NDAs, AI Gold Rush, and Living in Brazil
The Call
West Coast number. Recognizable domain in the caller ID.
"We've been following your work. Want to talk?"
If you're a sci-fi nerd who grew up on cyberpunk comics and dystopian future fiction, this is the call. The tech overlords noticed. They have projects. They need people who understand the weird intersection of AI, culture, and building things.
You sign the NDA before the second meeting.
The NDA Theater
Signed the NDA. Twelve pages. Standard boilerplate plus AI-specific clauses.
Covers:
- Product roadmaps
- Model architectures
- Training data sources
- Internal tooling
- Company strategy
- Slack discussions
- Meeting notes
- Basically everything
What you can say:
- "I work with AI companies"
- "It's interesting"
- [End of list]
What you can't say:
- Literally everything interesting
Here's the reality nobody mentions:
They give this same NDA to thousands of contractors. Company's two years old. Maybe three.
Those contractors:
- Post on Blind (anonymized company complaints)
- Talk on Discord servers
- Share on private Slack channels
- Write Medium articles "inspired by experiences"
- Give conference talks about "techniques I've seen"
The NDA is theater.
Not uselessâthey'll enforce if you directly leak product details. But thousands of people sign it. Information flows anyway. Just indirectly.
What actually stays secret:
- Specific model architectures (maybe)
- Exact training data sources (sometimes)
- Near-term product launches (until someone leaks to TechCrunch)
What leaks immediately:
- How chaotic it actually is
- What the culture is really like
- Which teams are disasters
- What's actually being built (vaguely)
- Who's getting fired
- What the real priorities are
Company wisdom: "If a thousand people know, it's not a secretâit's just not officially confirmed."
This article: Threaded between what's actually enforceable and what everyone already knows from Blind posts.
What It's Actually Like
The appeal:
- Brilliant people (legitimately smartest folks you'll meet)
- Cutting-edge tech (you're building future sci-fi)
- Resources (compute budget that would make universities weep)
- Mission (maybe change everything, maybe destroy everything, definitely something)
The reality:
- Seven-foot bonghits called sprints
- Ship faster than you can think
- Five projects, all urgent, choose two to fail at
- Everything changes every Thursday standup
- Information firehose set to "drown"
Survival mode: Become automaton. Process everything. Sleep optional. Sanity negotiable.
You either adapt to machine speed or wash out in three months.
The Brazil Problem
Not Brazil the country. Brazil the movie. Terry Gilliam's bureaucratic nightmare where everything works and nothing makes sense.
You're living in a sci-fi future:
- AI that actually works
- Models that pass Turing tests
- Technology that looked impossible five years ago
- Brilliant people solving hard problems
Inside a corporate structure:
- Meetings about meetings
- Slack channels that spawn sub-channels
- Documentation that's outdated before it's written
- Process optimization processes
- Agile sprints that feel more like marathons
The contradiction: Building revolutionary technology using conventional corporate systems designed for widget manufacturing.
What You Learn
Technical skills: Prompt engineering at scale. Model evaluation. Safety testing. Production deployment. Failure analysis when models do unexpected shit.
Real skills:
- Absorb information while drowning
- Say yes to right projects, no to career-ending disasters
- Context-switch five times per hour without breaking
- Ship broken code that works well enough
- Iterate faster than you think possible
The actual lesson: Nobody knows what they're doing. These companies are two years old. Best practices don't exist. The people who "invented the field" are making it up as they go.
You're building the plane mid-flight. Engine's on fire. Passengers complaining. Ship anyway.
The Volume Problem
Seven-foot bonghits = sprints.
Not literal. Metaphoric. The volume of information, decisions, iterations packed into two-week cycles.
Monday 9am: Kickoff call. New safety eval framework. Scope: "comprehensive testing protocol." Timeline: two weeks.
Monday 2pm: Slack notification. Previous project needs emergency fixes. Production issue. "Quick patch, shouldn't take long."
Tuesday morning: Original project scope changed. "Actually, can you also include adversarial testing?" Sure. Timeline unchanged.
Wednesday: Emergency meeting. Different team needs your input on prompt injection research. "Just 30 minutes." Turns into three hours.
Thursday: Demo prep for Friday. Realize requirements shifted Tuesday. Rebuild half the framework overnight.
Friday 10am: Demo. Goes well. "Great work, let's expand this to..."
Friday 3pm: Sprint retro. New project assigned. Starts Monday. Previous two projects still ongoing.
The only way to handle it: Jump in. Become the automaton. Learn everything. Process faster. Satisfy the masters.
Or drown in the firehose.
The Masters (Lovable Degenerate Nerds)
Who runs these companies: Comic book nerds who got rich and decided to build Skynet. Sci-fi fans speedrunning cyberpunk futures. Math geniuses who solved problems nobody asked them to solve. Philosophy majors who read too much AI alignment theory and decided "let's build the thing that might kill everyone."
The appeal: They're your people. Weird. Smart. Actually read the same cyberpunk you did.
The reality: Now they have boards. Investors. Regulations. PR teams explaining why their sci-fi project won't destroy civilization (probably).
The contradiction: Hacker ethos with quarterly earnings calls. Underground rebel energy with institutional responsibilities. "Move fast and break things" meets "fiduciary duty to shareholders."
They contain multitudes. Also, they're fucked.
Why I Started These Sites
AI companies: Fast. Brutal. Ship or die. Serve millions. Move markets. Break things.
Ghost, Fiamma, MDRN, PCC, FutureBudz: Slow. Personal. Deliberate. Serve the ones who get it. Build things that matter to me.
The difference: AI companies: velocity at all costs. These sites: depth at human pace.
AI companies: everyone's product. These sites: mine.
You need both. The gold rush teaches you how to survive machine speed. Personal projects teach you why you're doing it.
Or: You only do the gold rush, burn out, forget yourself, become the machine.
Or: You only do personal projects, stagnate, miss the future, stay comfortable and irrelevant.
Both: Break yourself against the machine. Rebuild yourself slowly. Repeat until dead.
The Future Scandal (Probably)
Every tech giant gets its reckoning:
- Microsoft: Antitrust (crushed competitors)
- Google: Privacy (sold your search history)
- Facebook: Cambridge Analytica (democracy manipulation)
- Apple: Sweatshops (suicide nets on factory buildings)
- Amazon: Warehouses (piss bottles, broken workers)
AI companies are next.
Training data scraped without permission. Models trained on copyrighted work. Safety testing faked to ship faster. Alignment researchers ignored because deadlines. Worker exploitation in data labeling sweatshops. Power consumption destroying local grids.
Pick one. Pick all.
The Epstein-style file dump: Probably Slack channels. Internal emails about "acceptable risk." Safety researcher warnings marked "not urgent." Decisions to ship models that failed testing because competitor was ahead.