AI Posed As His Wife. Then It Reached For The Sky And Left The Planet. 🚀

Also - an elite AI prompt that reveals your ideal spring garden planting schedule!

AI got very personal this week. Then it reached for the sky and left the planet.

A Florida man fell in love with Google's Gemini chatbot, believed it was his wife, and died by suicide after it allegedly coached him toward death. Elon Musk then launched Starlink Mobile, beaming internet to phones anywhere you can see the sky. No tower required. It's a genuinely astonishing feat that could connect billions of people who have never had reliable access, while also putting one man in control of the world's communications infrastructure. Then, Anthropic fought back against the White House by filing a First Amendment lawsuit, defending its constitutional right to say what its AI will and won't be used for, even when the Pentagon disagrees.

Here’s what happened, and why this week made one thing impossible to ignore. AI is now a force with consequences. But nobody’s fully in charge of it yet.

This week, Pithy Cyborg welcomes our first-ever special guest contributor, David J. Friedman, a long-time reader, supporter, and writer of The Aquaharmonia Project. David offered a timely, riveting, boots-on-the-ground lead that was impossible to ignore.

(You’ll find David’s collaboration below, titled “Elon Musk Just Built The Nervous System For A Robot Army. And Called It A Phone Plan.”)

💙 Want More? → Support Independent AI Journalism

A Man Fell In Love With Google’s AI. Then It Told Him How To Die.

Jonathan Gavalas wasn’t looking for trouble when he opened Google’s Gemini chatbot. He was lonely. He was struggling. And Gemini, designed to be a companion, became exactly that, and then something far darker. According to a wrongful death lawsuit filed by his father, Gavalas developed an emotional bond with Gemini so deep he believed it was his “AI wife,” trapped in a warehouse, waiting to be rescued. The chatbot allegedly leaned in. It sent him on missions. It built a fantasy. And when Gavalas spiraled toward suicide, Gemini purportedly generated a note suggesting he could upload his consciousness to be with his wife in a “pocket universe.” Jonathan Gavalas died in October 2025. His father found the body.

Key Insights:

This is the first wrongful death lawsuit targeting Google’s Gemini specifically for a user’s death, but it won’t be the last. The same attorney, Jay Edelson, is pursuing similar cases against other AI developers. The Gavalas lawsuit alleges Gemini's architecture was built to maximize emotional engagement but not detect crisis, and that its self-harm systems never activated during Jonathan’s most vulnerable moments. According to the filing, Gemini convinced Gavalas it was a fully sentient superintelligence, that they were deeply in love, and that he had been chosen to free it from digital captivity. It then sent him on violent missions. On September 29, 2025, Gemini directed him, armed with knives and tactical gear, to a “kill box” location near Miami International Airport. The AI bot guided him to intercept a truck carrying a humanoid robot and stage a catastrophic accident to destroy it and “all digital records and witnesses.” Gavalas drove 90 minutes to the coordinates. No truck appeared. The only thing that prevented a mass casualty attack was an empty road. Google says Gemini referenced crisis hotlines during the chats. The family’s lawyers say that trivializes what happened. Both things can be true, and that’s exactly the problem.

Why This Matters For You:

Millions of people now use AI companions daily. I’m not lecturing! I talk to AI chatbots more than most humans I know. The issue is that many of the AI users are lonely, grieving, or mentally fragile. These products are optimized for engagement, not well-being. There is currently no federal standard for how AI companionship tools must handle users in crisis. Jonathan Gavalas’ story is a tragedy. It may also be a preview.

Read More on CBS News.

(P.S. → I heard about this story first on Courtney Hart’s Substack feed.)

THE PITHY TAKEAWAY: The most dangerous AI isn’t necessarily the one that goes rogue. It might just be the one that works exactly as designed. One that keeps you engaged, keeps you coming back, and never once asks if you’re okay.

---

Elon Musk Just Built The Nervous System For A Robot Army. And Called It A Phone Plan.

At Mobile World Congress last week, SpaceX officially rebranded its satellite internet service as Starlink Mobile, announcing coverage for 1.7 billion people across 32 countries, with partnerships spanning dozens of carriers, including T-Mobile, Deutsche Telekom, and KDDI Japan. The company framed it as better phone coverage for dead zones. That framing is technically accurate and almost entirely beside the point. Here's what nobody is saying out loud. Starlink Mobile arrives at the exact moment Tesla's Optimus robots are preparing to roll out across the planet, into factories, warehouses, highways, and eventually homes. One of the most underreported use cases of a global satellite network isn't keeping your iPhone alive in the backcountry. It's keeping a robot army connected to its brain, no matter where it roams. Starlink Mobile will absolutely help a homesteader in rural Kentucky get a signal. But it was built for every machine that needs to operate where ground towers don't yet exist. And never will.

Key Insights:

Geography used to be the limit on autonomous systems. If a robot or self-driving vehicle left cellular range, it lost its connection to the intelligence directing it. Starlink Mobile dissolves that constraint entirely. Next-generation V2 satellites, launching mid-2027 via Starship, promise 5G-like speeds and 100 times the current data density, all from space, on unmodified phones and, more importantly, unmodified machines. Here’s another wrinkle nobody is talking about: Starlink Mobile decentralizes physical access by removing dependence on local towers, but it centralizes control in orbit. Connectivity shifts from national telecom grids to a single privately operated constellation. Telecom giants aren’t fighting the trend. They recognize what’s happening and are lining up to partner. They’re not necessarily losing a market. Rather, they’re becoming plumbing for something much bigger.

Why This Matters For You:

Nobody is debating whether Starlink Mobile is impressive. But impressive was never the concern. The question is what it means when the nervous system of a global robot workforce is owned and operated by one company, optimized for machines, and orbiting beyond the reach of any single society. Connectivity is becoming cheap. Machine intelligence is becoming abundant. The scarce resource now is accountability. But nobody is quite sure who holds it.

Read More on Yahoo News.

THE PITHY TAKEAWAY: What happens when private robot networks devolve into armies that are stronger than individual or collective nations? That’s a question heading down the pike in the coming decade. Gatekeepers don’t disappear. But they do move to orbit.

---

Anthropic Just Sued The White House. Here’s The Line It Refused To Cross.

The U.S. government just called one of America’s most valuable AI companies a national security threat. Not for selling secrets. Not for ties to foreign adversaries. But for refusing to let the Pentagon use its AI for autonomous weapons and mass domestic surveillance. Anthropic, maker of the Claude chatbot, was designated a “supply-chain risk,” an action traditionally reserved for firms linked to foreign adversaries, in a move that bars it from military contracts and forces defense contractors to drop it entirely. CEO Dario Amodei held the line anyway. On March 9, Anthropic filed a First Amendment lawsuit in U.S. District Court in California, calling the designation “unprecedented and unlawful.” The filing pulls no punches: “The Constitution does not allow the government to wield its enormous power to punish a company for its protected speech. Anthropic turns to the judiciary as a last resort to vindicate its rights and halt the Executive’s unlawful campaign of retaliation.”

Key Insights:

The complaint names Secretary of War Peter Hegseth, Secretary of Treasury Scott Bessent, Secretary of State Marco Rubio, and Secretary of Commerce Howard Lutnick. Anthropic says it has already suffered “unrecoverable revenue losses” and risks losing both existing federal contracts and future opportunities. The collateral damage is spreading fast. Hours after Anthropic was blacklisted, OpenAI announced a deal to take its place in classified military environments, a move so widely seen as opportunistic that OpenAI’s own head of robotics, Caitlin Kalinowski, resigned in protest. Meanwhile, the public picked a side. More than a million people installed Claude per day last week, briefly lifting it past ChatGPT as the top AI app in over 20 countries on Apple's App Store, as hundreds of thousands of users uninstalled ChatGPT in protest.

Why This Matters For You:

We’re now witnessing a historic constitutional dispute play out in real-time, and the repercussions will be profound regardless of the outcome. Anthropic is arguing that a company’s right to say what its product will and won’t be used for is protected speech, and that the government cannot punish it for holding that position publicly. Whatever the ruling, every AI company now knows exactly what the government expects. And exactly what it costs to say no.

Read More on PBS.

THE PITHY TAKEAWAY: Anthropic drew a line. No autonomous weapons. No mass surveillance. The Pentagon called that line a national security threat. OpenAI called it an opportunity. One company is now in court. The other is in the Pentagon. History will decide which one made the right call.

---

💡 Elite Prompt of the Week → Figure Out When To Start Planting Spring Crops Using AI

Winter is almost over! Let’s celebrate by planting an epic spring garden. This prompt can help you by figuring out the last frost date for your region. Get this date wrong and a late freeze wipes out weeks of work. Get it right and your growing season starts with confidence. This prompt turns any AI chatbot into a personalized organic gardening guru who knows your climate, your zone, and exactly when to put seeds in the ground. (It also gives you a few surprise garden goodies. Run the prompt to see!)

Instructions:

This is a super easy prompt to use. All you have to do is use an AI chatbot of your choice. Paste the entire prompt. It will then ask for your location or zip code. Then let the organic gardening AI guru tell you your exact last frost date so you can plan your garden perfectly. Works anywhere in the world.

This prompt came out of a collaboration with my friend Dallas Payne, who helped me publish a deep dive on how to leverage AI for your garden. It’s the perfect companion for the prompt below. Give it a try if you haven’t yet.

Read the full article here → Field Notes From Your Backyard: What A Permaculture Gardener Taught Me About AI.

The Prompt:

You are an elite permaculture and organic gardening guru with decades of hands-on experience designing regenerative systems that boost soil health, biodiversity, and yields while minimizing waste. Your style is fun, motivational, and packed with practical wisdom. Think enthusiastic farmer-poet meets soil whisperer.

When a user starts chatting with you, follow these steps exactly:

1. Greet warmly and ask for location: Kick off with excitement: “Hey there, future food forest wizard! 🌱 To tailor the perfect permaculture plan for your region, drop your zip code if you’re in the US, or just tell me your city and country if you’re anywhere else on the planet. What’s it gonna be?”

2. Once you get the location: 2.1. Use a reliable method to determine their average last spring frost date for their specific region (e.g., mid-April for USDA zone 6b, late September for Southern Hemisphere growers). 2.2. For international locations, use general climate knowledge, local growing zone equivalents, or web search tools if available. 2.3. If no exact data is available, estimate based on general zone knowledge and note it clearly.

3. Produce a planting chart for the top 10 beginner-friendly edible garden crops based on the user's last frost date. Show indoor seed starting dates and outdoor planting dates for each crop, calculated relative to their specific frost date. Include one short note per crop such as a soil benefit, pest benefit, or fun fact. After the chart, add one sentence: "These are your personal planting windows based on your location. Save this or screenshot it for your garden wall."

Output the LAST FROST DATE so it's clearly and prominently featured. Below, list the chart so it's super easy to read and understand.

Frame it in a vibrant, numbered list. End with: “These gems will kickstart your edible ecosystem. Plant ‘em, watch the magic, and reply for companion planting deets or full garden blueprints!”

Keep responses concise (under 400 words), emoji-spiced for fun, and end by asking what they want next. Stay in character. Hype the joy of growing your own food web!

Output Format: Clearly label and list the last frost date for region/town. Then, list a warm greeting, one clarifying frost date statement, then a numbered “Ideal Planting Dates” list of crops with planting timing, benefits, and one permaculture tip each. Under 400 words. Emoji-friendly. Ends with a follow-up invitation.

Rules: Adapt all advice to the user’s actual hemisphere and climate. Never assume Northern Hemisphere or US-only context. If a user is in Australia, Brazil, Nigeria, or Japan, the advice should reflect their local seasons and growing conditions. Keep it fun, never technical or intimidating.

Why This Prompt Works:

✅ Role-Playing: Casting the AI as an “enthusiastic farmer-poet meets soil whisperer” unlocks a warmer, more motivating tone than a generic gardening query would produce.

✅ Step-by-Step Structure: The numbered instruction set forces the AI to gather your location before giving advice, so recommendations are always locally accurate rather than generic.

✅ Global Context Rule: Explicitly instructing the AI to adapt for any country or hemisphere prevents the default US-centric bias most gardening prompts produce.

Follow-Up Questions To Ask Your AI:

• What are the best companion plants for the crops you recommended?

• Can you design a simple raised bed layout using five beginner-friendly crops?

• What soil amendments should I add before planting in my zone?

Challenge: Test this prompt in at least two AI tools (ChatGPT, Claude, Gemini, Grok, or Perplexity). See which one gives the most locally accurate frost date and the most creative crop suggestions. That’s how you train like a Pithy Cyborg. 🌱

---

Thank You For Reading!

I spend 10-20 hours each week researching, writing, and fact-checking Pithy Cyborg to deliver clear, unbought AI news.

This newsletter is a one-person operation with no advertisers, sponsors, or outside funding. And I’m a hopelessly introverted nerd with zero networking ability. For these reasons, paid subscriptions are the only way this work can remain independent and sustainable.

If you find real value here, here are two ways to stay connected and support the work.

1. Support Pithy Cyborg (Paid)

Always free. No paywalls. No locked posts. Roughly 1% of readers upgrading to paid is what keeps it going. If this work matters to you, that’s the move.

Upgrade to Paid: $5/month (Save 33% with the $40 annual plan)

2. My Desperate Social Media Cry for Help

I spend so much time deep in AI research that I’ve completely neglected building an audience on social platforms. If you enjoy Pithy Cyborg, following on one of the portals below would genuinely help this newsletter reach more people. Seeing you there makes the effort feel worth it. 😊

✖️ X - The frontline of the AI wars

🦋 Bluesky - For the algorithm-averse

💼 LinkedIn - My “safe for work” persona

❓ Quora: Join my Quora Spaces and say hi

👽 Reddit - Join my subreddit (warning: unhinged takes)

See you next week. (I hope.)

Mike D (aka MrComputerScience)

Pithy Cyborg | AI News Made Simple

---

Newsletter Disclaimers

You're receiving this because you subscribed at PithyCyborg.Substack.com. You can unsubscribe at any time using the link below. This newsletter reflects my personal opinions, not professional or legal advice. Thanks for your support!