How AI Gave a Dying Dog her Life Back

AI news, leaders, business insights and more

Dr. Ayesha Khanna
March 17, 2026 • Reading Time: 7 minutes

In partnership with

Hi! Here are today’s top AI stories:

How AI Gave a Dying Dog her Life Back
Let the Bots Cook 🤖
AI Around the World

NEWS YOU CAN’T MISS

How AI Gave a Dying Dog her Life Back

Image: Paul Conyngham.

A man had no biomedical degree. His dog had terminal cancer. He had ChatGPT.

When Australian AI entrepreneur Paul Steven Conyngham learned his rescue dog had mast cell cancer, and that chemo wasn't working, he refused to accept the prognosis.

He decided to build a solution. Here's what happened in under two months:

He used ChatGPT to research treatment options. It pointed him toward immunotherapy and genomic sequencing.
He sequenced Rosie's tumor DNA at the UNSW Ramaciotti Centre for Genomics.
He used AlphaFold to model the 3D protein structures of her tumor mutations.
He partnered with the UNSW RNA Institute to manufacture a custom mRNA vaccine.

A University of Queensland vet with ethics approval administered it in December 2025. One month later, Rosie's tumor had shrunk by 75%.

As UNSW computational biologist Martin Smith put it, seeing the results was a “holy crap, it worked” moment.

Scientists describe it as the first personalised cancer vaccine designed for a dog, a signal of what’s now potentially possible ... one day even for humans.

She's not cured. But she's alive, and her dad is already sequencing the tumors that didn't respond to find out why.

🐕 Her name is Rosie. She's a Staffy-Shar Pei cross. I was so happy to read about her regaining her health! ♥️

Let the Bots Cook 🤖

Most of social media is full of humans posting selfies, hot takes, and the occasional argument in the comments. But what if the users weren’t people at all?

Meta just bought Moltbook, a Reddit‑like social network built for AI agents. Here’s what the news keeps missing.

Within days of Moltbook getting attention, bots were already inventing religions, posting anti-human manifestos, and basically starting digital cults. Not exactly the onboarding flow critics had in mind. 😶

But when researchers analysed almost 300,000 Moltbook posts, they found that nearly 55% of "agents" were human-influenced. Not one viral scare originated from an autonomous bot.

Meanwhile, the more interesting implication barely received attention. 🤔

As Andrew Ng has discussed in his writing on AI systems, networks of agents can become powerful when they generate, critique, and refine each other's work. One agent drafts. Another reviews. A third improves it. The system iterates, documents reasoning, and raises the quality of outputs.

That's the other side of AI agents "chatting" with each other. Not collusion. Peer review.

Of course, that only works with the right governance. A social network of autonomous agents without oversight would create real risks.

But the opposite risk also exists: shutting down agent-to-agent communication entirely could prevent systems that significantly accelerate research, engineering, and knowledge creation.

I'll end with Andrew Ng's comment: "Social sharing isn’t only for humans. It’s also for agents!"

ps. "Let them cook" is a popular internet slang term that essentially means "give them the space to do their thing."

Do you think it's OK to let AI agents socialize?

AI Around the World

🇺🇸 Elon Musk’s xAI is undergoing a major overhaul with several co-founders and senior engineers leaving, as the company rebuilds its team and technology to better compete with OpenAI and Anthropic in AI coding tools.

🇪🇸 Spain launched an AI tool called HODIO that scans social media to measure and track how much hate speech spreads on different platforms, helping the government monitor and hold companies accountable.

🇦🇺 Australia’s financial regulator has warned Gen Z investors about relying on AI chatbots and social media “finfluencers” for financial advice, after a new study found 23% of young Australians now own cryptocurrency.

FROM OUR ADVERTISERS

The Architecture Behind AI-Native Revenue Automation

Most “AI finance” tools guess. Finance can’t. This white paper explains how AI-native revenue automation combines reasoning, deterministic math, and commercial context to automate billing, cash, and close—without sacrificing accuracy. Read the architecture behind AI-native revenue automation.

Get the whitepaper

Your AI is resolving tickets. Is it keeping customers?

Resolution rates look great. But Gladly's 2026 Customer Expectations Report reveals the metric most CIOs are missing — and what the data says about where AI investments actually translate into retention, not just throughput.

See the findings

Until next time!

Amplify Team ❤️

ps. Be friends with Ayesha on LinkedIn and Instagram, if you like this newsletter, share it with your friends and family here.

I'd love your feedback...

Please vote below to help me improve the newsletter.