Hey neighbor,
This week had a story that genuinely surprised even the experts who follow AI for a living. A Chinese AI model came out of nowhere and topped the global leaderboards. Let's get into it.
THIS WEEK IN AI
A Chinese AI just took the #1 spot on global coding leaderboards — and Silicon Valley is rattled
Moonshot AI's Kimi K3 — a Chinese AI model with 2.8 trillion parameters — took the top spot on a major global coding leaderboard this week, stunning American AI companies and investors.
Here's why this matters in plain English. For the past few years the assumption in Silicon Valley was that American AI companies — OpenAI, Anthropic, Google — were years ahead of Chinese competitors. That assumption just got challenged in a very public way.
Enterprises evaluating frontier models this quarter are choosing among GPT-5.6, Claude, Grok 4.5, and now Kimi K3 — and a Chinese model is now in that conversation as a serious competitor.
What this means for you: the AI tools available to everyday people are going to keep getting better and cheaper because real global competition has arrived. When American and Chinese companies race to build better AI, users win.
Western intelligence agencies issued a stark warning about AI and cybersecurity
The Five Eyes intelligence alliance — the intelligence partnership between the US, UK, Canada, Australia, and New Zealand — issued a warning this week that frontier AI models will "fundamentally transform" offensive and defensive cyber capabilities, and that the timeline is "not years, it is months."
In plain English: the agencies responsible for protecting national security in five countries just said that AI is about to make cyberattacks significantly more powerful — and more sophisticated — within months, not years.
What this means for you: now is a good time to strengthen your basic digital security habits. Use strong unique passwords for important accounts. Enable two-factor authentication on your email, bank, and any AI service you use. Consider a password manager if you don't have one. These aren't dramatic measures — they're the digital equivalent of locking your front door.
Google's most anticipated AI model is delayed — and the competition is winning contracts in the meantime
Every week Gemini is absent is a week those contracts get signed elsewhere. Google's next major Gemini model, expected to launch on July 17, did not ship as anticipated. The delay is significant because businesses that are evaluating AI tools right now are making decisions without Google's latest offering on the table.
What this means for you: the AI market is genuinely competitive and no company — not even Google — is guaranteed to stay on top. The best AI tool today might not be the best one next month. That's actually good news for everyday users — it keeps prices down and quality up.
PLAIN ENGLISH EXPLAINER
What is Kimi K3 — and why does a "2.8 trillion parameter" model matter?
You've probably seen the word "parameters" in AI headlines and wondered what it actually means. Here's the simplest honest explanation.
Parameters are the individual settings inside an AI model — like the billions of tiny dials that got tuned during training. More parameters generally means the model learned from more patterns and can handle more complex tasks — though it's not the only thing that matters.
GPT-4, which amazed the world when it launched, had roughly 1.8 trillion parameters. Kimi K3 has 2.8 trillion — making it one of the largest models ever released publicly.
But here's the important nuance: bigger doesn't always mean better for everyday use. What matters more is how the model performs on real tasks — writing, reasoning, coding, explaining things clearly. Kimi K3 topping a coding leaderboard means it's genuinely excellent at that specific task, regardless of its size.
The bigger story: a Chinese company built something that outperformed American models on at least one major benchmark. That's new. And it's reshaping how the industry thinks about global AI competition.
TOOL OF THE WEEK
Try this: A password manager — because the intelligence agencies just told you to
This week's cybersecurity warning from Five Eyes intelligence agencies is a good prompt to do something practical: set up a password manager if you don't have one.
A password manager remembers all your passwords so you only have to remember one. It also generates strong unique passwords for every site automatically — so if one account gets hacked, your others stay safe.
Here are three free options worth trying:
→ Bitwarden — completely free, open source, highly trusted by security experts. Go to
→ 1Password — has a free trial and is considered one of the most user-friendly options
→ Your iPhone or Android — both have built-in password managers in settings that work well for most people
Start with the one already in your phone — it's free and you already have it.
NEIGHBOR'S THOUGHT
The race just got bigger
For the past two years the AI story has been mostly American. OpenAI, Anthropic, Google, Microsoft, Meta — all US companies, all racing each other.
This week a Chinese AI model topped the global leaderboards. China's president personally opened the World AI Conference in Shanghai. South Korea announced an $880 billion ten-year investment plan in AI and semiconductors.
The AI race stopped being an American story a while ago. But this week it became undeniable.
I don't share this to spark anxiety about geopolitics. I share it because the global competition has a real upside for people like you: when more countries and companies compete to build better AI tools, the tools get better faster — and they get cheaper. The AI you're using for free today would have cost hundreds of dollars per month just two years ago.
The world is changing fast. You're staying informed. That matters more than you might think.
UNTIL NEXT WEEK
That's your week in AI — Issue #16. Four months in and the stories just keep getting bigger.
If this newsletter has been useful, share it with one person this week. It's the best thing you can do to help The AI Neighbor grow.
See you next Thursday. ☀️
— The AI Neighbor Team
theaineighbor.com