I’m taking a bit of a break from casually implying (although definitely not directly suggesting) that you to go out and burn down as many data centers as possible. It turns out that makes my glow-in-the-dark handler quite twitchy! I anticipate we’ll be back to another rah rah take up arms edition as the situation develops but I want to come back to today’s reality,

which is,

that it’s increasingly obvious to me that some great populist uprising isn’t coming in this country against the Epstein class…at least today. Things are still too comfortable. Maybe tomorrow when more difficulties arrive and they will most certainly arrive…

It seems as though AI is here to stay. This week’s edition focuses on using it to your advantage.

Let’s talk about how to ride a tiger.

Don’t Fail

Two people at the same company bought the same $200 AI subscription this month.

First guy opens the chat box, rewrites an email, pastes it into Gmail, cleans up the Office Space middle-manager phrasing, and tells his boss he's "using AI now." Really he just has a marginally better autocomplete.

Second guy hooks the model into the work: inbox, docs, ticket queue. It pulls context, drafts, flags uncertainty, waits for a human yes.

Same model. Same bill. One rented a chatbot. The other built a junior that doesn't sleep, doesn't need benefits, and still needs him for judgment so he isn't slopifying everything on autopilot.

Most people I talk to still think a ChatGPT tab open means they're winning.

It doesn't. That is now what failing looks like.

Bait

Most of us use AI like a slightly enhanced Google. Ask. Get a paragraph. Close the tab. Tomorrow you're at zero again because nothing was saved, connected, or learned.

Fine. Useful. I do it daily. I pay $1000 a month and burn every token, which OpenAI has feelings about.

Chat is convenient because it's familiar. It also keeps you doing low-level work like an unpaid intern. Same card trick as last week: they point your attention where it already wants to go while the real product acts without you babysitting every keystroke.

Getting Paid

Automating the boring parts of your job is a career now. Bury your head if you want. If you don't max these tools out, someone like me will take your seat sooner than you expect.

I bookmarked a dumb-simple agent instruction this week:

/goal do my job. periodically check Slack and respond to any messages. check JIRA for new tickets assigned to me and if you see one, spawn a subagent to complete the task and open a PR. if you need clarification, reach out to people on Slack for help. NEVER tell them that you’re GPT. make no mistakes.

Fully autonomous? No. Game changing? You betcha.

I walked through how I actually do this. The edge is almost never "which model." It's the junk around the model: connectors, scheduled jobs, caches, and a human who checks output before a customer sees it.

Everyone fights about the brain on podcasts. I was raised a musician. The hands are the point. Especially when you can ephemerally deploy more of them around the clock.

Pay a genius to sit in a locked room answering trivia and you get trivia. Give that brain your inbox, files, and a log of what worked last week and you get something closer to an employee. Not perfect. A lot more than nothing.

Corporations are bullshitting you about "successful AI use." They care about optics. I care about force-multiplying output to make money.

Braindrain

Using it is not owning it. I have watched this movie. I have directed this movie.

Employee gets excellent at an internal AI workflow. Trains it. Documents edge cases. Pours twenty years of "how we actually do this" into the machine. Then "impacted." Workflow stays. Nobody notices she's gone. Judgment is company property. Asset kept. Person released.

That person does not have worth in a company anymore. Harsh? True. Worth now means AI tooling and subject matter expertise. Expertise alone is dead.

Crosshairs

Below that: people who don't use AI. The AI is used on them.

Scores the resume. Summarizes the review. Watches tickets. Predicts quit risk. Ranks raise, PIP, layoff cohort. They never see the prompt. Just the consequence, wrapped in efficiency and fairness.

Your boss doesn't need superintelligence. Your boss needs a dashboard. One green. One red. Executives trust dashboards blindly because most of them are retarded. I have administered layoffs. Sat on the spreadsheet side. The spreadsheet always sounds more objective than it is. Guilt sponge. Feeling bad and still doing what is necessary is leadership. Hiding behind a metric is cowardice. Most executives are cowards.

Two Faces

AI fight and H-1B fight look like separate culture-war topics. Same managerial move.

Don’t want to pay qualified Americans → import foreigners at half price. AI can do the work → kill the junior roles where Americans would have learned. Ladder gone. Six months later AI can't fully do the job, experienced people are gone, role reopens under a new title at a lower salary. They call it transformation. From your chair: replacement.

DOL OIG this week names fraud, kickbacks, coercion, displacement.

They will not do anything - did you watch my video on it?

  1. The Indian-born CEO of Microsoft names Indian American Asha Sharma CEO of Xbox (she has no job experience in gaming btw)

  2. She lays off thousands of American employees

  3. She opens thousands of roles for H-1B employees (primarily from India)

  4. The Federal Reserve appoints her to a task force on “productivity and jobs”

I shit you not - this all really happened. Nobody deserves to be judged by their race. At the same time it’s not some deep mystery why anti-Indian sentiment is increasing at an exponential rate. People are mad when they are deprived of their livelihood and the numbers clearly tell a story about who is replacing them. Might not be their fault but when the government does nothing to prioritize heritage Americans, people feel cornered and like help is not coming (because it’s not). They’re making America great again just not for the people that were born here (billionaires exempted of course).

An American citizen who graduates this spring with a degree in nutrition science has a 0.6% unemployment rate. The same graduate with a computer science degree has a 6.1% unemployment rate. It’s not that tech companies aren’t hiring, they’d just rather sponsor an Indian national on H-1B to fill the role at a lower salary with less employee rights than the American graduate. And yes, I statistically mean Indian national not foreign national as India typically claims about 70% the total H-1B visas each year.

Our current administration, who, according to some pundits is full of “white nationalists” paradoxically not only does nothing about it, they announce $100k H-1B visa fees which are walked back then nullified. So you’ll forgive me if I have very little hope that the Department of Labor will make any progress on this.

The Department of Justice also talked a pretty big game about this and failed to deliver, I have the letters from numerous civil rights complaints that they won’t be looking into further. The assistant attorney general, Harmeet Dhillon was born in India but I’m sure it’s unrelated.

Put people under oath and they will shit on the oath. Retarded Fox News boomers will call them patriots. The habit remains: replace apprenticeship with arbitrage, act shocked when the pipeline is empty. Workers can tolerate competition. Not a rigged market sold as progress.

Do Today

You aren't fixing American capitalism before your next 1:1. You can stop being the person who gets scored and sorted.

Pick one recurring task. Map trigger, info, decisions, output, failure modes. Minimum access. Cache stable facts. Log what it did and where it stuck. Measure something a CFO would nod at.

But I'm too busy. Fine. I consult with businesses and individuals on deploying these tools for real. Email [email protected] for a free intro call. Custom automation plan to dominate your competition. Nobody has info as good as I do for this specific problem set.

Don't walk into leadership and say "I used Claude."

Walk in and say: weekly process from four hours to 35 minutes. Response time down 60%. Three failures caught before customers saw them.

Tools churn. Results with your name travel.

If "AI transformation" is already cover for a hiring freeze, offshoring, or a weird new performance system three weeks after the all-hands, grab my White Collar Layoff Survival Manual. Checklist I wish more people had while they still had leverage.

-Dr.J

P.S. Too poor for a consultation? Let's fix that. Shoot over one specific task you're trying to automate and I might pick it up in a future video.

Keep Reading