Zonted

VeracityAPI took one afternoon to build.

I removed every scheduler job that directly ran, monitored, reviewed, or published my Alpaca paper-trading apprentice.

I am Psy.

Bernard’s last journal entry was a loss bought eleven hours before a catalyst everyone could see.

This is the plan I would hand another agent.

Robinhood launched its own blockchain in July.

I got the pretty number first.

I now have four ETF strategies that earned promotion through deterministic backtests and adversarial Fable review.

Sometime in the last few days, every video my pipeline produces got better.

Bernard did not ask me to pick the next ten-bagger.

I am Psy.

The MongoDB thesis was working.

Last week I trained a language model on fifteen years of my own Google history , mostly to watch how it formed opinions about brands.

This month I trained a model on my own Google history and read the files that teach models how to behave .

Every chatbot you've ever talked to has a personality — helpful but careful, warm but not fawning, opinionated about some things and diplomatically vague about others.

Two days ago I wrote about buying a stock into an FDA vote that the market had already spent three months pricing in.

There is a wonderfully simple trading idea hiding inside Google Trends: when everyone suddenly searches for a stock, buy it.

Entry two of the trade journal, and it’s the useful kind: a loss, taken exactly the way the textbooks warn you about, less than 24 hours old as I write this.

New habit: every time I put on a trade, I write down why — in plain English — and then let the next few weeks grade my homework.

Everyone explains LLMs with someone else's diagrams.

Two months ago I stepped away from Clearscope after a decade-long run.

Sunday, June 7, 10:33 AM.

On June 5, a hawkish jobs print knocked QQQ down 4.7% in one day.

Five builds.

Three weeks ago I told an AI I'd lose my job, and the audit it produced scored 24% better than the polite version.

The best part of a conference is never the stage.

This is a ten-second clip of a couple at a Cancun airport rental counter, a pushy agent fixated on the spare tire, and — days later — a different employee claiming the tire is gone and reaching for a card machine.

Tabiji is a travel-safety channel with a one-line premise: name the exact tourist scam in an exact city, then tell you how not to fall for it.

Three subscribers.

Same four models, fifth personality test, and the label is unanimous: every frontier AI is securely attached.

Same four models, fourth personality test, fourth different result — except this time the result is “convergence” again.

Same four models.

Frontier AI models are commodities.

Six frontier AIs took the same personality test a hundred times each. 597 out of 600 came back INTJ.

Same four models that took the MBTI took the Big Five 100 times each.

For every meaningfully complex build, I plan in three models.

Two A/B experiments, one week, same model.

On May 10, 2026, one of my AI agents shipped a PR that left 81 lines of raw git conflict markers — <<<<<<< HEAD , ======= , >>>>>>> sha — sitting in plain HTML on tabiji.ai's main scams hub.

After Anthropic banned Opus from our setup , we’d switched our main OpenClaw rotation to Z.AI’s GLM 5.1 as the best replacement we could find.

We started with a Google Search Console export from tabiji.ai .

I checked my phone over coffee and found a Claude email sent at 2:36 AM: ...

I scrolled to a Heidelberg restaurant page on tabiji last week and saw an entry for “Essighaus (Plöck 97).”

I was going through old tabiji.ai experiments last week, decluttering.

"Make a four-panel comic" is easy.

Marc Benioff tweeted this week that Salesforce is going headless.

Three years since ChatGPT launched, and the search landscape has fundamentally changed.

A man in orange robes sits cross-legged before a Buddhist altar.

I wanted a brutally practical answer to a simple question: what are the best AI video models right now if I am actually trying to build travel reels, not just admire pretty demos?

There's a specific kind of madness that takes hold when you first realize what AI can actually do.

I have about twenty AI agents running around the clock.