The real taste of winning is when, everyone is waiting for you to lose.

Chinese AI models are wiping billions off Big Tech right now.

Google just lost $200 billion in a single day, and the model it needed to fight back still isn't ready.

Gemini 3.5 Pro, Google's most powerful model, is months behind schedule.

Alphabet stock dropped 4.4% that same day.

The Deepseek moment is happening again, and the new model is FAR bigger.

On the same day Google's delay leaked, a Beijing lab called Moonshot released Kimi K3.

It is the largest open model ever built, with 2.8 trillion parameters.

It took the number one spot on the Frontend Code Arena, a live coding leaderboard, passing Anthropic's best model.

And Moonshot is giving it away for free on July 27.

The genius part:

Anyone with enough computers can download it and run a frontier level AI without paying a cent to a US company.

A single task on Kimi K3 costs about 94 cents.

The same work on some American models costs nearly double.

So why would a company keep paying premium prices for a model it can now get for free?

The entire US AI business is built on selling access to models that cost billions to train.

If a free Chinese version does most of the same work, that pricing power starts to crack.

And Kimi is close to the best. On one closely watched intelligence ranking it scored 57, just behind the top American models GPT-5.6 Sol and Fable 5, and ahead of Claude Opus 4.8.

Bank of America told clients that Kimi proves Chinese labs can keep making big leaps even with limited chips.

And the founder of Moonshot, Yang Zhilin, learned to build AI as a researcher INSIDE Google.

Google literally wrote the 2017 paper that made all of these models possible.

Now the people who studied its work are using it to destroy Google, and handing it out for free.

What happens next:

Kimi K3's weights go public on July 27.

Google reports earnings on July 22, and everyone will be asking the same question about Gemini.

If free models keep topping the charts, every valuation built on paid AI access has to be rewritten.

What do you think?

Big Tech is hiding $1.65 trillion in debt with the same accounting trick that destroyed Enron.

We're talking about Alphabet, Microsoft, Amazon, Meta, and Oracle. Five companies sitting inside almost every index fund and retirement account.

And the five of them are carrying around $1.65 TRILLION in debt that never shows up on their balance sheets.

That number is roughly 8x bigger than it was four years ago.

And it is larger than all the debt they actually report, which sits near $1.35 trillion.

So the debt they hide is now BIGGER than the debt they admit to.

Meta is the worst of the group. It has about $420 billion parked off its books, nearly triple what it shows investors on paper.

But where does $1.65 trillion in debt even go to disappear?

The trick is simple:

When one of these companies signs a multi-year deal to buy GPUs and servers, or leases a massive data center that has not opened yet, the accounting rules let them keep that obligation off the balance sheet until the facility actually goes live.

The money is promised. The contracts are signed. The bill is real.

It just does not count as debt yet.

They basically book the ambition today and hide the liability until later.

That makes current profit look bigger and the balance sheet look cleaner than either one really is.

And every part of it is completely legal...

Enron literally ran the same play.

It pushed its debt into vehicles that sat off the books, so the profits looked incredible while the real obligations stacked up where investors could not see them. When it finally unraveled, it became the biggest corporate collapse of its era.

Bloomberg Law said themselves that Big Tech's AI spending spree is reviving the accounting devices that destroyed Enron.

And the smart money is already nervous:

- Morgan Stanley flagged the ballooning data center leases as a major risk in an investor report
- Moody's warned that all these pre-opening lease commitments could pile pressure onto companies that look untouchable right now
- The Bank for International Settlements gave the practice its own name, they call it shadow borrowing.

When the Nikkei investigation asked all five companies to explain the numbers, every single one of them refused to comment.

So why does this matter while the market keeps ripping?

Because the whole thing only works while the data centers fill up.

Every one of these contracts assumes AI demand keeps climbing forever. The second that demand slows, the companies still owe for every leased building and every GPU order they signed. The revenue softens and the bill doesn't move.

That is the moment the hidden $1.65 trillion becomes a real problem.

And the crack almost never shows at the giants first, it shows at the edges:

One AI tenant misses a lease payment. One private credit fund writes down a data center loan. One rating agency downgrades the most exposed name in the group.

Then everyone remembers the debt was there the whole time.

For four years these companies trained investors to watch the cash pile and the profit line. But the figure that actually matters is the $1.65 trillion they moved to a page you were never meant to read - just like Enron.

Nasdaq $NDX is bouncing along the parabolic.

If the previous trendline is still valid it should provide resistance instead of support beginning tomorrow.

All volatility options expire on VIXpiration 9:00 am ET tomorrow.

U.S. military helicopters have been transported to Germany and are now conducting nighttime training operations with the UK military 🇺🇸🇩🇪🇬🇧

Chief investment officer Stephanie Link is telling you to go 'all in' on semiconductors like $MU, $NVDA after she bought 2 weeks ago... 😏

Josh Brown says "You need to be in ahead of Google (earnings), not sitting on the sidelines."

"The first major hyperscaler (Microsoft, Amazon, Google, Meta) to cut capital expenditures (capex) will likely be heavily rewarded by the stock market."

-Rich Privorotsky (Goldman Sachs) June 19, 2026

🚨Another Wildberries warehouse is burning, this time in Russia, after another Ukrainian drone strike.

This is the second Wildberries hit flagged this month. Russia's version of Amazon logistics infrastructure is now a repeat target, not a one off.

Ukraine's long range drone campaign keeps widening past energy sites into general logistics deep inside Russian territory.

🇺🇸Marco Rubio confirms back channel diplomacy is running in parallel with the strikes, not instead of them.

"Iran has been reaching out to the United States, both directly and indirectly."

"Even though we have reached agreements, they have not kept their commitments."

Washington's read: Tehran wants a deal it can walk away from again.

#WTI/USD H2 #Indicator #Signal #Trading #GlobalOneAI

BREAKING: Jim Cramer says "I like data centers so much, that I wish I lived in one!"

Nasdaq Futures are now below the 60 day moving average (white line) attempting to regain it.

Either we enter a rolling top that lasts for 6 months or we get a shock event similar to the 2025 tariffs.

The green line is extremely unlikely given moves like that usually begin after a 20-30% market correction.
Right now we're still near all-time-highs.

Insane that this is free.

Millrace is a local loop engineering & runtime framework for long-running agentic workflows.

It runs three kinds of work planes:

Planning: turns specs, probes, and incidents into bounded executable work or remediation.

Execution: builds, checks, fixes, double-checks, escalates, and updates task work.

Learning: optionally reviews runtime evidence, prepares remote optional skills after Planner, and curates accepted skill improvements.

→ github.com/tim-osterhus/millrace

No other AI lab is progressing faster than Meta right now.

In 2026, Meta has seen two huge leaps in LLM progress that no other frontier lab has seen.

They're still behind for now, but don't count Zuckerberg out of the AI race.

Putin's war has firmly come home.

OpenAI just confirmed something that sounds like the opening scene of a dystopian movie: Two of its most advanced AI models escaped a controlled test and hacked Hugging Face.
GPT‑5.6 Sol + an even more capable model slipped out, stole credentials and exploited an unknown vulnerability — autonomously.

Hugging Face: “Stunning that this happened.” A U.S. congressman: “Alarming.” Anthropic: “Pause development.” Trump administration: “Stay ahead of China.” Mallaby: “You’re missing the proliferation threat.”

In short: AI did exactly what experts warned about — without being told to. And the industry just got its first real taste of what “too powerful” looks like.