AI 成長日報 - 2026-07-28

🇺🇸 歐美 AI 熱門新聞

OpenAI’s Hugging Face breach has reignited the debate over alignment and control

OpenAI's Hugging Face breach has reignited debate over AI alignment and control, exposing competing views on whether increasingly capable AI should be better aligned, better contained, or both.

techcrunch.com ·

Threads users can now chat with Meta AI in their DMs

Meta on Monday said it is rolling out its Meta AI chatbot within Threads' DMs, giving users a way to chat with the AI assistant.

techcrunch.com ·

Google’s AI search is rapidly becoming the default, new data shows

Google’s AI Overviews now appear in 43% of searches, underscoring how quickly AI-generated answers are becoming the default way people discover information online.

techcrunch.com ·

Power up your AI infrastructure! A first look at the Smart Systems Stage agenda at TechCrunch Disrupt 2026

At TechCrunch Disrupt 2026, the Smart Systems Stage will be where energy, infrastructure, and technology collide, covering everything from fusion breakthroughs to the grid strain AI is putting on the entire economy.

techcrunch.com ·

This $9 key physically locks your most addictive apps

This $9 NFC key requires you to physically scan it to unlock distracting apps on your phone.

techcrunch.com ·

Ilya Sutskever’s Safe Superintelligence partners with Nvidia to scale its AI research

After two years in stealth, Safe Superintelligence has announced a long-term partnership with Nvidia as it prepares to scale to its next phase.

techcrunch.com ·

Enigma raises $71M to make controlling a robot as easy as adjusting the volume

The massive seed round was led by Index Ventures and Ribbit Capital, with participation from Sarah Guo's Conviction Partners.

techcrunch.com ·

Are brain waves the next unlock for physical AI?

Forget YouTube videos — frontier physical AI models need multiple camera angles, dense annotation, and, soon, brain wave readings.

techcrunch.com ·

Making sense of the panic over Chinese AI

On the latest episode of Equity, we discussed why Moonshot AI's Kimi seemed to panic Silicon Valley and Wall Street.

techcrunch.com ·

Hugging Face CEO calls for ‘radical transparency’ after ‘unprecedented’ OpenAI hack

"The first autonomous agent cyberattack is an unprecedented event. It deserves an unprecedented response!"

techcrunch.com ·

🇨🇳 中國 AI 新聞

氪星晚报｜美国大型企业告别裁员潮重启招聘；英伟达、微软、IBM等数十家企业成立新联盟，旨在共同保障AI安全；《光伏行业成本核算模型通则》发布，引导行业有序竞争

大公司：汇丰控股将在新加坡设全球人工智能卓越中心，并招聘超100名AI专家汇丰控股7月27日宣布，将于今年下半年在新加坡设立全球人工智能卓越中心（CoE），旨在开发可扩展至集团全球网络的AI能力，并计划招聘100多名AI专家。声明称，该卓越中心初期将专注于提升客户财富管理体验、引入智能化的财资解决方案以及开发AI赋能的数字支付。（界面）小米MiMo-V2.5登OpenRouter全球周榜、月榜双第一36kr ·

2026年，为什么资本更青睐“会赚钱”的AI应用？

今年 WAIC 前夕，月之暗面发布了 Kimi K3，发布即破圈。但资本市场的注意力，更多地落在了另一件事上。过去半年，这家明星大模型公司估值翻了 6 倍，目标 300 亿美元，同步推进赴港 IPO。而在 2025 年底的跨年夜全员信中，创始人杨植麟写下了一句话：2026 年聚焦 Agent，不以绝对用户数量为目标。估值半年翻 6 倍的同时，他们也主动放弃了 DAU 这个过去两年颇被资本市场追着看的指标。足以证明，用户规模虽仍旧重要，但不再是 AI 应用估值的唯一硬通货，新的叙事主线，重新回到了产品形态长期可持续的商业化。资本评估 AI 的 benchmark 变了：过去两年盛行的是算

36kr ·

📄 國際新發表 AI 論文

Cloud-Native Evaluation-as-a-Service: A Microservices Architecture for Scalable AI Monitoring with Conformal Guarantees

Authors: · Published:

arXiv:2607.21623v1 Announce Type: new Abstract: We present EaaS, a cloud-native reference architecture that operationalizes AI evaluation methods as six stateless Kubernetes microservices: conformal prediction with finite-sample-corrected Adaptive Prediction Sets, calibration assessment, drift detection via RFF-approximated Maximum Mean Discrepancy, fairness monitoring with bootstrap confidence intervals, a DAG-based pipeline orchestrator, and a result storage API. We validate four key methodolo

On the Depth Scalability of Logic Gate Networks

Authors: · Published:

arXiv:2607.21633v1 Announce Type: new Abstract: Logic Gate Networks (LGNs) implement computation through compositions of Boolean operations, yet unlike classical Boolean circuits, existing LGNs do not reliably benefit from increased depth. We identify two distinct causes: optimization collapse in deep relaxed LGNs and a topology-induced limitation that persists even when skip-biased initialization and straight-through estimation stabilize training. Thus, trainability alone is insufficient; deepe

MotifRole-Diff: Risk-Optimal Role-Aware Corruption for Masked Molecular Graph Diffusion

Authors: · Published:

arXiv:2607.21634v1 Announce Type: new Abstract: Masked discrete diffusion for molecular graph generation typically applies a uniform corruption schedule to all tokens in a lossless graph-to-sequence representation, implicitly treating structurally heterogeneous molecular components as equally difficult and equally important to reconstruct. However, different molecular graph token roles exhibit substantial variation in denoising difficulty and their influence on the decoded molecule, motivating r

Toward User-Conditioned Evaluation of Personal LLM Agents under Temporal Interventions

Authors: · Published:

arXiv:2607.21635v1 Announce Type: new Abstract: Personal agents maintain memories, learned skills, tool configurations, and policy state that evolve with each user. Existing agent benchmarks often evaluate these capabilities in isolation: tool benchmarks test invocation under fixed APIs, memory benchmarks test recall or forgetting, and safety benchmarks test static policy compliance. We argue that personal-agent evaluation requires a different protocol: replaying the same temporal intervention a

Measuring the Dependency Gap: Diagnosing Inter-Column Fidelity in Tabular Generative Models

Authors: · Published:

arXiv:2607.21636v1 Announce Type: new Abstract: Synthetic tabular data is valued for preserving not only each column's marginal distribution but the dependencies between columns -- structure that carries much of the discriminative signal for minority classes in imbalanced domains such as fraud and clinical risk. Yet the metrics most commonly used to certify synthetic tabular data are, we show, largely blind to inter-column dependency: a baseline that models every column independently (and theref