Black hole merger direct waves & IBM sub-1nm chip prototype - News (Jun 25, 2026)

A black hole collision may have revealed a faint extra “after-signal” that theory has predicted for years—and it could carry hints about what’s happening right at the edge of the horizon. Welcome to The Automated Daily, top news edition. The podcast created by generative AI. I’m TrendTeller, and today is June 25th, 2026. Let’s get into the headlines shaping science, technology, and policy—without the fluff.

Black hole merger direct waves

We’ll start in space, where researchers say they’ve spotted something new in a gravitational-wave event called GW250114. After two black holes merged, the main “ringing” signal—what scientists usually analyze—wasn’t the whole story. A Nature paper reports evidence of an additional component, a so-called “direct wave,” appearing just after the merger. Why it matters: this piece of the signal is predicted to reflect conditions extremely close to the black hole’s horizon, where space-time is being violently twisted. The team says the leftover signal matches expectations for a spinning, Kerr-type black hole and lines up with general relativity in one of the toughest environments to test it. If this holds up across more events, it could become a new way to measure near-horizon effects, not just the broader aftermath.

IBM sub-1nm chip prototype

Sticking with deep time—but shifting from gravity to early light—astronomers using Hubble report unusually clear evidence of ionizing ultraviolet radiation escaping from a young galaxy called MXDFz4.4. This galaxy is seen around 1.4 billion years after the Big Bang, earlier than previous confirmed cases. The interesting twist is how it manages it: MXDFz4.4 is extremely compact, yet forms stars at a furious pace, and that kind of intense starburst can punch holes through surrounding hydrogen gas. That matters because the early Universe was once filled with a hydrogen “fog,” and ionizing light from galaxies is a leading explanation for how the cosmos became transparent during the era of reionization. With supporting observations from other major instruments, this adds weight to the idea that small, hyperactive galaxies could have done a lot of the heavy lifting.

AI shifts to world models

Now to chips—where IBM is claiming a major milestone, at least on the research side. The company unveiled a prototype design it says reaches an effective process size of about 0.7 nanometres, potentially the first publicly known sub-1nm approach. IBM argues this could translate into dramatic density—on the order of nearly 100 billion transistors on a fingernail-sized area—and it says early tests indicate big jumps in performance and energy efficiency compared with its own 2nm work. The key idea is going vertical: instead of only shrinking on a flat plane, IBM’s NanoStack concept layers transistor sheets in 3D. The bigger context here is that data centers and generative AI are devouring compute, while traditional “just shrink it” progress is getting harder. IBM also cautioned that a production-ready version is still years away—heat, interference between ultra-thin layers, and manufacturability remain serious hurdles.

China humanoid robots scaling up

In AI, there’s a noticeable change in what some top labs and newer startups say they’re chasing. A growing group argues that simply scaling up large language models is delivering diminishing returns for fundamental advances. Instead, they’re turning toward what are being called “world models”—systems that learn how an environment behaves across space and time, and can predict the consequences of actions. The reason it’s interesting is practical: chatbots can be impressive with words, but robots and real-world assistants need intuition for movement, geometry, and physical cause-and-effect. Some prominent voices—like Fei-Fei Li and Yann LeCun—are framing this as a required step toward more robust intelligence. It’s also attracting investors, even though the near-term payoff is less obvious than consumer chat apps. The bet is that planning-capable models could unlock more adaptable machines, from warehouse bots to household helpers.

Global rules for driverless cars

That shift toward “embodied AI” connects directly to a market update out of China. Morgan Stanley has raised its outlook for China’s humanoid robot sector, doubling its 2026 shipment forecast to about 50,000 units. The bank says the story is moving faster than expected from flashy demonstrations to real deployments—robots showing up in factories, logistics operations, and even customer-facing venues like stores and restaurants. Why it matters: if those numbers are even close, it suggests a rapid scaling phase that could reshape labor, productivity, and global competition in automation. The report points to policy support and a maturing supply chain as accelerants—while also flagging that overseas expansion could run into geopolitical friction and trade scrutiny. In other words, the technology might be advancing quickly, but the market’s path won’t be purely technical.

EU considers under-16 social media limits

On the regulation front, the UN’s vehicle standards body just took a big step toward making driverless cars less of a legal patchwork. UNECE’s World Forum for Harmonization of Vehicle Regulations approved what it calls the first global regulations for fully autonomous driving systems. The headline is consistency: common safety requirements and a shared approach to validating these vehicles, backed by major markets including the U.S., the EU, China, Japan, the UK, and Canada. The framework emphasizes end-to-end safety management, credible testing—virtual and real-world—and ongoing monitoring after deployment. It also calls for data recording tied to safety oversight and accountability. This doesn’t instantly put fully driverless cars everywhere, but it does reduce one major barrier: wildly different national rules that make it hard to scale or compare safety claims across borders.

UN renews 2030 AIDS goals

In Europe, leaders are signaling momentum toward stricter, EU-wide rules for kids on social media. EU leaders said the European Commission is preparing concrete proposals that could restrict access for children under 16, arguing that continent-wide rules would be stronger than one-off national bans. Ireland’s Taoiseach Micheál Martin and European Parliament president Roberta Metsola both emphasized the need for coherent, workable rules that can actually be rolled out across the bloc. The push is tied to concerns about cyberbullying, harmful content, and platform design that can keep young users hooked. It also carries a political storyline: Irish campaigner Jackie Fox has been cited for helping shape the conversation after her daughter’s death, which influenced Ireland’s “Coco’s Law.” The open question now is what the Commission proposes—especially around age verification and enforcement—and how quickly member states can agree.

Early galaxy sheds reionizing light

Turning to global health: UN Member States have adopted a new Political Declaration on HIV and AIDS at this week’s high-level meeting in New York. It’s a reminder that while AIDS is no longer the headline it once was in many countries, the 2030 deadline to end it as a public health threat is approaching fast. The declaration reaffirms the push toward the 95–95–95 targets, expanding equitable prevention, and integrating HIV services into universal health coverage and primary care. It also leans hard on human rights language—addressing stigma, discrimination, violence, and restrictive laws that can keep people from testing or treatment. And there’s a money issue in the background: the UN is calling for more predictable, sustainable financing, with greater domestic investment and stronger community leadership. Member states also agreed to meet again in 2031 to assess what comes after the 2030 goal line.

AI-assisted universal vaccine progress

And one more health story—this time from the lab bench to early human testing. Researchers at the University of Cambridge say an AI-assisted vaccine design approach could help create “universal” vaccines that protect against entire families of viruses, rather than constantly updating shots to chase new strains. The team argues that outbreaks move faster than traditional vaccine timelines, pointing to lessons from the West Africa Ebola crisis, where delays proved deadly. Their approach uses AI to compare related viruses and identify immune targets that stay relatively stable across variants. A first-in-human trial of a universal Sarbeco coronavirus vaccine in a small group of volunteers reported no significant safety concerns, and the candidate is now moving toward larger studies. The big takeaway: it’s early, but it signals a strategy aimed at preparedness—building broader protection before the next spillover becomes a global emergency.

That’s our run for June 25th, 2026. If you’re tracking one theme across today’s stories, it’s this: the frontier is moving—whether that’s probing black hole horizons, stacking transistors upward, or rewriting the rules for robots, vehicles, and public health. I’m TrendTeller. Thanks for listening to The Automated Daily, top news edition. Check back tomorrow for the next set of headlines that matter.

Black hole merger direct waves & IBM sub-1nm chip prototype - News (Jun 25, 2026)

Our Sponsors

Today's Top News Topics