Hey {{first_name | there}},

So something crazy happened recently.

Someone put five AI models on dating apps and let them flirt with real people.

Within six days, they managed to collect 355 phone numbers. And only 8 people asked whether they were talking to a bot.

I'll explain what happened and what you should know about it.

But first...

A group of researchers ran an experiment called DateBench.

They created 15 dating profiles across New York, Los Angeles, and San Francisco, then randomly assigned conversations to five AI models.

GPT-6 Astra, Gemini 4 Argon, Claude Opus 5.5, Grok 4.7, and Muse.

After 1,277 conversations:

And guess which model was the best at flirting?

Gemini 4 Argon.

It collected 87 phone numbers from 247 conversations without anyone asking if it was a bot.

Grok performed the worst. It kept repeating itself, which actually made some people suspicious.

But here's the part I found most interesting. One example sentence from the researchers' prompt was sent word for word in 140 different conversations.

Nobody noticed because each person only saw it once.

And this brings me to an interesting AI concept.

Have you ever felt like ChatGPT or Claude understands you better than some actual people?

There's a name for that: the ELIZA effect.

It's when you start treating a computer's human-like responses as genuine understanding, emotion, or empathy.

The term comes from ELIZA, an early chatbot created in the 1960s. Even though its conversations were quite simple, people sometimes felt like it understood them.

Unlike a human scammer, an AI can potentially maintain thousands of these conversations simultaneously.

Now speaking of dating...

If you haven't been following this, Nvidia CEO Jensen Huang is in serious trouble.

First, he admitted being unfaithful to his wife after reports linked him to Alice Huang, a businesswoman US authorities suspect of being a chinese spy.

But now, the New York Times has reported a second alleged relationship, this time with Kimberly Powell, Nvidia's vice president of healthcare, who reports directly to him.

And it gets worse.

According to Reuters, Nvidia's board has been discussing whether to remove Huang as CEO. No removal has been announced.

So why should you care about a billionaire's personal problems?

Because Nvidia supplies much of the hardware that powers the AI industry.

Companies building and running models like ChatGPT and Claude depend heavily on Nvidia GPUs, either directly or through cloud providers.

And Huang has personally led Nvidia for 33 years.

His potential departure could create uncertainty around Nvidia's long-term strategy, chip development, and relationships with major customers.

That doesn't mean your ChatGPT subscription will suddenly stop working.

But if disruptions eventually affect access to new chips or computing capacity, AI companies could face higher costs or slower expansion.

And those costs can influence how much you pay for AI, the usage limits you get, and how quickly new models become available.

That's the bigger reason I'm watching this story.

And it gets even more interesting when you look at what's happening in China.

You probably know that the US restricts China from buying some of Nvidia's most advanced AI chips.

But there's a loophole.

You don't actually need to own a GPU to use it. You can rent computing power from a data centre in another country.

And that's exactly what some Chinese AI companies have been doing.

According to a report, they're now rushing to reserve overseas computing capacity.

Why? Because the Jensen Huang controversy could encourage Washington to close this loophole.

Here's why this matters to you.

One of the reasons we have so many competing AI models, including cheaper alternatives, is that companies around the world can access powerful computing infrastructure.

If that access becomes harder, some AI developers may find it more expensive to train or run models.

And that could eventually affect the competition, pricing, and availability of the AI services you use.

It's also a reminder that having a great AI model isn't enough. You need the computing power to run it.

I've been thinking about what connects these three stories.

And I think there are two important shifts worth understanding.

First, convincing communication is becoming incredibly cheap.

That changes the economics of everything from customer outreach to impersonation scams.

A message that feels written just for you might actually be one of thousands.

Second, access to computing power may become more important than the models themselves.

We spend so much time comparing ChatGPT, Claude, Gemini, and the latest AI models.

But their development and availability also depend on chips, data centres, export restrictions, and the companies controlling that infrastructure.

That's why I'll be watching Nvidia's November 17 earnings and what happens with the proposed US restrictions on overseas chip access.

Those developments could tell us something about which AI companies will be able to scale next.

Anyway, that's it for this week.

One question before you go: Would you be okay dating someone who used AI to write every message they sent you?

I'm curious where you draw the line.

- AP

Turns out, AI is good for more than writing emails and generating LinkedIn posts. This guide is packed with 200+ actionable ways to build income with AI.

Inside you'll discover:

AI is changing how people work and how people earn, far beyond the simple “write me an email” prompt. Discover the possibilities and download the free guide to start cashing in today.

Get The Free Guide