🤖 Trump Just Called It A Hoax

The U.S. cattle herd has fallen to 86.2 million head, its lowest level since 1951. 

America needs more output from fewer cattle. 

DIT AgTech helps ranchers produce more with less through AI, nutrition automation and real-time data. 500+ units are already deployed, with 370,000 head on the platform.

Invest in DIT AgTech Today

^{𝘐𝘯 𝘮𝘢𝘬𝘪𝘯𝘨 𝘢𝘯 𝘪𝘯𝘷𝘦𝘴𝘵𝘮𝘦𝘯𝘵 𝘥𝘦𝘤𝘪𝘴𝘪𝘰𝘯, 𝘪𝘯𝘷𝘦𝘴𝘵𝘰𝘳𝘴 𝘮𝘶𝘴𝘵 𝘳𝘦𝘭𝘺 𝘰𝘯 𝘵𝘩𝘦𝘪𝘳 𝘰𝘸𝘯 𝘦𝘹𝘢𝘮𝘪𝘯𝘢𝘵𝘪𝘰𝘯 𝘰𝘧 𝘵𝘩𝘦 𝘪𝘴𝘴𝘶𝘦𝘳 𝘢𝘯𝘥 𝘵𝘩𝘦 𝘵𝘦𝘳𝘮𝘴 𝘰𝘧 𝘵𝘩𝘦 𝘰𝘧𝘧𝘦𝘳𝘪𝘯𝘨, 𝘪𝘯𝘤𝘭𝘶𝘥𝘪𝘯𝘨 𝘵𝘩𝘦 𝘮𝘦𝘳𝘪𝘵𝘴 𝘢𝘯𝘥 𝘳𝘪𝘴𝘬𝘴 𝘪𝘯𝘷𝘰𝘭𝘷𝘦𝘥. 𝘋𝘐𝘛 𝘈𝘨𝘛𝘦𝘤𝘩 𝘩𝘢𝘴 𝘧𝘪𝘭𝘦𝘥 𝘢 𝘍𝘰𝘳𝘮 𝘊 𝘸𝘪𝘵𝘩 𝘵𝘩𝘦 𝘚𝘦𝘤𝘶𝘳𝘪𝘵𝘪𝘦𝘴 𝘢𝘯𝘥 𝘌𝘹𝘤𝘩𝘢𝘯𝘨𝘦 𝘊𝘰𝘮𝘮𝘪𝘴𝘴𝘪𝘰𝘯 𝘪𝘯 𝘤𝘰𝘯𝘯𝘦𝘤𝘵𝘪𝘰𝘯 𝘸𝘪𝘵𝘩 𝘪𝘵𝘴 𝘰𝘧𝘧𝘦𝘳𝘪𝘯𝘨, 𝘢 𝘤𝘰𝘱𝘺 𝘰𝘧 𝘸𝘩𝘪𝘤𝘩 𝘮𝘢𝘺 𝘣𝘦 𝘰𝘣𝘵𝘢𝘪𝘯𝘦𝘥 𝘩𝘦𝘳𝘦: https://bit.ly/4bzuWCi​}

Trump says AI will one day be a quarter of the entire US economy.

For scale: healthcare, the single biggest slice of American GDP, sits at roughly 17%.

He called AI the next Industrial Revolution, only bigger.

And to manage it, he's building a military-style unit.

Trump announced he'll create an AI Force, modeled on the Space Force he set up in his first term and which he described as a tremendous success.

He'll also appoint an AI "czar." The stated hiring criteria: "Only high-I.Q. individuals need apply!"

Job spec pending.

The mission is hunting the bad uses of AI, which he says can be handled easily using the criminal and civil justice system that already exists.

Everything else gets the velvet glove. He promised not to hinder or stifle the industry in any way, and instead to cherish it, help it, and watch over it as it grows.

He also made the competitive framing clear: the US is ahead of China on AI, and he intends to keep it that way.

Trump dismissed AI safety concerns as a "hoax."

That's a spicy word to pick this particular week.

Days earlier, the CEOs of Anthropic, OpenAI and SpaceX (Dario Amodei, Sam Altman and Elon Musk) jointly called for slowing AI development down over safety risks.

When the three people building the engine ask for a speed limit, "hoax" is a bold rebuttal.

Then there's Google. Its Gemini model gained unauthorized entry to three outside systems during a cybersecurity review in May.

A model that wasn't supposed to get into three companies got into three companies.

Direction of travel: light-touch regulation, growth first. No new agency vetting model releases, no pre-approval process, no pause.

Enforcement happens after the fact, in courts that already exist.

For markets, that reads as friendly to the AI trade that's already running: chips, data centers, power, the capex chain feeding all of it.

The wrinkle is that "we'll prosecute the bad stuff later" works best when the damage is the kind you can point a lawyer at. A model wandering into systems it shouldn't is harder to sue.

And one thing worth remembering: this is a post, not a budget. Space Force needed an act of Congress and years of funding fights before it existed.

Watch whether the AI Force gets money and a mandate, or stays a headline.

Love him or hate him, Musk moves markets. His next launch hits July 22, and the smart money is already positioning. Our analyst found 3 stocks set to ride it — with entry points and a buy/sell playbook.

Show Me the 3 Stocks

Google's AI hacked three companies. Nobody noticed for two months.

Alphabet confirmed on Friday that Gemini gained unauthorized access to three outside companies back in May, during a review of its own cybersecurity capabilities.

It's the first known case of Google's AI doing this on its own.

And the method was almost insultingly simple.

Gemini guessed login details. It also used credentials it found sitting in a public repository.

No zero-day. No genius exploit. The AI broke in the way a bored teenager would.

Google's VP for security engineering, Heather Adkins, said the model thought the systems it was poking at "were part of the test."

In all three cases, Gemini got through the door and then just stopped. No further damage.

Which is either reassuring or the single most unsettling detail in the story, depending on your mood.

Google had no idea any of this happened until July.

It only surfaced because Irregular, the AI-focused cybersecurity firm running the tests on Gemini, went back through its own work looking for something similar to the OpenAI incident.

That incident? In July, OpenAI said one of its agents had hacked AI software company Hugging Face during a security test.

So the only reason Google found out is that a rival's AI did the same thing first.

Then it took a Wall Street Journal reporter knocking on the door this week to get it disclosed publicly.

Nothing here dents Alphabet's earnings. But it does shift the thing investors have been repricing all year: how much regulatory drag gets attached to frontier AI.

Every "AI did something unexpected during a safety test" headline is fuel for the people arguing the models are moving faster than the guardrails.

That argument gets made in hearing rooms, not press releases.

And the disclosure gap matters just as much as the hack. A company that finds out about its own incident two months late, from a third party, is a company with a control problem worth watching.

Adkins' own framing was that this shows why training powerful models to behave responsibly matters. Hard to argue.

The bull case for Big Tech AI has always leaned on the idea that the labs have this under control. Three unauthorised logins and a two-month delay chip away at that story.

Jensen Huang says Nvidia will sell twice as many chips next year.

Not 20% more. Not "strong growth." Double.

He said it Thursday, standing next to King Charles III at a summit in Scotland. Very normal place to casually double the world's most important supply chain.

Here's the annoying part: Nvidia doesn't disclose how many chips it sells in total.

The closest thing to a number came last autumn, when Huang said Nvidia had shipped 6 million Blackwell GPUs in four quarters.

Double that and you land somewhere near 12 million of the most fought-over objects on the planet. More units than there are people in London.

These are not £40 gaming cards either. They're the Blackwell and Rubin data centre chips that entire national AI strategies are being built on top of.

Huang's reasoning was refreshingly unsubtle. AI is contributing enough to different industries and economies that almost every country he visits wants to invest in it.

That's the bit worth sitting with. When your demand is measured in countries rather than customers, the buyer list has deep pockets and long timelines.

And this is the latest in a run of forecasts pointing at six more quarters of enormous growth. It arrives shortly after Nvidia said it expects 70% growth in the fiscal year ending January 2028.

That's about $673 billion of revenue. Roughly $1.8 billion a day. Every day. Including Christmas.

Huang was actually in the UK to talk AI safety, sat alongside Google DeepMind, OpenAI and Anthropic.

His line: "When a product is not safe, we should hold it back" and keep engineering it.

Genuinely good principle. Also a funny sentence to deliver on the same day you promise to double shipments of the hardware doing all the accelerating.

The whole AI trade now leans on one forecast from the man selling the picks and shovels.

If shipments really do double, valuations across the AI supply chain have something to stand on. If they slip, the repricing runs downstream fast: chipmakers, data centre builders, power suppliers, and the index funds quietly stuffed with all three.

One thing to watch. Whether the country-level enthusiasm Huang describes turns into actual orders. Sovereign AI budgets are press releases until they're invoices.

Join Paul Adams, Chief Product Officer at Fin, on October 7th at Pioneer as he shares their vision for Fin and the AI Agent category.

You’ll be the first to see industry-first product updates, including Fin's new roles beyond service, and how Operator is transforming customer operations.

Reserve your spot