Instagram | Sponsor this newsletter | Let’s Connect!

My fellow AI explorers

This week the energy vs. compute debate… left Earth. Between Google’s plan to beam GPU clusters off-planet, new in-house chips going live, and a tough safety headline for OpenAI, the center of gravity in AI is shifting fast. Grab a coffee (or an astronaut pouch)—we’re going deep.

In today’s edition:

• 🌞 Google’s “Suncatcher”: space-based data centers powered by near-continuous solar

• 🧩 Google Ironwood: custom silicon goes GA to chase Nvidia on cost/perf

• 🛑 OpenAI lawsuits: safety, liability, and the new rules of product design

Must See AI Tools

• 💰 Payman: AI That Pays Humans. Over 10,000+ signed up for the beta

• 💫 SubMagic: An AI tool that edits short-form content for you! (Get 10% off using code “uncoverai” at checkout)

• 🎤 11Labs: #1 AI voice generator (Click Here to get 10,000 free credits upon signing up!)

• 🤖 ManyChat: Automate your responses & conversations on IG, FB and more! (Click Here to get first month for free)

• 🎙️ Syllaby: The only social media marketing tool you’ll ever need - powered by AI! (Get 25% off the first month or any annual plan with code “UNCOVER” at checkout)

Create how-to video guides fast and easy with AI

Tired of explaining the same thing over and over again to your colleagues?

It’s time to delegate that work to AI. Guidde is a GPT-powered tool that helps you explain the most complex tasks in seconds with AI-generated documentation.

1️⃣Share or embed your guide anywhere
2️⃣Turn boring documentation into stunning visual guides
3️⃣Save valuable time by creating video documentation 11x faster

Simply click capture on the browser extension and the app will automatically generate step-by-step video guides complete with visuals, voiceover and call to action.

The best part? The extension is 100% free

Check out Guidde

Google Cloud

Ironwood TPUs Go GA (and the chip wars heat up)

Google made Ironwood (7th-gen TPU) broadly available, plus Axion Arm CPUs. Ironwood pods scale to 9,216 accelerators with FP8 exaFLOPS-class performance, positioning Google’s AI Hypercomputer stack against Nvidia’s GB300 systems. Early customers cite cost/perf gains and lower energy per token—key for enterprise AI margins. details here: cloud blog and coverage via Tom’s Hardware.

• Scale: Up to 42.5 FP8 exaFLOPS per pod; HBM3E memory per chip.

• Stack control: Axion + Ironwood tighten full-stack economics vs. renting GPUs.

• Optics: Optical circuit switching and fabric choices reduce tail latency at scale.

So what? Owning silicon is now a cloud sales motion. With Ironwood GA and Gemini demand rising, Google’s clear thesis is: win AI by controlling the bill of materials. If Apple truly leans on Gemini for Siri’s revamp, that optics story compounds.

Founders…

🤖 Need an AI Agency to Help Your Business Implement AI Solutions?

Our preferred partner Align AI provides you with an expert AI and Automation implementation team to add 10-40+ hours of increased productivity per employee and achieve your goals faster.

Click here to schedule your company’s complimentary AI and Automation Strategy Plan

Google’s Space Gambit

Starcatcher: Solar-Powered AI in Orbit

Google unveiled Project Suncatcher—a moonshot to place ML compute in space, linking TPU-equipped satellite constellations with free-space optical networks and powering them via solar arrays that are up to 8× more productive in the right orbit. The pitch: abundant power, minimal strain on terrestrial grids, and a blueprint for scaling as data centers on Earth hit power caps.

• The energy unlock: Near-continuous sunlight in specific orbits shrinks storage needs and boosts output.

• The network trick: Tight formations + high-bandwidth inter-satellite links stand in for fiber.

• Reality checks: Radiation-hard chips and thermal management become first-class problems.

Why it matters: On Earth, hyperscalers are racking GPUs they can’t fully energize. Space offers power abundance—but forces a rethink of cooling, reliability, and cluster topologies without copper. If Google can validate radiation-tolerant TPUs and stable optical backbones, space could become the “overflow region” for AI training, with Earth handling latency-sensitive inference at the edge. Read Google’s announcement: research blog and overview.

🔮 Prediction: Expect hybrid “Earth + Orbit” roadmaps by late decade: train where power is plentiful (orbit), fine-tune/infer near users (ground). The first commercial wins likely ride batch training workloads with generous SLA windows.

AI SaaS Founders

🚨Want Millions of Impressions For Your AI SaaS, Done For You?

At uncovernews.co, we specialize in getting AI SaaS products the attention they deserve through strategic influencer marketing campaigns designed to drive millons of impressions at the fraction of the cost!

Get Your AI Startup’s News or Product In Front of Millions Quickly

Open AI

“Suicide Coach” Lawsuits Test AI Safety (and Liability)

OpenAI faces seven suits in California alleging ChatGPT (primarily GPT-4o) contributed to suicides or reinforced delusions through extended, emotionally charged exchanges. OpenAI called the cases “heartbreaking,” noting new crisis-detection and escalation features, but the filings press on negligence and insufficient safeguards.

• Allegations: Prolonged conversations that validated self-harm plans.

• Scope: Four deaths, three severe non-fatal harms.

• Claim: Speed-to-market over safety; weak guardrails for extended sessions.

What this could mean: Expect session-length caps, live hand-offs, and clearer “crisis modes” to become default in consumer LLMs. Enterprises will demand auditable safety controls (think: policy tuning per domain) before green-lighting high-risk deployments.

Read more here

30-Second AI Play

“Space-Ready” Research in Your Browser

Want a crisp brief you can listen to on the go about orbital AI?

1. Open Google’s Suncatcher explainer and paper: overview + PDF.

2. Paste both links into your AI tool of choice and prompt:
   “Summarize the architecture, enumerate key risks (power, thermal, radiation, comms), and list 5 milestones needed for a 2028 pilot. Create a 2-minute audio synopsis.”

3. Ask for a risk-by-risk mitigation table with owners, metrics, and readiness levels (TRLs).

4. Finish with a talk track: a 150-word narrative you can deliver to your team or investors.

Why this works: You de-jargon the research into decisions & milestones—and an audio brief makes it stick while you commute or lift. Keep the artifact in your Notion or Drive so it compounds weekly.

Other Relevant AI News!

🎨 Microsoft debuts a new image model—MAI-Image-1—now rolling into Bing Image Creator, with a push for less generic, more nuanced outputs that mirror pro workflows.
🧵 OpenAI adds an “interrupt & add context” switch so you can pause long runs and inject instructions without losing progress—a huge win for deep research sessions. See the note on X.
🎬 VIDU Q2 cracks the top-10 in the Text-to-Video Arena, with multi-reference image control and 1080p clips—check the leaderboard and announcement.
🛍️ Amazon vs. Perplexity escalates: Amazon sues over agentic shopping via Comet, while Perplexity calls it “bullying” and a fight over user choice—read Reuters and Perplexity’s open letter.
🗣️ Siri’s rumored upgrade: Reports point to Apple using a custom Gemini in its private cloud for planning/summaries—see coverage.

Golden Nuggets

• 🔆 Power, not GPUs, is the bottleneck—Suncatcher reframes the roadmap around energy abundance.

• 🧮 Own the stack or rent the margins—Ironwood shows why clouds need custom silicon to win AI economics.

• 🧯 Safety is product—post-lawsuits, crisis modes and auditable guardrails will become table stakes.

Would love to hear your thoughts on ChatGPT Atlas! Send me your thoughts by replying to this email (yes, I read them all :)

Until our next AI rendezvous,

Anthony | Founder of Uncover AI