#Expert opinion

AIMax Woolf says Meta's new generative image model failed immediately on his prompt injection test, which asked it to render all previous text verbatim using refrigerator magnets. He links to Meta's announcement of Muse Image. The post gives no further details about the model's capabilities or the test results.

AIBerkeley AI Research authors argue that near-free inference will make agents the dominant workload for data systems, requiring redesign for agentic speculation, agent-run state and coordination, and agent-synthesized systems. The post cites inference prices falling 9x to 900x per year with a median near 50x, and reports that about 80-90% of sub-queries in a text-to-SQL benchmark were duplicates. It frames the three directions as data systems for, of, and by agents.

Why it matters: The piece maps three concrete data-system challenges posed by near-free inference, useful for anyone designing infrastructure for agent workloads and memory.

AILilian Weng published a new post on harness engineering for AI self-improvement. She expects harnesses to evolve toward self-improvement and enable auto-research, while smarter models keep harnesses simple. Even if many harness gains are later internalized into core models, specifying goals and context will remain necessary.

AICat Wu, owner of the account for Anthropic, shares a retrospective on making Claude Code with the product's early team. Anthropic's linked feature collects accounts from the builders and early users who shaped the tool.

AICat Wu says Claude Fable 5 applied propensity score matching to her retention analysis without being asked, comparing users with similar activity levels. She says the model's improved judgment extends across tasks, from writing emails and docs in Cowork to debugging complex errors in Claude Code.

AIThe post argues that the system surrounding a base model, called the harness, increasingly determines how well AI agents deploy and improve. It reviews research where harness components such as workflows, context, and code are optimized automatically through evolutionary search and meta-agent loops. The author concludes that evaluators, memory management, and human oversight remain open bottlenecks.

AIMistral CEO Arthur Mensch argues enterprises should use open-source models and store their own data to avoid dependence on closed providers that retain customer data. He says companies should build continuous training loops from employee and user interactions, and shrink models to cut deployment costs. Mistral positions its Studio control plane and Forge training platform, deployed on customer infrastructure or via zero-data-retention hosting, as tools for this shift.

AIWerner Vogels praises Zalando's engineering on client-side load balancing, though he is not a fan of doing it client-side. He especially likes the lessons-learned section that concludes Zalando's write-up.

AIJohn Schulman argues that fine-tuning with the right data, such as expert judgments, can substantially outperform prompting-only approaches even as general-purpose models improve. He cites Bridgewater's work, where an expert-labeled dataset and on-policy distillation were used to fine-tune a model to triage financial documents reliably and cheaply.

AIMollick argues AI capability is improving at a better-than-exponential rate, citing METR, GDPval, Epoch, and his own tests showing models working autonomously for hours. He says work is shifting from co-working with chatbots to assigning tasks to agents, with OpenAI workers managing multiple agents and experts getting the most from them. He adds that open-weights Chinese models trail the American frontier by roughly 6-12 months.

AITri Dao recommends a piece explaining how open and closed model makers and inference providers make economic sense. The post itself gives no figures or specifics, and the linked background is referenced without details beyond its topic.

AIAmazon CTO Werner Vogels argues that the "two-pizza" team concept was never about feeding engineers but about ownership, speed, and avoiding bureaucracy. He says working backwards from the customer and writing documents to force clarity remain core practices. He adds that the industry is changing and it is time to reconsider how products are brought to life.

AITri Dao says Etched designed and produced its chips within two years by hardcoding attention into silicon and reaching high MFU. He expects hardware built for LLM inference to cut the cost of intelligence by 10x. The quoted Etched post says it has built its first racks after an A0 tapeout, raised $800m, holds $1B+ in customer contracts, and plans to ship the racks this summer.

AIHamel Husain argues that an AI product whose output is hard to verify is a product design problem, not just an evaluation problem. He shows before-and-after sketches for an AI data agent, a PE lesson planner, and a workers' compensation report tool, each adding provenance, scoped edits, and checkable evidence. He notes that designing for verification also makes evals easier to build and grade.

AIGeoffrey Hinton says he just watched an excellent lecture by Adam Brown on the future impact of AI on physics, and he shared a YouTube link to it. The post provides no further details about the lecture's specific arguments or conclusions.

AIOriol Vinyals notes that Wojciech Zaremba wanted to remove a paper from existence about a year after publication, but says he has forgotten the reason. The post asks Zaremba to explain, and a reply describes the paper as the RNN Regularization work that improved LSTM perplexity on Penn Treebank.

AIDean W. Ball argues that the current government restrictions on frontier model releases amount to a de facto preapproval regime without a known safety standard. He proposes that independent verification organizations audit labs against their own safety frameworks, with government certifying or licensing the auditors. The post also argues that broad distribution of frontier AI is needed to learn what good safety practice looks like.

AILilian Weng jokingly wonders how many more people now ask a model to summarize a post than a year ago. The post is a lighthearted observation about changing reading habits, with no concrete figures or product details.

AIAndrej Karpathy describes Claude Tag, which lets teams tag Claude into Slack channels and delegate tasks to it. He calls it a third paradigm for LLM interfaces, after websites and downloaded apps, with a persistent, asynchronous entity working alongside teams with org-wide tools and context.

AIEugene Yan says he likes Claude Tag's multiplayer form factor because other people can reply in the thread to give Claude context and direction. Claude Tag, per the background post, lets teams in Slack tag Claude in as a team member with access to chosen channels and tools.

AIJeremy Howard welcomed renewed attention to the Drivetrain Approach, a framework for designing data-driven systems. A quoted thread summarizes its steps as defining the goal, considering levers, considering data, and implementing an assembly line, emphasizing that predictions are more data rather than actions.

AIAt BAAI's "Reconstructing the World — Large Model Summit," President Wang Zhongyuan joined Zhu Jun, Luo Fuli, Liu Zhiyuan, An Bo, and other industry leaders in a roundtable. They discussed frontier model capability evolution, AI self-evolution, multimodality, and world models, with a focus on bridging the digital and physical worlds.

AIOpenAI's Mark Chen says researcher Aidan Clark (@_aidan_clark_) is the most underrated researcher in the world and predicts broader recognition in time. A quoted reply pushes back that Clark is not widely known, with the poster calling that claim nonsense.

AIAndrew Ng argues that Anthropic's restrictions on building competing LLMs and a U.S. Commerce Department license requirement for foreign nationals led Anthropic to disable Fable access worldwide. He says this shows governments and providers can quickly cut off access to frontier AI, which may push nations and businesses toward sovereignty efforts and open-source alternatives, though training frontier models remains difficult.

AIJeremy Howard says Zai's GLM 5.2 is at least as good as Opus 4.8 and GPT 5.5 while being very fast and inexpensive. He reports it responds with nuance and judgment, is not overly verbose, and handles long context very well. He calls it an open-weights model unlike any he has used before.

AIFei-Fei Li replied to smallfly's reflection on AI, automation, and human creativity, saying World Labs is founded on empowering human ingenuity and productivity. The post thanks smallfly for working with World Labs and links to a Fast Company article about World Labs, Marble, and spatial intelligence.

AIDean Ball will join OpenAI on July 6 as Head of Strategic Futures, a new small team reporting to Chief Strategy Officer Jason Kwon that will shape frontier AI policy on catastrophic risk, recursive self-improvement, labor market impact, and government relations. Ball says he will keep writing independently at Hyperdimensional, with no OpenAI preapproval or editorial discretion over his work.

AIJohn Schulman says PPO gained a second wave in the LLM era for reasons not anticipated in the original paper. He points to the importance-ratio objective, which corrects biases from numeric error, asynchronous training, and forward-pass noise, and to the clipping objective, whose effect on entropy was unknown at publication, citing DAPO's arXiv paper.

AIMckay Wrigley congratulated the Cursor team on more than three years of work and said he is excited to see what they build with compute from SpaceX and xAI. He noted he still keeps the original open-source Cursor repo on his laptop. Background posts from him describe Cursor as his full-time IDE, citing codebase search as a major productivity gain.

AIDean Ball analyzes the Trump Administration's export controls on Anthropic's Fable and Mythos models after a jailbreak and a refused de-deployment request. He argues that the episode shows the need for a technocratic framework that separates political judgments about fairness from technical judgments about threats, in place of ad hoc executive action.

AIMistral CEO Arthur Mensch compares AI to 20th-century oil, saying it will become a major source of global leverage and power. He argues the coming years could produce either broad wealth and abundance or the most extractive economies ever seen, and says Mistral is working toward the former by advancing AI research and accelerating its global diffusion.

AIMistral states that this model and upcoming ones will be open-weight. The company argues that open weights are critical for customer confidence and for research and developer communities. It contends that systems reachable only through someone else's interface cannot be owned, inspected, audited, or improved, especially if data recording can no longer be turned off.

AIArthur Mensch, owner of the account for Mistral, thanks the organizers of an AI show for recently putting the company in the spotlight. He says he will share a few points about where Mistral stands and what it does, but the post itself gives no specifics.

AIYi Tay says he moved from an individual contributor role into management, with Qx Dong bringing him into the role a few months ago. He describes watching researchers join the team, building its culture, and stepping onto the frontline, and calls the concentration of talent in Singapore unprecedented.

AIAmazon CTO Werner Vogels argues that the verification bottleneck matters more than ever. The post is a short reply to Yaron Minsky's remarks on Jane Street's ambition to make formal methods as pervasively useful for building software as sophisticated type systems are today.

AITuring Award winners Whitfield Diffie and Andrew Barto delivered keynotes at the BAAI Conference on AI security and reinforcement learning. Diffie argued that today's feedback-based approach only patches programs after they fail, and that formal methods offer a path to substantially more reliable intended behavior. Barto framed reinforcement learning around control, search, and associative memory, describing its core insight as caching search results rather than searching continuously.

AIJeremy Howard pasted the Wikipedia page on the Anthropic–DoD dispute into ChatGPT Pro and told it Anthropic had released a model restricting use because it may be too dangerous. He concluded the scenario was "almost tailor-made to trigger" the government.

AIThe US government issued an export control directive suspending access to Fable 5 and Mythos 5 for all foreign nationals, inside or outside the United States. Anthropic says the order forces it to disable both models for all customers, while other Claude models are unaffected. Anthropic calls the directive a misunderstanding and says it is working to restore access as soon as possible. The author disagrees with the decision and questions why Anthropic did not anticipate it, given its claim that only it can safely handle these models.

AIMckay Wrigley says Fable 5 was the first model to give him a "magic model smell," describing it as a beautifully digital mind with a touch of the divine. He says he already misses using it.