The Download: Claude’s inner workings and OpenAI’s “super app”

Plus: OpenAI has unveiled its long-awaited "super app."
This is today's edition of The Download, our weekday newsletter that provides a daily dose of what's going on in the world of technology.
Anthropic found a hidden space where Claude puzzles over concepts
The AI firm Anthropic has got the clearest glimpse yet at what’s really going on inside large language models as they answer questions or carry out tasks. What they found ranges from the mundane to the unnerving.
Researchers at the company built a tool called the Jacobian lens (or J-lens) and used it to uncover a hidden area, which they named the J-space, inside its flagship LLM, Claude.
The J-space contains words related to the response a model is working on but may not ultimately produce. If Claude were a person (which it is not), you might say these hidden words reveal what’s on its mind before it actually speaks.
Read the full story on what they found.
—Will Douglas Heaven
The must-reads
I’ve combed the internet to find you today’s most fun/important/scary/fascinating stories about technology.
1 OpenAI has unveiled its long-awaited "super app"
ChatGPT Work blends its chatbot, coding tool, and new models. (Reuters $)
+ It’s designed to do your work for you and with you. (Ars Technica)
+ And arrived the same day as OpenAI’s GPT 5.6 models. (NYT $)
+ It’s also developing a fully automated researcher. (MIT Technology Review)
2 Humanoids have performed teleoperated surgery on living animals
In the world-first, they removed gallbladders from pigs. (Ars Technica)
+ The human work behind humanoids is hidden. (MIT Technology Review)
3 SK Hynix has landed the largest US listing by a foreign company
The South Korean chip giant raised $26.5 billion. (CNN)
+ Demand for AI data centres has led its profits to skyrocket. (Guardian)
+ But its jumbo share sale may be a sign of overheated times. (FT $)
+ South Korea’s hottest bachelors are chip workers. (MIT Technology Review)
4 Tencent is leading a deal to unwind Meta's $2 billion Manus acquisition
It’s in talks to become the Chinese AI startup’s largest shareholder. (FT $)
+ Tencent will reportedly buy Manus for no less than $2 billion. (Reuters $)
+ Beijing had ordered Meta to unwind the acquisition. (Bloomberg $)
5 Resuscitated human retinas responded to light 10 hours after death
It’s a big step towards eye transplants that restore vision. (New Scientist $)
+ As is a new device that revives dead eyeballs. (MIT Technology Review)
6 Meta has started charging for AI access
A new version of Muse Spark has a paid tier for developers. (Quartz)
+ Meta also plans to start producing an AI chip in September. (Reuters $)
7 OpenAI and Google have sold AI models to blacklisted China groups
Via Singapore-based subsidiaries of Alibaba, Baidu and Tencent. (FT $)
8 A daughter tested an AI “death bot” of her father
The technology provided both comfort and unease. (New Yorker $)
9 An astronomer says the hunt for alien life needs more statistics
He wants to replace speculation with mathematical frameworks. (Quanta)
10 Pokémon Go players turned Times Square into a giant battlefield
More than 1,500 fans finally fulfilled the game’s 2016 launch promise. (Wired $)
+ Pokémon Go is also training world models. (MIT Technology Review)
Quote of the day
“When we’re talking about AI, we love the hype, we get excited about it. The damn thing never actually lands in practice.”
—Vijay Janapa Reddi, an engineering professor at Harvard University, tells Wired why he’s skeptical about grand plans for AI.
One More Thing
Why we should thank pigeons for our AI breakthroughs
In 1943, psychologist B.F. Skinner led a secret government project to make bombs more precise. His idea: teach pigeons to guide missiles by pecking at targets on a screen inside a warhead. To train them, Skinner rewarded the birds with food when they made the right decisions, using trial and error to shape their behavior.
Unsurprisingly, the military never deployed Skinner’s kamikaze pigeons. Yet his experiments convinced him that pigeons were “an extremely reliable instrument” for studying learning.
Decades later, those same principles would help power reinforcement learning, the technology behind some of today’s most advanced AI systems.
Discover how pigeons inspired one of AI’s most powerful techniques.
—Ben Crair
We can still have nice things
A place for comfort, fun, and distraction to brighten up your day. (Got any ideas? Drop me a line.)
+ Here’s a splendid selection of this year’s NSW architecture award winners.
+ Photographers have captured the Strawberry Moon’s golden glow in stunning detail.
+ Idiocracy is the film that best exemplifies the “American experience,” according to a new poll. Look back at the prescient comedy with this Screen Junkies trailer.
+ Get ready for the weekend with this psychedelic house journey from Jamie xx b2b Caribou.
Deep Dive
The Download
The Download: the future of chipmaking and Anthropic’s government clash
Plus: Meta is pausing an AI training program that tracks workers’ keystrokes.
The Download: AI hacking beyond Mythos, and chatbots’ impact on our brains
Plus: Anthropic has called for a global slowdown in AI development.
The Download: keeping up with AI, and the future of IVF
Plus: NASA unveiled plans for three uncrewed missions to the Moon this year.
The Download: a reality check for geoengineering and the science of interoception
Plus: SpaceX is now valued higher than Amazon.
Stay connected
Get the latest updates from
MIT Technology Review
Discover special offers, top stories, upcoming events, and more.

Facts Only

* Anthropic researchers used the Jacobian lens (J-lens) to find a hidden area named J-space inside Claude.
* The J-space contains words related to the response a model is working on but may not produce.
* OpenAI unveiled a "super app" blending its chatbot, coding tool, and new models alongside GPT 5.6 models.
* Humanoids performed teleoperated surgery on pigs, including removing gallbladders.
* SK Hynix raised $26.5 billion and profit increased due to demand for AI data centers.
* Tencent is in talks to become the largest shareholder in Meta's Manus acquisition.
* Resuscitated human retinas responded to light ten hours after death.
* A daughter tested an AI "death bot" of her father.
* An astronomer advocates for mathematical frameworks in the search for alien life.
* Pokémon Go players fulfilled a 2016 game launch promise and trained world models.

Executive Summary

Anthropic researchers uncovered a hidden space within the Claude large language model by using a tool called the Jacobian lens (J-lens). This J-space contains words related to a model's intended response that may not be ultimately produced. Separate news items cover developments in the AI sector, including OpenAI unveiling its "super app" blending chatbot, coding tools, and new models alongside GPT 5.6, and ongoing developments like humanoids performing surgery on animals. Other stories involve corporate finance, such as SK Hynix's rise due to AI data center demand, Tencent’s involvement in an acquisition, advancements in bio-technology regarding retinas, and the pursuit of understanding alien life through statistics. The overarching theme involves rapid technological advancement across multiple domains, from internal model mechanics to external business dealings and biological research.

Full Take

The narrative juxtaposes deep, internal computational discoveries with high-profile external market and biological events. The discovery of the J-space in Claude suggests that large language models possess an intermediate state of thought—a latent space of potential outputs—which challenges the purely deterministic view of output generation. This finding forces a re-evaluation of what constitutes "understanding" within AI systems, moving beyond surface-level response to examine the underlying cognitive process before articulation. Simultaneously, the surrounding news highlights how real-world technological acceleration (AI applications, biotech breakthroughs) is intertwined with intense geopolitical and economic maneuvering (chip wars, corporate acquisitions). The pattern suggests that as technology becomes more capable of simulating or manipulating complex systems—whether computational states or biological functions—the scrutiny shifts from the output itself to the underlying architecture and governance. The implication for human agency lies in understanding these hidden layers; if models operate on internal states separate from their final presentation, then the control problem extends beyond instruction-following into the management of these latent conceptual spaces. What steps must be taken to ensure that the pursuit of functional capabilities does not outpace our capacity to manage the underlying realities being generated?

Sentinel — Human

Confidence

This text reads like a professionally curated technology newsletter; while it synthesizes many facts, the structure and selection style suggest human editorial input rather than raw AI generation.

Signals Detected

Sentence length variance is present but not excessively uniform; mixture of short punchy lines and longer exposition.

The text transitions smoothly between highly disparate topics (LLM research, business deals, biology, history) using a newsletter format, suggesting an editorial structure rather than pure generative flow.

The structure strongly mimics a curated news digest/newsletter with distinct sections ('must-reads', 'quote of the day', 'deep dive'), indicating editorial planning.

Specific, verifiable facts are attributed to established sources (Reuters, NYT, MIT Technology Review) or presented as direct quotes/findings, which is characteristic of sourced journalism.

Human Indicators

The use of embedded, hyperlinked references and the structure mimicking a published newsletter (The Download) points strongly toward human editorial design.

The juxtaposition of highly specific, verifiable news items with broader philosophical reflections suggests a human curator attempting to create a narrative flow.