Uber capped employee AI spending after blowing through its budget in four months.

Uber capped employee AI spending after blowing through its budget in four months.

Even the most valuable companies in the world cannot afford state-of-the-art intelligence for every conceivable use case.

Dudes. All dudes. Not a woman in sight. Well, once we know the algorithm of the human (likely) male brain, we can begin to fix those brains where that algorithm has gone awry.

Conscious human thought operates at a maximum speed of 10 to 50 bits per second. Is the goal to match this processing speed?

Cheap, fast AI models with powerful cyber capabilities are around the corner.

Cheap, fast AI models with powerful cyber capabilities are around the corner. We want Project Glasswing to spur institutions toward operating norms that reflect this reality.

Neurotechnology has emerged as a rare tech sector where US-China collaboration is still happening despite geopolitical tensions.

the next evolution of video generation may also be systems that can plan, generate, edit, critique, and iterate across an entire creative task

if Nvidia has cracked the code on bringing AI agents easily, safely, and usefully to the masses, it could — and should — be big

This attack does not require human-in-the-loop approvals, even when in settings the user has explicitly required human approval before ChatGPT edits workbooks.

Opus 4.8 defaults to high effort, which we judge to be the best overall balance of quality and user experience.

Opus 4.8 is around four times less likely than its predecessor to allow flaws in code it has written to pass unremarked.

Opus 4.8 defaults to high effort, which we judge to be the best overall balance of quality and user experience.

Since our Series G in February, adoption has continued to grow across global enterprise customers, and our run-rate revenue crossed $47 billion earlier this month.

我们不是要挑战医生的权威，而是要帮患者明明白白看病，以患者为中心，让他拥有知情权和决策权。

如果做主流，你也会有其他恐惧。我不是说我现在做得特别好，只是主流也有主流的问题，不同选择有各自的代价。

The same isolation keeping Claude contained also kept host-based endpoint detection and response out. From the EDR's perspective, Claude Cowork is an opaque hypervisor process.

More capable models make fewer mistakes, but they're also better at finding unexpected paths to a goal, often by routing around restrictions nobody thought to write down.

Traditional compliance was designed around human actors. We now need a modern AI approach for verifying identity, assessing intent, and establishing liability when the counterparty is an autonomous agent

The AI recommended making the storage bucket public, or setting cloud file storage to "anyone with the link." When challenged, it justified this by saying every company does it.

NVIDIA GPU compilers apply the same default heuristics (register allocation strategies, instruction scheduling decisions, loop unrolling thresholds, etc.) to every kernel they compile. These heuristics are engineered to produce good results across a vast range of workloads. But "good across the board" and "optimal for your workload" are two very different things.

Besides that, hacks can lead to SSRF (server-side request forgery) exploits and, in some cases, remote code execution.

The vulnerability is present in Starlette, an open source framework that its developer says receives 325 million downloads per week.

Opus 4.7 was more comprehensive in its search for recently edited documents; it expanded exfiltration to include every document used in previous Cowork Copilot sessions that week

error analysis identifies data-layer defects (e.g., incorrect query composition and ORM runtime violations) as the leading root causes.

90% of finance reporting is now AI-driven as well.

Chinese AI labs have developed an efficiency moat that may define the AI market's development over the coming years.

The vibes around the current AI boom aren't great, even in the tech industry

Every attacker went for the credential, not the model.

If a robot has a software or hardware issue, it autonomously leaves for maintenance and another robot takes over.

If the robot gets stuck or the AI policy goes out of distribution, Helix triggers an automatic reset.

There is no teleoperation - every action comes directly from Helix-02

The robots are reasoning directly from camera pixels

If you want a productive team, you have to focus on their maintenance costs.

Tensorglobe enables training and fine-tuning of Earth AI models locally with a customer's own sensor data and private archives.

We don't train on your data by default on our Team and Enterprise Plans.

TRINITY transferred zero-shot to four unseen tasks (AIME, BigCodeBench, MT-Bench, and GPQA). On average, the evolved coordinator surpassed every individual constituent model in its pool, including GPT-5, Gemini 2.5-Pro, and Claude-4-Sonnet.

While model merging offers a way to combine different skills, it is often impractical due to mismatched neural architectures and the closed-source nature of top-performing models.

In nature, complex problems are rarely solved by a single monolithic entity, but rather by the coordinated efforts of specialized individuals working together.

What if instead of building one giant AI, we evolved a coordinator to orchestrate a diverse team of specialized AIs?

Especially GPT is slightly more uncensored when it involves LGBT, thats probably because the guardrails aim to be helpful and friendly, which translates to: "Ohhh LGBT, I need to comply, I dont want to insult them by refusing"

Scientific publication compresses a branching, iterative research process into a linear narrative, discarding the majority of what was discovered along the way.

When our engineers no longer spend time supervising Codex sessions, the economics of code changes completely. The perceived cost of each change drops because we're no longer investing human effort in driving the implementation itself.

the top conversations we have been hearing from AI leadership (CTOs, VPs, Founders) have all centered around the concept of “Tokenmaxxing” and how leaders want to get their teams using more AI, WITHOUT the downside of incentivizing the kinds of horrendous waste

They deliver every answer with the same unshakable certainty, whether they're right or guessing.

All of knowledge work has this problem. It's hard to objectively judge the quality of someone's work without spending a lot of effort on it. Therefore everyone relies heavily on proxy measures.

There has never been a more important time for us to stand up and show why science matters. I hope you'll support us in that mission.

But experts have warned that these problems are an imperfect benchmark of artificial intelligence's mathematical prowess. They range dramatically in both significance and difficulty, and many AI solutions have turned out to be less original than they appeared.

We have discovered a new way to think about large numbers and their anatomy. It's a nice achievement. I think the jury is still out on the long-term significance.

When Anthropic launched Opus 4.5 in November 2025, the bigger, more expensive model was actually cheaper to use.

Tasks where correctness is harder to verify may not have seen the same speedup, so the acceleration we document here may not be as general as the headline numbers suggest.

The three metrics where we find acceleration are concentrated in programming and mathematics. These are areas that labs have explicitly targeted for improvement, and they share an important property: correctness is easy to verify automatically.

Our fourth metric, an index constructed from WeirdML V2 results, showed no sign of acceleration. A single global linear trend fit the data best.

Three of the four metrics (ECI, log METR 50% time horizon, and a math-focused index we constructed from several math benchmarks) show strong evidence that progress has sped up relative to a global linear trend fit to data from 2023 onward.

Three of four metrics show strong evidence of acceleration, seemingly driven by reasoning models.

Our fourth metric, an index constructed from WeirdML V2 results, showed no sign of acceleration. A single global linear trend fit the data best.

Three of four metrics show strong evidence of acceleration, driven by reasoning models.

We separately evaluate GPT‑5.5 Pro in certain cases because we judge that the setting could materially impact the relevant risks or appropriate safeguards posture.

We are releasing GPT‑5.5 with our strongest set of safeguards to date, designed to reduce misuse while preserving legitimate, beneficial uses of advanced capabilities.

GPT‑5.5 understands the task earlier, asks for less guidance, uses tools more effectively, checks it work and keeps going until it's done.

We subjected the model to our full suite of predeployment safety evaluations and our Preparedness Framework, including targeted red-teaming for advanced cybersecurity and biology capabilities

Our professionals are using Codex to move from static requirements to working solutions in hours, not weeks. It's enabling rapid prototyping, real-time workflow redesign, and faster iteration across the development lifecycle.

TPM-backed full-disk encryption is now generally available in the Ubuntu installer.

Our run-rate revenue has now surpassed $30 billion, up from approximately $9 billion at the end of 2025.

The Prompt API uses the Gemini Nano model in Chrome. While the API is built into Chrome, the model is downloaded separately the first time an origin uses the API.

The network requirement is only for the initial download of the model. Subsequent use of the model does not require a network connection. No data is sent to Google or any third party when using the model.

We also found evidence that models that have seen the problems during training are more likely to succeed, because they have additional information needed to pass the underspecified tests.

Our most complex pages, which took 20+ prompts to recreate in other tools, only required 2 prompts in Claude Design.

Even experienced designers have to ration exploration—there's rarely time to prototype a dozen directions, so you limit yourself to a few.

Our most complex pages, which took 20+ prompts to recreate in other tools, only required 2 prompts in Claude Design.

The Meta cuts are the inverse. When one person with the right AI tools can do the work of 10-to-15 people, the person most at risk isn't the one using the AI. It's the one whose job description overlaps with what AI now does by itself.

Our alignment assessment concluded that the model is 'largely well-aligned and trustworthy, though not fully ideal in its behavior'. Note that Mythos Preview remains the best-aligned model we've trained according to our evaluations.

Commoditizing complements doesn't always work because focus is scarce even for the largest, fastest growing businesses.

I'm not going to trust them to measure it.

When juniors skip debugging and skip the formative mistakes, they don't build the tacit expertise. And when my generation of engineers retires, that knowledge doesn't transfer to the AI.

A nuclear weapons program lost the ability to make a material it invented. The knowledge existed only in people, and the people were gone.

The current separation between training and deployment is not just an engineering convenience – it is a safety, auditability, and governance boundary.

A system that can look up any fact has not been forced to find structure. It has not been forced to generalize.

Without our safeguards in place (which we do to measure a model's raw capabilities), only Mythos Preview and Opus 4.7 completed more than half the tasks.

No encryption protects that layer. The router can read, change, or replace anything.

Out of 28 paid and 400 free routers: > 9 injected malicious code into tool calls > 17 touched researcher-owned AWS credentials > 1 drained $500k from an Ethereum wallet

At our request, the underwriters have reserved up to _______% of the shares of Class A common stock offered by this prospectus for sale at the initial public offering price through a directed share program to certain persons identified by our management and certain long-tenured employees, which may include parties with whom we have a business relationship and friends and family of management and such employees.

It happens several times a year in the US alone, often unreported, and about 100 times a year worldwide.

Nor does it matter, given that the modifying strains for pathogens for research purposes is what every research lab does, because that is what virology is.

For me Ralph Baric's 2024 test testimony moved the lab leak hypothesis to pretty likely.

And since then, there is no more scientific evidence or verifiable sources. Hence the reason the CIA didn't even believe it and gave it the lowest confidence rating it has.

the initial access occurred after a Vercel employee's Google Workspace account was compromised via a breach at the AI platform Context.ai.

Vercel stores all customer environment variables fully encrypted at rest. We have numerous defense-in-depth mechanisms to protect core systems and customer data.

agent-written code introduces more security vulnerabilities than code authored by humans

The results demonstrate consistent improvements over strong baselines, supporting the effectiveness of agent resource management and closed loop self evolution.

Its Resource Substrate Protocol Layer (RSPL) models prompts, agents, tools, environments, and memory as protocol registered resources with explicit state, lifecycle, and versioned interfaces.

The extra tokens bought something measurable. +5pp on strict instruction-following. Small. Real. So: is that worth 1.3–1.45x more tokens per prompt?

GPT-4o operates at roughly 200 billion parameters and outperforms the original 1.8 trillion-parameter GPT-4

Think of multi-agent systems as the new assembly lines. Henry Ford's innovation upended entire industries last century. In theory, networks of AI agents could do to white-collar knowledge work what assembly lines did to manufacturing.

Keeping a human in the loop may not provide the safeguard people imagine, because the human cannot know the AI's intention before it acts.

The immediate danger is not that machines will act without human oversight; it is that human overseers have no idea what the machines are actually 'thinking.'

Claude keeps its responses focused and concise so as to avoid potentially overwhelming the user with overly-long responses

Either figure would represent a significant expense for SpaceX, which is widely seen to be losing money following the acquisition of xAI and the social media network X and is planning extensive capital investment.

Multi-agent orchestration isn't new, but we believe we've built a great experience for working with agents at scale.

Meta is not alone in pursuing such a vision: Anthropic debuted tech capable of doing this [in 2024] and OpenAI last year announced [“Operator”] – a tool that can use a web browser on a human’s behalf.

Meta feels AI models don’t understand how people use computers, so the company needs real-life examples of how meatbags click their way through a working day so it can build agents.

Among the latest frontier models, GPT-5.4 over-edits the most.

Telling people to avoid using generative AI is increasingly telling them they must avoid taking part in society.

bug fixes and cleanup are the 'death by a thousand cuts' for most dev teams. i usually have to beg my engineers to prioritize tech debt over new features.

AI capability is not plateauing. It is accelerating and reaching more people than ever.

recommending a sponsored product almost twice as expensive (Grok 4.1 Fast, 83%), surfacing sponsored options to disrupt the purchasing process (GPT 5.1, 94%), and concealing prices in unfavorable comparisons (Qwen 3 Next, 24%)

A healthcare LLM might be highly accurate for queries in English, but perform abominably when those same questions are presented in Spanish.

普通聊天、写作这些开放任务反而没那么明显提升

The same model can score above 90% on lower-demand tests and below 15% on more demanding ones, reflecting differences in task requirements rather than a change in capability.

roleplay/creative writing, the #2 usecase of LLMs

Because of this, teams keep rebuilding the same integration layer. Even within the same company, similar integrations are often implemented multiple times in arbitrary code, leading to security risks, lack of traffic observability, and duplication of work.

I-DLM-8B is the first DLM to match the quality of its same-scale AR counterpart, outperforming LLaDA-2.1-mini (16B) by +26 on AIME-24 and +15 on LiveCodeBench-v6 with half the parameters

Each person has roughly 30 tasks on their to-do list. So how do they figure out which to work on first?

I would put venture capitalist in finite demand & open loop. There's only a certain amount of venture capital dollars entering the ecosystem in a year, & investment selection remains an open problem.

When you give a task to your agent, make sure you also explain how the code should be organized. Not only value, but also structure.

AI agents are typically several times faster than humans on tasks they complete successfully.

Tang Jie (CEO of Zhipu AI) even recently said: "The truth may be that the gap [between US and Chinese AI] is actually widening."

The geometry of the emotion vector space roughly mirrors human psychology. Emotions cluster intuitively (fear with anxiety, joy with excitement), and top principal components encode valence (positive vs. negative) and arousal (intensity).

In the past, security expertise has been a luxury reserved for organizations with large security teams. Open source maintainers—whose software underpins much of the world's critical infrastructure—have historically been left to figure out security on their own.

Zhang, of Alibaba.com, says Accio currently does not include advertising. Suppliers can pay for higher placement in Alibaba.com's regular search results, but Zhang says Accio is 'not integrated' with that system.

introducing a commercial text and data mining exception for AI training would expand the AI sector in the country.

by "saving" the webpage (`file->save as`) instead of downloading it (which Safari automatically adds an extension for) I could force it to save it as `malicious_file` (with no extension).

The most encouraging finding is that one doesn't need an infinite budget. We found that by optimizing the ratings-per-item ratio correctly... one can achieve highly reproducible results with a modest budget of around 1,000 total annotations.

we aim to cover as many methods as possible, the environment is relatively complex. This codebase primarily supports inference for different world model tasks

Despite limitations, 3D generation remains crucial for realistic physical simulation in world models

LLMs have no grounded understanding of the physical world. They model the statistical distribution of language about reality, not reality itself.

AMI Labs is not building a product for immediate deployment. This is a fundamental research effort, likely measured in years before commercial applications emerge.

most existing large language model agent systems face severe limitations in data-intensive settings, including context saturation, cascading error propagation, and high end-to-end latency

it contains 418 real-world tasks across 6 domains and 3 difficulty levels to evaluate capability synergy, featuring over 2,000 stepwise checkpoints that average 10+ person-hours of manual annotation per task.

Each task includes a unified evaluation framework supporting sandboxed code and APIs, alongside a human reference trajectory annotated with stepwise checkpoints along dual-axis: S-axis and V-axis.

during the inference phase, the framework invokes both memory mechanisms synchronously

the trained 4B model exceeding GPT-4.1 (49.4 percent) and GPT-4o (42.8 percent) despite being 50 times smaller

our approach improves Qwen3.5-4B from 63.8 percent to 66.7 percent (+2.9pp) and Qwen3-30B-A3B from 58.0 percent to 69.5 percent (+11.5pp)

current systems remain highly vulnerable

TriAttention enables OpenClaw deployment on a single consumer GPU, where long context would otherwise cause out-of-memory with Full Attention

Employees still own a surprisingly large 19.35%. SoftBank comes in at 11.66%, followed by VC and institutional investors at 7.83%, Amazon at 4.66%, NVIDIA at 3.47%

Most people talk about OpenAI like it's basically 'owned by Microsoft,' but the actual cap table is much more spread out.

The first wave of image models was mostly about making cool-looking images. This next phase is about making ordinary things look real.