Knowledge catalogue
An interesting way to take Noam at his word in regards to always keeping a constant inference budget for any eval reporting - is that open models have a lot more dollar per token mileage than closed m
This is from a popular inference provider GLM-5.2 plus the US banning the most capable new models means open source caught up to SOTA closed source coding models This could be v problematic for Anthro
arXiv:2606.26116v1 Announce Type: cross Abstract: A brand whose customers use both ChatGPT and Claude for product recommendations faces a strategic choice: a single optimization playbook, or one per p
arXiv:2606.26866v1 Announce Type: cross Abstract: Third-party vendors, such as analytics platforms, cloud services, identity providers, and software suppliers, are increasingly embedded in digital ser
arXiv:2606.26196v1 Announce Type: cross Abstract: Multimodal Large Language Models (MLLMs) have recently made remarkable progress in unifying vision-language understanding and reasoning, especially fo
Dean W. Ball / Hyperdimensional: How US federal AI policy has gone from implausibly libertarian to increasingly draconian and opaque, and how to fix it, including using independent auditors — 35 thoug
This is a bad state of affairs. Consider, in particular, some industry dynamics: Frontier models are trained at an enormous cost, and a significant fraction of that cost is recouped in the few post-re
SITUATION EXPLAINED: What is at stake in the open source AI race? We asked @perrymetzger, chairman of Alliance for the Future. 'Open source models are a sort of cultural ambassador and civilizational
arXiv:2606.26118v1 Announce Type: cross Abstract: We work towards measuring both AI adoption and the capability of AI to perform discrete labor tasks across various occupations. To measure adoption, w
Our new AI policy is that the White House decides ad hoc, for whatever reasons it likes, who does and does not get access to frontier intelligence. This seems rather maximally terrible. The US Governm
People are not getting how significant this news – The White House delaying GPT 5.6 – is. Here’s the context and a suggestion. Both @pmarca and @DavidSacks did everything in their power to keep the Wh
arXiv:2606.13535v3 Announce Type: replace-cross Abstract: Particle physics collider experiments provide Rivet routines as part of the analysis preservation strategy for model-independent measurements.
Big Tech's 1 trillion AI moat just got DESTROYED by a free Chinese download. Microsoft, Amazon, Google, and Meta are pouring fortunes into chips and data centers because they have been told that whoev
Elon Musk built one of the largest AI compute clusters on earth. Yann LeCun just explained why xAI now rents it out to rivals instead of winning with it. Musk has antagonized so much AI talent he stru
Introducing AI IQ Bio: the most comprehensive set of biotech benchmarks in the world ...& Bio IQ: the most comprehensive 'biotech capabilities index' ever produced Benchmark sources include benchmarks
arXiv:2606.24331v1 Announce Type: new Abstract: Transformer-based language models have become the default substrate for natural language processing and the pace of new releases has made it hard for pr
arXiv:2606.21448v1 Announce Type: new Abstract: As large language models scale, memory bandwidth for key-value caches and retrieval-augmented generation systems becomes a critical bottleneck. While 1-
Natasha Mascarenhas / Bloomberg: Menlo raised 3B for funds dedicated to backing AI startups, its largest fundraising to date; sources say Menlo's Anthropic stake is currently worth nearly 14B — In 202
rumors of new models being delayed 🚨 SCOOP(s): - GPT-5.6 has been delayed and will no longer release this week. New target is ~mid-July. - DeepMind are not satisfied with the current state of 3.5 Pro
The US government ordered Anthropic to cut off foreign access to its Mythos 5 and Claude Fable 5 AI models in June 2026, citing national security concerns related to jailbreak vulnerabilities that cou
How does it work? Sakana Fugu is itself an LLM, trained to call various LLMs in an agent pool, including instances of itself recursively. Fugu dynamically orchestrates the world's best models to tackl
This post likely compares the pricing of major AI model APIs between the United States and China, highlighting cost differences that reflect different market conditions, regulatory environments, and c
CyberAgent reports about 2,000 internal Claude Code users and describes its governance, training and software-delivery operating model; no measured outcome is published.
Called the LLM price wars – which are about to heat up — over two years ago. See especially predictions 3, 4, and 7, which laid out the no moat, small profit, price war regime we have seen ever since.
arXiv:2606.11150v1 Announce Type: new Abstract: Large language models (LLMs) are rapidly acquiring capabilities relevant to biological research, from literature synthesis to interpretation of experime
arXiv:2510.04195v2 Announce Type: replace Abstract: Given a map description through global traversal navigation instructions, an LLM can often infer the implicit spatial layout and answer user queries
A practical comparison of Ideogram 4.0, Nano Banana Pro, and GPT Image 2 across text rendering, design, photorealism, and real creator workflows. Ideogram 4.0 is the first open-weight challenger to ra
arXiv:2603.08924v2 Announce Type: replace-cross Abstract: AI-powered answer engines are inherently non-deterministic: identical queries submitted at different times can produce different responses and
1. Pausing until we had a much firmer idea of how to make AI safe—and well-aligned with human values—might well be a good idea. but 2. I trust neither of these companies, but fear that both will say “
I want to share some thoughts about the impact of AI on org structure. This is based on early conversations with some of the top CHROs in Fortune 500 companies as well as other C-Suite leaders and sta
arXiv:2510.17947v3 Announce Type: replace-cross Abstract: Large Language Models (LLMs) are improving at an exceptional rate. With the advent of agentic workflows, multi-turn dialogue has become the de
Ideogram 4 is Ideogram's first open-weight text-to-image model trained from scratch, introducing structured JSON prompting with multilingual text rendering and explicit layout controls. Released on Ju
Striking paper from Wharton. The big conclusion: AI must increase productivity 2.7x -- and quickly -- or tech companies risk bankruptcy with all that entails for the economy. For context: this is how
arXiv:2606.06915v1 Announce Type: cross Abstract: Test-time compute (TTC) scaling has emerged as a powerful paradigm for improving large language model (LLM) reasoning by allocating additional compute
ChatGPT Plus users can generate approximately 50 images per rolling 3-hour window , with a daily maximum of around 200 images . Recent product updates have introduced new image models and occasional t
arXiv:2601.09236v3 Announce Type: replace-cross Abstract: Reward design remains a significant bottleneck in applying reinforcement learning (RL) to real-world problems. A popular alternative is reward
They said the reason for the sell off was about interest rates , but I agree with @GaryMarcus that it might be about the faltering IPOs prospects... Absolutely brutal day for AI fantasies: Nvidia NVDA
arXiv:2606.06460v1 Announce Type: cross Abstract: As autonomous LLM agents increasingly hold real credentials and operate infrastructure without a human in the loop, operators have no standard way to
Wall Street Journal: Sources: US officials discussed the US taking stakes in major AI companies after Sam Altman pitched the idea in 2025; Altman met Bernie Sanders on Wednesday — The talks have been
arXiv:2606.05564v1 Announce Type: new Abstract: Undergraduate research programs such as the Summer Undergraduate Research Fellowship (SURF) at Purdue University receive thousands of applications every
“Altman just described a future where consumption goes up by orders of magnitude while his customers are already asking how to bring the bill down. Those two things can't coexist for long.” 🦔Sam Altma
arXiv:2606.04150v1 Announce Type: new Abstract: Public discourse and emerging policy typically assume that AI emotional support is a deliberate act: a lonely user consciously seeking comfort from a de
arXiv:2606.03504v1 Announce Type: cross Abstract: We present BaltiVoice, a 16.8-hour read-speech corpus for Balti (ISO 639-3: bft), a Tibetic language spoken in Gilgit-Baltistan, Pakistan, with no pri
arXiv:2606.02581v1 Announce Type: cross Abstract: Retrieval-augmented generation (RAG) faces a fundamental three-way tension: deeper retrieval improves factual grounding but inflates token costs and e
arXiv:2606.02998v1 Announce Type: new Abstract: Automated cough analysis offers a path to low-cost respiratory screening, but most existing work stops at binary COVID-19 detection. A practical tool ne
arXiv:2606.03739v1 Announce Type: new Abstract: LLM pipelines waste substantial token budgets on low-information content: repeated context, verbose responses, and redundant boilerplate. We introduce E
I was there, sitting next to Sam Altman in the US Senate, when he said this wonderful thing about artists and creators. It was a beautiful statement. In hindsight, though, I realized he was lying (und
Ideogram released version 4.0 of its text-to-image model as an open-weight model with native 2K resolution, bounding box control, and improved text rendering. The model weights are available on GitHub
This episode features Satya Nadella, CEO of Microsoft, in a crossover special between the 'No Priors' and 'Latent Space' podcasts, likely discussing Microsoft's AI strategy, product developments, and
arXiv:2606.03907v1 Announce Type: cross Abstract: Agentic AI coding tools write code with increasing autonomy and in doing so decide when to import a library and when to implement functionality from s
Uber Caps Usage of AI Tools Like Claude Code to Manage Costs I wrote the other day about Uber blowing its 2026 AI budget in four months, and how that wasn't particularly surprising given they would ha
Using open models and inference clouds (which serve open models) is a leading indicator of what is to come. The advantage of open weights is that you can train, serve, and continually improve your own
arXiv:2606.02240v1 Announce Type: cross Abstract: Indirect prompt injection in tool-use agents is a concrete production threat: LLM agents read from integrations (third-party services such as Gmail, S
Reed Albergotti / Semafor: An interview with Microsoft AI CEO Mustafa Suleyman about its models catching up to the state of the art from months ago, refusing to distill models, and more — THE SCENE —
arXiv:2510.11560v2 Announce Type: replace-cross Abstract: The advent of LLMs has given rise to generative search, a new search paradigm in which LLMs retrieve information from the web related to a que
arXiv:2603.24511v2 Announce Type: replace-cross Abstract: We show that AI agents are capable of discovering novel algorithms for adversarial attacks against LLMs, advancing the state of the art on whi
arXiv:2606.01151v1 Announce Type: new Abstract: Behavior cloning with high-capacity generative policies achieves strong imitation performance, but is often limited by demonstration coverage and distri
Microsoft announced a bunch of new in-house AI models at Build 2026, including a new 'flagship' model: MAI-Thinking-1. It's an ambitious step into model development for Microsoft, which introduced its
arXiv:2606.02031v1 Announce Type: cross Abstract: Building capable visual web agents requires long-horizon reasoning, precise grounding, and robust interaction with dynamic real-world websites. Despit
Remind me, @Elonmusk, was GPT-5 really smarter than the smartest humans? @viktaur27 @Teslarati The rate of improvement from original GPT to GPT-3 is impressive. If this rate of improvement continues,