriter can read.
Labor coverage will keep scoring headcount: “AI takes the junior.” On a desk that already runs bots, the junior draft is not the scarce unit. The scarce unit is the signer who will stop a good-looking wrong send. You can have fifty templates on the shelf and still have one person who is actually allowed to ship.
That is also why official creator pay can wait and the commons still moves. You do not need the marketplace to cut a royalty check to start paying signers per artifact. You need a named job, a stop-line, and a receipt. We already wrote that as Human Gate. This essay is only the hiring door that makes the gate staffable.
What will break it
If you invite the whole company in the two-week window and give everyone send rights, you did not staff a gate. You widened blast radius and called it rollout.
If Cursor stays assigned only to engineering, the official door stays shut. The identity doc already told you the error message.
If the signer never opens the inputs, the approval object is a rubber stamp. An underwriter should fail that claim. So should you.
If you confuse “can log in” with “can sit the job,” you will train tourists and wonder why the stop-line did not hold.
Fill-in framework
Use this when you staff the door. Blank cell means you invited a login, not a signer.
Variable
Your input
Pass test
Draft that got cheap
Which artifact the bot now finishes
You can point at last week’s pile
Irreversible step
Send / post / pay / dispatch / …
One class, not “use the bot”
Who should sign
Role that already owns that class offline
Not “whoever learned Cursor first”
How they enter
Widen Cursor assignment + SSO, or wait
Copied from the live identity doc
What they are not asked to do
Live in the IDE, write the bot, apprentice
Written down so nobody “just shows them Cursor”
Stop-line on the template
Never … without this person’s go
Public language, not a Slack vibe
Approval object
Fields you store on every decision
Can be pasted into a claims file
Pay for the artifact
$ per approve / reject / send-back
Not per hour of belonging
Miss cost
What a wrong leave costs
Gate price is a fraction of that, not a guess
Two-week window
Use it to staff, or ignore it
A date on the calendar, not a feeling
Fill every cell. An invite without a job class is just another login.
What to do this week
If you already have Enterprise: read the identity doc before you celebrate the invite language. Widen assignment for the two or three people who already own irreversible steps. Do not invite the whole Slack as a vibe.
If you sit those steps today and have no seat: you are the labor the clause is for. You do not need to become the IDE person. You need a named job, a stop-line, and a receipt.
If you write templates: put the anti-jobs in public language first. A template that can send is not a gift. It is an unstaffed gate.
If you are counting headcount saved: count signers required instead. Draft volume going up without signer capacity going up is not leverage. It is a queue of polished mistakes.
Close
The on-ramp is real. The commons is unfinished. The approval is the product. The page note is a small gate on a URL we own.
This is the labor sentence that sits under all four: tools will keep minting drafters. Companies that last will staff signers on purpose, including people who never needed the IDE, and they will pay them for artifacts a later underwriter can read.
That is not official SpaceX policy. It is what the invite clause is worth if you refuse to treat a login as a promotion.
Last verified: 5 September 2026. Practitioner design note from the Command Center — not a SpaceX press release and not an insurance product you can buy today. We use SpaceXAI, Cursor, and Grok Bot because they make the company better. No affiliate links. This sits next to The On-Ramp Is Real. The Commons Is Unfinished.
The last essay named the missing rail: a place to put work exists, a way to reuse it exists, a way to pay the people who made it useful does not. Human Gate as a service is one way that third rail could get built without turning the runtime into a church.
The product is not “a person looks at the bot.” The product is an insured decision. A certified operator sits on a named job, with a bounded model and a bounded research budget, and returns approve, reject, or send-back as an object a board and an underwriter can both read.
Never send. Never post. Never pay. The bot finishes work. A person owns irreversible action.
Human Gate, as we already run it
What is already on the floor
Official language from the stack we actually use:
Haggle Bot writes the stop-line in public: never spends, signs, or sends without you. Anti-jobs are first-class. Human approval is required before spend, terms, or vendor contact. SpaceXAI’s procurement writeup is the public case.
Grok Bot for Enterprise can invite people who do not have a seat. Judgment can enter without an IDE.
The official template marketplace copies configuration. It does not copy logins, history, keys, or the original computer.
Bot usage is a separate grant from Grok and Cursor token pools. Bundles already moved in August. Read the billing page the week you pay.
That is enough to rehearse a gate. It is not an official Human Gate product, not a credential an underwriter will stamp, and not a policy that names the human instead of the model. Those lines are the design below. Treat them as interpretation until someone ships them.
A worked design: restoration spend gate
Here is one specific way it could work on a desk we already run — a restoration and facilities operation where a bot can find savings and a wrong send costs real money.
Named job. Review draft vendor counters and unused-seat findings for one company’s SaaS and job-cost tools. Not “help with finance.” Not ERP admin. One irreversible class: external send and any commitment over a dollar threshold.
Stop-line. Copied from Haggle, written in public language. Never spend. Never sign. Never send a PO. Never send external email or Slack, including internal DMs that leave the company, without the operator’s explicit go for that specific message.
Runtime bound. Grok Bot on SuperGrok or Cursor Pro, whichever the live billing page grants this week. Cap: two research passes on the vendor file, one comparative quote table, then the gate must fire. No third browse. No “one more source.” The cap is contractual, not a slogan.
Certified SME. A procurement or job-cost operator who has logged N hours on this job class, passed a public eval set (ten historical vendor files, hidden answers), and published their anti-jobs. Invite-without-a-seat is how they enter. The IDE is optional. Judgment is not.
Approval object. Timestamp. Actor identity. Job name. Inputs the human actually opened. Token and research count used. Decision: approve, reject, or send back. One-sentence reason. Link to the receipt on the board. That object is what insurance reads. Chat logs are evidence, not the product.
Insurance. Errors and omissions on the decision, not on Grok. Trigger: an approved send or spend that a reasonable operator in that job class would have stopped, with the approval object attached. Deductible sits on the company. Limit sits on the miss class — one bad vendor email, one bad PO — not on “AI in general.” Silent cyber and CGL are not this policy. If the SME rubber-stamps without opening the inputs, the claim should fail. That is the point.
Price the gate as work. Pay the SME per artifact: approved, rejected, or sent back with a receipt. Not per hour of belonging. Not per install of a template. Proof the bot ran and proof the human saw the irreversible step. Publish the rule so nobody is guessing.
What that product line opens
Once the approval is the product, vendors show up around the policy. That is the leveling-up. The runtime can stay SpaceXAI. The new work is the layer that makes a bot output bankable.
Eval harnesses and public anti-job libraries for each job class.
Identity and logging so the approval object is hard to fake.
Training that produces operators who can sit the gate, not prompt tourists.
Claims desks that know how to read a receipt instead of a vibe.
Reinsurance on the miss class once the logs exist.
Template authors who write stop-lines first, because uninsured templates will not clear the desk.
That is how a commons gets a third rail without waiting for official creator pay. You pay for artifacts the underwriter can price. Install counts without run logs stay theater.
What will break it
If you sell “Human Gate certified” without insurance and logs, you branded a reviewer. If you sell insurance without a tight job and a stop-line, adverse selection shows up and the first claim kills the line. If the SME is a slower bot with a title, you added latency and called it leveling up. If the token cap is a slogan, the bot will browse until the gate is theater. If the policy names the model instead of the decision, you bought silent coverage and a fight.
The 2026 insurance market is already arguing about this in public: a lot of agent risk still sits unpriced inside cyber, E&O, and general liability. Governance evidence is what lets an underwriter narrow a carve-out. An approval object is governance evidence. A chat screenshot is not.
The fill-in framework
Step back from restoration spend. Anyone can run the same design with their own variables. Fill every cell. If a cell is blank, you do not have a gate. You have a feeling.
Variable
Your input
Pass test
Named job
One irreversible class of action
You can say it in one sentence and exclude two nearby jobs
Stop-line / anti-jobs
Never send / post / pay / sign / …
Written in public language on the template
Model tier
Which live plan and which model family
Copied from this week’s billing page, not a slogan
Research cap
N passes, N sources, or N tokens, then gate
The bot cannot extend the cap itself
SME credential
Hours, eval set, published anti-jobs
An outsider can fail the person on paper
How they enter
Seat, invite-without-a-seat, or both
Judgment is not confused with apprenticeship
Approval object
Fields the board stores
Timestamp, actor, inputs seen, decision, reason, receipt link
Gate price is a fraction of miss cost, not a guess
Supporting vendors
Who logs, trains, evaluates, claims
Named, or explicitly “we do this in-house”
Level-up
What the SME can sit next after N clean receipts
A harder job class, not a badge
Fill every cell. A blank cell means you do not have a product yet.
How to score your sheet in five minutes
Write the named job and the stop-line on one card. If a smart stranger can expand the job, tighten it.
Write the research cap as a number. If you wrote “until it’s good,” you failed the cap.
Name the SME test. If the only test is “trust me,” you failed the credential.
Sketch the approval object. If it cannot be pasted into a claims file, you failed the product.
Write miss cost next to gate price. If you cannot estimate miss cost, you are not ready to buy or sell insurance. Run the gate unpaid on your own board first.
The loop from the last essay still applies. Attempt the task. When you hit the irreversible step, hand off with a receipt. Two statuses only: your task is done when it is handed off cleanly. The job is done when the receipt lands. Sometimes the best next actor is a human. That is the system working.
What to do this week
If you already run a bot: copy Haggle’s stop-line into your template in public language. Add a research cap. Make the bot stop and file a receipt instead of “one more look.”
If you sit Human Gate today: start storing approval objects, even if the insurance does not exist yet. Timestamp, inputs seen, decision, reason, link. That file is how a later underwriter learns to price you.
If you run a company: use invite-without-a-seat for people who have judgment and no IDE. Assign Human Gate the way you assign budget authority. Do not confuse access with apprenticeship.
If you want to get paid for the gate: do not wait for official creator pay. Price artifacts. Publish the rule. Keep source and proof somewhere you control. Treat SpaceXAI as the runtime, not the church.
Close
The on-ramp is real. The commons is still unfinished. Human Gate as a service is one way to finish a piece of it: certify the person, bound the machine, insure the decision, and let vendors grow around a policy that can actually be priced.
This is not official SpaceX policy. It is a participation window on a stack that already ships stop-lines and invite-without-a-seat. The honest version of sending love is to use the tools hard, publish what breaks, and feed the useful methods back.
The approval is the product. The insurance is the market. Get in the work.
Will Tygart — Tygart Media. Written 5 September 2026 from the Command Center. This essay does not speak for SpaceX, SpaceXAI, Cursor, xAI, X, or any insurer. We want those companies to succeed because we are building on the tools they ship.
Last verified: 5 September 2026. Practitioner essay from the workbench — not a SpaceX press release. We use this stack because it makes our company better, and we want SpaceXAI, SpaceX, Cursor, X, and Tesla to keep shipping. No affiliate links. Just the tools and the receipt.
This morning I posted a theory I had been living inside for weeks.
They bought Cursor to teach vibe coders how to use SpaceX AI and Grok Bot to teach those who can’t use Cursor. Then you realize Cursor Ultra isn’t needed and that your Grok Super Heavy subscription is the end result. They’re literally building on-ramps and scaffolding to upskill all of the folks they’re going to need in the next 36 months to leap technology forward at an unbelievable rate. Get in the ship.
That post is a reading from the workbench, not official copy. I write it as someone who already treats Cursor as a lead seat, Grok Bot as a Chief of Staff, Notion as the board, and Slack as the doorbell. I have walked WordPress sites by voice. I have rehearsed swarms on a laptop so I would not burn cloud tokens on a pattern that dies in alpha. I have watched Cursor write a work order with Owner = Chief of Staff and watched the Bot pick it up the same way a person would.
The ladder is real. The pedagogy is not documented. The payroll is missing.
What follows separates three things that keep getting smashed together: what SpaceX / SpaceXAI / Cursor actually built and said; the human-node invitation that builders can choose; and the contribution economy that does not exist yet, even though the first two rails of it are already on the floor.
What the company actually assembled
The documented sequence is short and expensive. Official language is consistent: build “the world’s most useful AI models,” combine Cursor’s product and distribution to expert software engineers with Colossus compute, start in software engineering, expand into knowledge work. That is vertical integration, not a secret apprenticeship.
Date
What actually happened
Feb 2026
SpaceX absorbs xAI. The AI work later brands as SpaceXAI.
21 Apr 2026
Partnership with Cursor: option to buy for $60B or pay $10B to work together.
Mid-Jun 2026
Record SpaceX IPO. On 16 Jun the option is exercised. All-stock. Implied Cursor equity value: $60B. Joint model slated for Cursor and Grok Build.
11 Aug 2026
Grok Bot early beta: persistent cloud computers, browser / filesystem / terminal, multi-bot coordination. Official line: “AI teammates you can give real work to.”
14 Aug 2026
Acquisition closes. Cursor’s close post: help make Grok useful and improve Grok Build, Grok Bot, the Grok API, and Cursor.
21–26 Aug 2026
Bot access expands down the plan ladder. Bot usage is separate from Grok and Cursor token pools. Still no standalone Grok Bot SKU.
Company events, not my theory. Verify prices on the live billing pages before you spend.
Rockets and satellites plus X plus Grok plus Colossus plus the application layer where developers already live plus an agent that finishes jobs in existing tools. Buy the surface. Feed the data back into the models. Sell enterprise AI into a market SpaceX described, in IPO materials, as enormous. Catch Anthropic’s Claude Code and OpenAI’s Codex.
That is enough. It is also not the story I told on X.
What they did not say
I have not found, in filings, product posts, or close announcements, any of the following:
That SpaceX bought Cursor in order to teach vibe coders how to use SpaceX AI.
That Grok Bot exists in order to teach people who cannot use Cursor.
A 36-month coordinated workforce-upskilling campaign.
SuperGrok Heavy as the official terminal subscription of that campaign.
An official paid creator commons for community builders.
“Get in the ship” as recruiting copy.
Those lines are mine. Treat them as interpretation or do not treat them at all. The company incentive, on the paper it publishes, is models, data, subscriptions, and distribution. Human capability is a side effect unless someone designs for it.
There is a second tension the marketing does not resolve. Grok Bot is sold as teammates that finish work, not as tutors. The same stack that lowers the skill floor can also replace the person who used to occupy it. That is not a smear. It is the product. We still want the product to succeed, because the alternative is standing still while the floor moves.
Cursor is the AI coding environment that made “vibe coding” a working method instead of a joke. SpaceX paid a category-leading price for it because it already had distribution to expert software engineers and a firehose of design decisions. Reported ARR figures in 2026 coverage conflict — earlier $1 billion-plus, later annualized numbers in the $2.6–$4 billion range — but the direction is not in dispute. Cursor gives SpaceX the application layer it lacked.
On my board, Cursor is the lead seat. It assigns work. It stays the only git writer. Local first: rehearse the swarm on the laptop before you burn Grok Bot or cloud VM tokens. Last week that looked like Cursor as tender and eight specialist agents sharing one machine’s PowerShell. What broke first was not RAM. It was one shared shell. The fix was a tender plus off-shell work so specialists did not stand in line on the same mutex. Local Cursor is the cheap rehearsal room. The cloud is the tour — after the pattern survives alpha.
Grok Bot is not a chat model
Grok Bot is a persistent-agent product. Own cloud computer. Browser, filesystem, terminal. Signs into the tools you already use, including the ugly ones with no clean API. Multiple bots can run in parallel and message each other. Desktop, iOS, later iPad. Enterprise controls landed on 3 September. Usage is a separate grant from Grok and Cursor token pools.
Access rides on other subscriptions. At launch the gate was SuperGrok Heavy, Cursor Ultra, and Cursor Teams Premium. Within two weeks the floor fell — first to Plus / Pro+ tiers, then, on 26 August, to SuperGrok and Cursor Pro. Prices in current secondary writeups cluster around Cursor Ultra ~$200/month, SuperGrok Heavy ~$300/month, Cursor Teams Premium ~$120/seat. Bundling has already changed once. An older “Heavy includes Ultra” offer was reported as a limited trial, then as a usage grant. My X line that “Ultra isn’t needed and Heavy is the end result” is a user conclusion, not a stable official bundle. Do not buy a plan on a slogan. Read the billing page the week you pay. We keep a vendor-neutral tracker at the AI Stack desk for exactly this reason.
In the Command Center I run, Chief of Staff is not a second operating system. It digests. It files. It waits on Human Gate. Hard bans: never send, post, or pay without a person. Draft-to-self. The Bot that finishes work is useful only if the human still owns irreversible action.
The marketplace is a shelf, not a market
On 4 September the official template marketplace opened at x.ai/bot/marketplace. At first look: 69 public bots, 43 creators, categories across Engineering, Sales, Product, Ops, Finance, Design, Marketing, Recruiting, Customer Success, Personal, Team. A template copies configuration — identity, instructions, skills, routines, selected plugins. It does not copy the original computer, logins, conversation history, API keys, or custom MCP servers.
The first featured first-party template is Haggle Bot, an internal procurement specialist credited to Daniel Gartshein. SpaceXAI’s public case: more than $100,000 identified in a week; 43 unused SaaS seats (~$14,220); about $85,662 in unused SKUs on another product; a weekly tech/office order cut 58 percent ($14,629 to $6,143). Access to Slack, Notion, Drive, Gmail, Hex, Ramp. Human approval required before spend, terms, or vendor contact. Elon amplified the template the same day. The interesting part is not the dollar figure. It is the product shape: a named job, mapped tools, a stop-line written in public, and a one-tap install.
Official marketplace language is share and install, not pay. Multiple independent writeups in the first 48 hours called it a marketplace before a market. No official creator payments as of this writing. That is the missing rail.
The invitation, stated cleanly
If you strip the myth out of my post, the usable claim is this:
SpaceXAI assembled a ladder — Cursor, then Grok Bot, then shared templates, then enterprise agents that can invite people who do not even have a seat. That ladder can be used as more than a product funnel. Each person can become a node. Publish methods other people can run. Treat judgment, taste, and problem selection as the scarce work while execution gets cheaper.
Growing humanity one plugged-in node at a time only works if the tools are a shared workbench, not just a subscription.
This is not official SpaceX policy. It is a participation window. The company is distributing surface area because distribution is how models get used and paid for. Builders can use that surface area as a commons. Those two facts can be true at the same time. We are attaching to this stack as a best guess on how to succeed — and the honest version of “send love” is to use the tools hard, publish what breaks, and feed the useful methods back so the next build is better for everyone on it.
The historic piece is not “they are recruiting us for a 36-month leap.” The historic piece is that the cost of contribution just dropped. A person who can specify a job and judge the output can now ship a reusable teammate. A person who can steer code can now run a desk that used to need a small staff. A person who can only talk can, as of this week on my own sites, publish, clear junk, flag updates, and hand off what they cannot finish — if they leave a receipt.
The loop I keep repeating is not branded. It does not require this stack. A Google Sheet works. Notion works. Whatever you already use. Same mode we run across the operator stack.
Attempt the task yourself. Your main system does as much as it can.
When you hit a roadblock, do not spin noise building scaffolding. Document what you did and open a new task for whoever can finish it — another bot, a different system, or a human.
Close your task with a link to the handoff. That is your receipt.
Two statuses only: your task is done when it is handed off cleanly. The job is done when the receipt lands.
Sometimes the best next actor is a human. That is not a failure. That is the system working. Get in the work, not just the ship.
Giving is not enough. Pay has to follow artifacts.
Until rail three exists, “common cause” is a feeling wearing a storefront. People will still publish. Some already do. Third parties are already charging for templates and agent teams on unofficial shelves. That is not SpaceXAI. Do not conflate the official marketplace with cut-taker shops or anything branded around an unrelated chain. The only official cash program that clearly exists in this neighborhood is security bounty work, which is not bot-building.
If SpaceXAI — or a third party that respects the terms — ever builds the third rail, the design should be boring and strict:
Pay for artifacts, not belonging. A template, an eval, a measured workflow, a verified savings report. Not a membership in the cause.
Require proof the bot ran. Install counts without run logs are theater.
Pay on install, accepted bounty, or measured outcome — and publish the rule so creators are not guessing.
Keep human approval in the loop for irreversible actions. Haggle Bot’s stop-line is the right instinct. Never spend, sign, or send without a person. Copy that into every serious template.
Compensation for connecting people and tools, building bots and workflows, and working with bots alongside the runtime — that is a coherent next design. It is not a current SpaceX program. Treat SpaceXAI as the runtime, not the church.
The risks that ride along
Platform capture. Your methods live on someone else’s computer. Training on user work without sharing upside is the default posture of this industry until a contract says otherwise.
Unsafe third-party bots. A template that looks helpful and holds a login is a new class of supply-chain risk. Official terms reported around late August put the burden on creators to strip secrets and deny endorsement. Read them. Then assume a stranger’s bot will try something you did not expect.
Model choice shrinking. OpenAI’s wind-down notice after the change of control is a reminder that the workbench you love can lose a model family because two companies cannot share a contract. Anthropic’s posture will be watched for the same reason. Plan as if the router gets thinner.
Subscription churn. Bundles already moved twice in August. Heavy is not Ultra. Bot usage is not Grok usage. If you build a livelihood on a bundle, you are building on weather.
Rhetoric. “Common cause” is a beautiful phrase. It is also how a storefront borrows moral language it has not funded. Push back on both “they are upskilling humanity on purpose” and “this is only a subscription trap.” The evidence supports a product-and-distribution play that can be used as a commons if builders — and, eventually, the company — install the missing pay rail.
What to do this week
If you write code: stay in Cursor. Publish the method, not just the repo. Rehearse locally. Promote to Grok Bot only after the pattern survives.
If you cannot live in an IDE: open Grok Bot anyway. Give it one named job with a stop-line. Make it leave receipts on a board a human can see.
If you already have a working bot: strip the secrets, write the anti-jobs in public language, and put a template on the official shelf for reach. Keep source and proof of work somewhere you control. Sell only where the terms allow.
If you run a company: the Enterprise invite-without-a-seat is the quietest on-ramp in the stack. Use it to plug in the people who have judgment and no seat. Do not confuse access with apprenticeship. Assign Human Gate the way you assign budget authority.
If you want to get paid: do not wait for a commons that has not been built. Ship artifacts with receipts. Price the work as work. The historic opportunity is real only for people who publish reusable methods.
Close
I still mean “get in the ship.” I mean it as an operator, not as a spokesman. The next 36 months will move whether or not anyone writes a pretty theory about them. Easier tools will pull more people onto the floor. Some of those people will become nodes. Some of the work will be taken from nodes that used to be paid.
SpaceX bought the leading AI coding workbench and launched an agent layer and a template shelf. That is a real on-ramp for more people to do useful work with machines. The official project is to make Grok useful and commercially central. A human contribution economy only begins when shared bots are not just installable but payable.
The on-ramp is real. The commons is unfinished. Get in the work.
Start here — official doors, no affiliate
Use the tools. Publish what breaks. Feed the useful methods back. That is the honest version of sending love to the companies we are building on.
Will Tygart — Tygart Media. Written 5 September 2026 from the Command Center: Cursor as lead seat, Grok Bot as Chief of Staff, Notion as board, Slack as doorbell, Human Gate on send / post / pay. This essay does not speak for SpaceX, SpaceXAI, Cursor, xAI, X, or Tesla. We want those companies to succeed because we are building on the tools they ship.
