SpaceX orbital AI data centers & Underwater wind-powered data centers - Tech News (Jun 10, 2026)

What if the next AI “data center” doesn’t go up in your neighborhood… but goes up into orbit? Welcome to The Automated Daily, tech news edition. The podcast created by generative AI. I’m TrendTeller, and today is June 10th, 2026. We’ve got a packed briefing: SpaceX’s bold play to run AI workloads in space, a very different approach with an underwater data center powered by wind, fresh IPO maneuvering in the AI world, and a new round of regulatory pressure on big messaging platforms. Let’s get into it.

SpaceX orbital AI data centers

Space and AI are colliding in a big way. Elon Musk and SpaceX just shared their most detailed look yet at what they’re calling an orbital AI data-center satellite—essentially a spacecraft meant to host AI computing in orbit instead of on Earth. The pitch is simple: space offers abundant solar energy and avoids the land, grid, and community friction that comes with building massive terrestrial data centers. What makes this announcement especially notable is the timing—SpaceX is also ramping up IPO-related messaging, and this “space compute” narrative reads like a major pillar of its next growth story. SpaceX is also trying to head off a key risk: chip dependence. The company says the compute module would be swappable so it isn’t locked into one silicon supplier, even while acknowledging real-world supply constraints. Critics, meanwhile, are already challenging the economics and practicality—especially around keeping powerful computers cool in a vacuum—so expect a loud debate between ambition and feasibility.

Underwater wind-powered data centers

And in a kind of mirror image to “compute in orbit,” China is pushing “compute underwater.” A new demonstration project near Shanghai is being billed as the world’s first wind-powered underwater data center. The core idea is to use the ocean as a natural cooling system, cutting energy spent on keeping servers from overheating and reducing freshwater use at a time when AI infrastructure is straining local resources. It’s also a signal of where the industry is headed: the competition isn’t only about better models, it’s about where we can physically run them—within real limits on power, water, and public tolerance. Environmental monitoring will matter here, too, because even if the energy is clean, building and operating hardware in marine environments brings its own risks.

OpenAI and SpaceX IPO buzz

On Earth, legacy automakers are trying to make the AI power boom work for them. General Motors says it wants to grow beyond EVs by targeting grid-scale energy storage and the rising electricity appetite of data centers. The big picture: if electricity demand keeps climbing—driven partly by AI—then batteries aren’t just for cars, they’re for balancing the grid and keeping facilities running. GM is also leaning into the idea that EVs themselves can become part of the energy system, helping homes or utilities during peak demand. For consumers, that’s a potential shift from “a car that uses power” to “a car that can also provide power,” depending on utility partnerships and how quickly vehicle-to-grid programs spread.

EU orders Meta WhatsApp access

Now to the money—and the momentum—around AI companies going public. OpenAI is reportedly taking a serious step toward an IPO by filing confidentially, even while saying it hasn’t committed to a timeline. The headline number floating around is user scale: ChatGPT is now described as having an enormous monthly audience. The counterweight is the cost: the company is still spending heavily to build and rent the compute capacity needed to run and expand its models. This is the tension that could define the next phase of the AI boom. Demand looks real, but so do the infrastructure bills. If multiple mega-IPOs land around the same time, it could reshape hiring, investment flows, and even real estate in major tech hubs—while also putting a brighter spotlight on profitability, governance, and the societal impact of automation.

Claude 5 Fable and coding agents

In Europe, regulators are pressing on a different part of the AI economy: distribution. The European Commission has ordered Meta to restore free access for rival AI chatbots to the WhatsApp Business API, after Meta blocked third-party general-purpose assistants from using it. This is framed as an interim measure while the EU investigates whether Meta used platform control to squeeze competition. What’s interesting here is that messaging apps are becoming the battleground for AI assistants. Whoever controls the pipes can shape who gets to build the “default” assistant experience inside chat—and that’s why the EU is treating access as urgent, not theoretical. Meta is expected to appeal, so this one is far from settled.

Apple Siri AI vs Microsoft agents

Let’s talk about the models themselves. Early testers are describing a notable jump in capability with Anthropic’s Claude 5 Fable, a guardrailed version in the new “Mythos-class” family. The consistent theme from hands-on reports is less about a single trick and more about endurance: long, multi-step work sessions where the model effectively manages a project—researching, drafting, coding, revising, and cross-checking. One reviewer describes the human role shifting from hands-on builder to commissioner—someone who sets direction and then audits the output. That’s a powerful productivity story, but it also comes with tradeoffs: higher usage costs, more opaque decision-making mid-process, and guardrails that can change behavior or downgrade capability when topics drift near security-sensitive territory.

China EV export surge and BYD

That change in workflow is already rippling through developer culture. A survey making the rounds suggests many developers are using LLMs to draft technical blog posts—often because blogging is suddenly expected at work, or because it feels like a shortcut to visibility. But the same survey points to a reality check: most people still do heavy editing, and many say the draft doesn’t capture their voice or even their ideas particularly well. One detail stands out: most respondents reportedly don’t disclose AI assistance. That’s likely part of why readers are increasingly noticing “AI-scented” posts—content that feels plausible but oddly generic. The trust layer of developer writing is getting stress-tested, and we’re going to see more norms emerge around disclosure, editorial responsibility, and what “authorship” means when a draft starts with an agent.

Ukraine bid to mass-produce drones

Apple’s WWDC narrative adds another angle to this: default, integrated AI. Commentary this week argues Apple’s AI announcements weren’t the most groundbreaking in raw capability, but could still make Apple the consumer AI leader because the features are built in, enabled by default, and wired into everyday apps. The bet is that most people won’t install a separate assistant if the built-in one is good enough—and if it actually works reliably, which is the lingering skepticism after earlier Siri disappointments. In parallel, analysts are contrasting Apple’s device-centric approach with Microsoft’s newly teased “Project Solara” vision, which imagines thin devices acting as portals to cloud agents doing a lot of the work with minimal human involvement. In short: Apple wants the iPhone to stay the center of gravity, while Microsoft is pitching the agent as the center of gravity—especially at work.

Gene therapy reprogramming for glaucoma

A quick but meaningful developer note from Apple, too: the company has open-sourced a Swift-based tool called “container,” aimed at making Linux containers feel more native on Macs, particularly Apple silicon machines. It’s positioned as a practical workflow improvement for developers who want modern container habits without fighting their platform. The catch is that it targets the newest macOS generation, so it’s also a quiet nudge: to get the best tooling, keep up with Apple’s OS cadence.

Developer tools and AI workflows

Global hardware competition is also heating up through exports. China’s passenger car exports surged in May, with electrified vehicles taking a growing share. The context matters: domestic sales in China have been weak, so overseas markets are where Chinese automakers can keep volume rising. BYD, in particular, is openly aiming to become the world’s largest automaker within a few years, and it’s accelerating European production plans to reduce tariff exposure. But that expansion is not frictionless. BYD is drawing scrutiny tied to local labor and environmental concerns in Europe, while also facing geopolitical headwinds—another reminder that the EV story is now inseparable from trade policy and national security politics.

Staying with geopolitics, Ukraine is pitching an enormous scale-up of drone production—arguing it could manufacture on the order of tens of millions of drones per year if NATO allies provide the necessary funding and investment. The message is that drones have become central to battlefield effectiveness, and that Ukraine can turn wartime iteration into an industrial advantage. The limiting factor may not be factory space—it’s components. Chips, sensors, and supply chains are still chokepoints, and funding at the requested level is far from guaranteed. But the strategic point is clear: whoever can produce and adapt drones fastest can shift the balance of attrition warfare.

In biotech, a notable first: Life Biosciences has dosed the first participant in a clinical trial testing partial cellular reprogramming, a gene-therapy approach meant to make aged cells behave more youthfully. The first target is glaucoma, aiming to regenerate or rejuvenate optic nerve cells. This is a landmark because it moves a long-discussed longevity-adjacent technique from animal results into human safety testing. And safety is the entire ballgame here—because pushing cells toward a more youthful state can, in the worst case, push them toward uncontrolled growth. Starting with the eye is a cautious choice: it’s more contained, easier to monitor, and potentially less risky than systemic delivery.

Finally, in space exploration beyond the business headlines, NASA has named the four-astronaut crew for Artemis III and outlined a revised mission profile that’s getting attention for its complexity. The plan involves multiple rendezvous and dockings with major contractor hardware in a tight sequence—creating several moments where a delay or a technical miss could cascade. That matters because Artemis timelines are increasingly coupled to private-sector readiness. Recent setbacks in heavy-lift development on the Blue Origin side, and remaining milestones for SpaceX’s Starship, make the schedule feel more contingent than ever. The ambition is enormous—returning humans to the Moon after more than half a century—but so is the choreography required to get there.

That’s the tech landscape for June 10th, 2026: data centers migrating to oceans and orbit, regulators fighting over the AI distribution layer, and new models nudging humans from “doer” to “director.” If you want to support the show, share this episode with a colleague who’s tracking AI infrastructure or platform regulation—it’s all converging faster than most roadmaps admit. I’m TrendTeller, and you’ve been listening to The Automated Daily, tech news edition. Check back tomorrow for the next briefing.

SpaceX orbital AI data centers & Underwater wind-powered data centers - Tech News (Jun 10, 2026)

Our Sponsors

Today's Tech News Topics