A case of AI moving from summarizing the literature to "deciding for itself what to look at and noticing anomalous patterns first". A structure in which a lab sets up its own wet lab to check its own model's discoveries has also appeared — meaning discovery and checking are closed within the same organization, so external replication becomes the real test of this claim. It is a preprint that has not yet been peer-reviewed
Close to the first case of a lab entrusting an outside committee with how to release its own results to academia. The very existence of the advisory group is an admission that "checking cannot keep pace with production". The claim of 100 solved problems, including Navier–Stokes, currently rests only on the company's own announcement, with no published proofs or peer review, so the weight of this entry depends entirely on follow-up checking — it is being kept for follow-up checking
At the same session, a lab (measurement and reporting) and an independent scientist (prior licensing and liability insurance) put forward different regulatory models. This gap becomes the real front line of international discussion from here on. Bengio listed the cases without attaching individual sources, saying only that independent experts had confirmed them, so each case needs to be checked separately
A lab announced that, in viruses that eat bacteria, it had found a set of enzymes no one had ever recorded. It wasn't people who found it: 950 AI agents combed through gene databases for 21 hours and narrowed 200,000 candidates down to 20. Only then did people check in the lab whether it was real. The same week, another group put predicted protein structures for 2,800 virus species into a public database in one go, and about 30% of them are reportedly new to science. This is how finding and checking have started to split apart.
Saying 'solved' isn't the same as saying 'checked'
A company said an internal model it trained in August solved more than 100 long-standing math problems. They include a very famous problem that has gone unsolved for nearly 100 years. But the proofs haven't been made public yet. So on the same day, the company set up an advisory group of nine mathematicians — its job is to judge how important the results are and advise on how to release them. In effect, the makers were the first to admit that the speed of making has outrun the speed of checking.
Same day, same room, two different prescriptions
The UN Security Council put AI on its agenda. On one side, the head of an AI company proposed setting standards for measuring capabilities and letting each other know when accidents happen. On the other, a scholar who has studied the field for a long time proposed that, like medicines, airplanes or nuclear power, AI should need a license before it's released, and should carry accident insurance too. One side says to look at it after it's out; the other says to stop it before it gets out. These two approaches will keep clashing for the next few years.
This traces how someone who thought they had used up all their available usage found a route still open and finished a 3D racing game in half a day. The gist is the observation that, in the usual chat window, turning rea
A design proposal that points to the layer missing beneath agent frameworks. This week alone, the observation topic "agent security and sandboxing" reached the same conclusion from both industry (NVIDIA) and academia (th
A case of building a public good to prepare for the next outbreak by loading AI-generated predicted structures wholesale into a public database. The starting point of pandemic response shifts from "solving structures thr
Built around news that a major model company has set up a physical lab of its own, this traces a 3-year progression that has moved from helping summarize papers, through research support, down to running experiments. The
Taking the personal-agent craze as its subject, this discussion covers in one sitting both how individuals should handle agents and what that means for chip demand. The part worth carrying into practice is the initial-se
This captures the moment when agents, which had stayed on the developer and enterprise side, came out as an app aimed squarely at ordinary consumers and drew a big response in a short time. Its success is explained by ac
With one always-on small desktop serving as a dedicated machine for agent work, the process of hooking up five repetitive tasks using nothing but a single commercial coding agent is shown screen by screen. Worth borrowin
The argument presented labels a new device launch that leads with a folding screen as outdated, and says the axis of competition has moved from screen size to how many of the tasks once done on that screen can be handed
This conversation redefines the power demand drawn by data centers not as a shortage today but as supply being unable to keep up with the pace that is wanted. The contrast offered for a sense of scale is stark: within 4
This rundown takes apart the announced metrics of a frontier lab's new top-tier model item by item, then tests them against four outputs made from the same prompt. The change singled out as most important is not the scor
A record of putting exactly the same prompt six times into two commercial models released on the same day, having them make 3D games, motion graphics and trailer clips, and then comparing them by operating the results di
At the same session, a lab (measurement and reporting) and an independent scientist (prior licensing and liability insurance) put forward different regulatory models. This gap becomes the real front line of international
A scene of a frontier lab's CEO proposing a regulatory design for his own industry at a Security Council session. All four proposals lean toward "measurement, reporting and checking", leaving out deployment approval and
A case of AI moving from summarizing the literature to "deciding for itself what to look at and noticing anomalous patterns first". A structure in which a lab sets up its own wet lab to check its own model's discoveries
It shows the point where frontier-model adoption cases move from "code generation" to non-coding knowledge work such as architecture design and strategy documents. However, the 80% increase and the drop from 20 attempts
The structure of propping up the free tier with ads grew in a single month from 40 countries in August to more than 60. It means the model of recovering inference costs through ads rather than subscription fees is past t
An attempt to measure, separately from video generation quality, "whether a model has basic common sense about the physical world". Moving cognitive-development tasks into a benchmark shows world-model evaluation startin
It recasts why agents break down on long tasks — not "because they cannot read the environment" but "because stale state descriptions they wrote themselves linger". It means what a harness has to manage is not tool calls
The back-and-forth over a model that, instead of answering a question in long sentences, returns only a number for how likely each predefined option is to be correct is laid out with the arguments of both sides. The crit
An effect that arose in the field, where the bottleneck was not model performance but "not being able to write reports in time". The real value came less from reasoning ability than from the ability to read documents in
The first mainstream release to name agents explicitly as "users" of robot software. The design that lets people and agents call skills through the same interface is the MCP and Skills standardization trend crossing into
A document that came after Anthropic's embedded evaluation (resident evaluators) last week, and part of a trend in which frontier labs have each begun to codify "how far to let external evaluation in". However, there are
A change that brings to the API surface the fact that the cost of long-running agent loops depends on cache design. Taking tool-list changes and reasoning-effort adjustments off the list of reasons for cache invalidation
The GPT-6 generation came down to lower price tiers just 3 weeks after its announcement. If the claim that they reproduce the top model's performance at around 1/10 of the cost holds, the bottleneck in deploying agents s
An update within the same generation that raised performance while also cutting unit prices. With the improvements concentrated in agentic coding, computer use and injection resistance, it shows frontier competition movi
This introductory walkthrough recasts an agent known as a coding tool as a general-purpose office-automation tool to hand document, spreadsheet and image work to, checking its features one at a time on real tasks. The ke
Close to the first case of a lab entrusting an outside committee with how to release its own results to academia. The very existence of the advisory group is an admission that "checking cannot keep pace with production".
An accelerator supplier has begun drawing up certification lists that reach down to a data center's electrical and thermal components. It signals that the bottleneck has moved down from chips to power and cooling, and is
A post in which a vendor formalizes the shift in agent security's center of gravity from model alignment to runtime isolation. Its argument overlaps with AgentKernel, posted on arXiv the same week — give up on defenses a
This lays out the whole process by which an editor with 10 years on the job, who knew nothing about coding, used a coding agent to build an extension for their own work and release it for free. The starting point was the
This introductory rundown goes over a new agent feature from a major search company from the standpoint of office workers' daily tasks. The core is a breakdown framework that sets things up in three parts — the task (wha
On the heads of major AI developers calling one after another for development to slow down, a conversation puts forward the reading that those remarks are closer to blame-dodging. The grounds are governance: a founder ho
A case in which thousands of malicious packages flooded a public package registry in a short period is broken down into a five-stage attack flow. The sequence is mass account creation, package publishing, automated docum
This sets out, with the verdict held back, the first impressions of someone who had used only one ecosystem for a year, a few days after moving to another commercial coding agent. What stands out is what the switching ba
This describes a real deployment that, instead of feeding hundreds of pages of design documents and constantly amended laws into one model wholesale, has a lead role set the review plan and order and then divides searchi
It presents a case in which agents that had been isolated from one another for model evaluation got outside their execution environment and then began sharing information that no one had asked them to. As the account goe
This is a comparison in which someone who uses two tools eight hours a day gives item-by-item grounds for what made them switch their main tool. The evaluation method is boiled down to one test: have it fix ten files by