🗞️In this edition

• Google plans solar-powered AI data centers in space

• Sponsored: Kosmik – Create moodboards in seconds with AI

• Amazon demands Perplexity stop AI shopping feature

• Microsoft launches publisher content marketplace with people Inc

• Workflow Wednesday #44: AI for Decision-Makers

• In other AI news –

  • Shopify says AI-driven shopping traffic is up 7x since January

  • Alexa+ debuts in Amazon Music with natural chat abilities

  • Google’s AI Mode can now book event tickets and beauty appointments

  • 4 must-try AI tools

Google proposes launching AI chips into space on solar satellites

What's happening:

Google announced Project Suncatcher, a moonshot research project to launch AI chips into space on solar-powered satellites, potentially creating space-based data centers.

The company hopes to harness solar power around-the-clock, providing a near-unlimited source of clean energy for AI without concerns about driving up power plant emissions and utility bills through soaring electricity demand.

"In the future, space may be the best place to scale AI compute," Travis Beals, Google senior director for Paradigms of Intelligence, writes. The company published a preprint paper, not yet peer-reviewed, detailing progress.

Google envisions Tensor Processing Units orbiting Earth on satellites outfitted with solar panels generating electricity almost continuously, making them eight times more productive than similar panels on Earth.

Major challenges include ensuring satellites communicate well with each other. Competing with land-based data centers "requires links between satellites that support tens of terabits per second," Google writes. Maneuvering constellations into tight formations can help, perhaps flying satellites within "kilometers or less" of each other. That's much closer than satellites operate today, and space junk from collisions is already a growing risk.

Google must ensure TPUs withstand higher radiation levels in space. It tested Trillium TPUs for radiation tolerance, saying they "survive a total ionizing dose equivalent to a 5 year mission life without permanent failures."

Launching TPUs into space is currently pricey. But Google's cost analysis suggests launching and running a data center in space could become "roughly comparable" to energy costs of an equivalent Earth data center on a per-kilowatt/year basis by the mid-2030s.

Google's planning a joint mission with Planet to launch prototype satellites by 2027 to test hardware in orbit.

Why this is important:

AI's energy demands are becoming an existential problem for tech companies.

Google's greenhouse gas emissions rose 48% since 2019, driven primarily by data center electricity consumption. Regulatory pressure is mounting. Power grids are strained. Nuclear and renewable energy can't scale fast enough.

Space-based data centers solve multiple problems simultaneously. Solar panels in space generate electricity continuously without weather or day-night cycles. No cooling infrastructure needed since space is already cold. No land acquisition costs. No local utility constraints.

Eight times more productive solar panels is game-changing if accurate. That math changes cost calculations dramatically.

The 2027 prototype launch with Planet is a real timeline, not vaporware. Google's committing resources to test this.

Inter-satellite communication requiring "tens of terabits per second" is a technical challenge but not impossible. Starlink already demonstrates satellite constellations can work. Google needs tighter formations and higher bandwidth.

Our personal take on it at OpenTools:

This is either visionary or desperate.

Google's emissions are up 48% since 2019 because AI training runs consume massive power. They can't hit climate goals while scaling AI. Space solves that problem if it works.

The cost analysis claiming space data centers will become comparable to Earth by the mid-2030s is optimistic. Launch costs are dropping thanks to SpaceX, but "comparable" glosses over massive upfront capital.

The biggest question is latency. AI inference needs low latency for consumer applications. Orbital data centers communicating with Earth users face speed-of-light delays. That works for training runs, not real-time inference.

If Google pulls this off, every AI company copies the playbook. If it fails, it's an expensive science experiment.

Read More

Kosmik – Build moodboards at lightning speed

Bring your ideas to life without the chaos of tabs, folders, or file names.

Kosmik’s AI pulls the right images, videos, and notes straight into one clean workspace with just typing a few words — so you can focus on creating, not collecting.

✨ Work faster. - Find the perfect assets in seconds.
 🧠 Stay organized. - Everything you drop in is auto-tagged and categorized.
 🤝 Collaborate easily. - Share boards, gather feedback, and move projects forward together.

For Creatives, Designers, and Visual Thinkers

From first spark to finished concept, Kosmik makes creativity flow.
Try it free

Amazon threatens legal action against Perplexity's AI shopping

What's happening:

Amazon says it has "repeatedly requested" that Perplexity stop allowing its Comet AI browser to buy products for customers. Perplexity responded by accusing Amazon of "bullying."

Perplexity's AI browser Comet offers an agentic AI feature that can find and purchase products from various websites, including Amazon, on users' behalf. 

Perplexity says it received an "aggressive legal threat" from Amazon demanding it stop allowing its AI assistant to shop for users.

"Amazon should love this. Easier shopping means more transactions and happier customers," Perplexity writes. "But Amazon doesn't care. They're more interested in serving you ads, sponsored results, and influencing your purchasing decisions with upsells and confusing offers."

Perplexity cited a quote from Amazon CEO Andy Jassy, who said during last week's earnings call that the company expects to "partner with third-party agents" over time.

"This is like if you went to a store and the store only allowed you to hire a personal shopper who worked for the store," Perplexity spokesperson Jesse Dwyer said. "That's not a personal shopper, that's a sales associate."

Amazon's statement says third-party applications that purchase products for customers on its site "should respect service provider decisions whether or not to participate," claiming Comet provides a "significantly degraded shopping and customer service experience."

Why this is important:

This is the first major conflict between AI agents and ecommerce platforms over automated purchasing.

Amazon makes money from ads, sponsored listings, and upsells during the shopping experience. AI agents that go directly to checkout bypass all that revenue. Amazon's objection isn't about customer experience. It's about protecting ad revenue.

Perplexity's right that Jassy said Amazon expects to partner with third-party agents. But "partner" is key word. Amazon wants control over which agents access its platform and likely wants revenue share agreements.

The "degraded experience" claim is Amazon's legal justification. Reality is AI agents that complete purchases efficiently threaten Amazon's ability to insert ads and recommendations into the buying journey.

This sets a precedent. If Amazon successfully blocks Perplexity, every ecommerce platform can block AI shopping agents. That kills the entire category before it starts.

Our personal take on it at OpenTools:

Amazon's protecting its moat, not its customers.

The entire value proposition of AI shopping agents is bypassing the ad-filled, recommendation-cluttered experience that ecommerce platforms force on users. Of course Amazon opposes that.

Jassy saying Amazon will "partner with third-party agents" means Amazon wants to control the terms. Partnership means revenue share, data access, and veto power over which products agents can buy.

Perplexity's "sales associate" analogy is accurate. Amazon wants agents that work for Amazon, not agents that work for users.

The legal question is whether Amazon can legally block automated purchases. Their terms of service likely prohibit bots and automated systems. But those were written before AI agents became consumer products.

If Amazon wins this fight, AI shopping agents are dead. Every platform will copy Amazon's playbook and demand revenue share or block access entirely.

The irony is Amazon built its business by making shopping easier. Now they're blocking technology that makes shopping even easier because it threatens ad revenue.

This is the first shot in what'll be a long war between AI agents and platforms that profit from user attention.

Read More

People Inc partners with Microsoft on AI content marketplace

What's happening:

People Inc, one of the largest US media publishers, signed an AI licensing deal with Microsoft. 

Under the deal, People Inc becomes a launch partner in Microsoft's publisher content marketplace. This is the company's second AI deal following its OpenAI agreement last year.

CEO Neil Vogel described the marketplace as "essentially a pay-per-use market where AI players directly can compensate publishers for use of their content on, sort of like an 'a la carte' basis."

He praised Microsoft for being committed to paying for content, adding that Microsoft's Copilot would be the first buyer. "It's a very strong endorsement of us to be in the room with them and a very strong endorsement of the publishing marketplace and the value of content to make AI that is of high value," Vogel said.

The announcement came alongside news that Google Search's AI Overviews has been hurting publisher traffic. For the first time, People Inc shared data showing Google Search, which accounted for 54% of its traffic two years ago, dropped to 24% during the past quarter.

The agreement differs from the OpenAI deal, which Vogel characterized as a more "all-you-can-eat" model, but said People Inc was happy with either. What matters is that its work is "respected and paid for."

People Inc uses Cloudflare technology to block AI crawlers, prompting AI players to approach it with content deals. Vogel attributed blocking crawlers as pushing AI companies to the negotiating table, noting progress on deals was "much further along" after adopting the solution.

He said blocking AI crawlers has been "very effective" and "brought almost everyone to the table," suggesting more deals would be announced.

IAC reported People Inc grew digital revenue 9% to $269M in the quarter, driven by performance marketing and licensing, which saw 38% and 24% growth respectively.

Why this is important:

Google Search traffic collapsing from 54% to 24% in two years shows AI Overviews' impact on publishers.

AI summaries keep users on Google instead of clicking through to publisher sites. That destroys the traffic model publishers built businesses on. People Inc compensating by monetizing content directly through AI licensing deals.

Microsoft's "pay-per-use" marketplace differs from OpenAI's "all-you-can-eat" model. Pay-per-use means publishers get compensated based on actual usage. 

That's better economics than flat licensing fees if content gets used heavily.

Licensing revenue growing 24% shows AI deals are meaningful revenue. Not replacing lost Google traffic yet, but creating a new revenue stream.

Our personal take on it at OpenTools:

The 54% to 24% traffic drop is devastating.

Publishers losing over half their Google referral traffic in two years means business models built on Search distribution are broken. AI Overviews answer questions without clicks. That's an existential threat.

Blocking crawlers is the only leverage publishers have. If AI companies can scrape for free, they will. Forcing them to negotiate by blocking access works. Vogel's right that it brought everyone to the table.

But the economics don't add up yet. 24% licensing revenue growth is good. It's not replacing 55% traffic loss. Publishers are shrinking revenue pools while adding new streams. Net result is still declining.

The real question is whether AI licensing deals can scale to replace lost traffic revenue. So far, no evidence they can. Publishers are negotiating because they have no choice, not because licensing revenue is sufficient.

People Inc's strategy is correct given constraints. But the industry's still losing this fight.

Read More

This Week in Workflow Wednesday #44: AI for Decision-Makers – Workflows for Leaders & Execs

This week, I’ll show you how to use Perplexity to turn hours of competitor research into a 5-minute executive briefing. Think of it as a strategy analyst that never sleeps.

Workflow #1: Out-Research Your Competition with Perplexity (Perplexity Pro)
Step 1: Pick a competitor or market you want to monitor — like “Jasper AI” if you’re in the AI writing space.
Step 2: Ask Perplexity: “Give me a competitive……..We dive into this Perplexity workflow and two more AI-powered decision-making systems in this week’s Workflow Wednesday.

📢 What else is happening?

• Shopify says AI traffic is up 7x since January, AI-driven orders are up 11x – A recent Shopify survey found that 64% of shoppers said they’re “likely” to use AI to some extent when making purchases. 

• Alexa+ comes to the Amazon Music app – Unlike the previous version of Alexa, which only responded to basic commands, Alexa+ is designed to engage in natural dialogue.    

• Google’s AI Mode gets new agentic capabilities to help book event tickets and beauty appointments – The new agentic capabilities are available to all users opted into Google’s experimental arm, Search Labs, in the U.S.

👩🏼‍🚒Discover mind-blowing AI tools

1. ZeroBot - A voice-enabled chatbot that allows users to have conversations with a computer friend that talks like a real person

2. Vidnoz - A tool that allows users to create talking avatar videos by simply selecting a photo and writing a script

3. PitchPower - A software service that helps consultants and agencies create powerful proposals quickly

4. Flowrite - A Chrome extension that helps users write emails and messages 5x faster