AI News

LLM agents are increasingly used for open-ended, data-driven scientific discovery, yet most of them evaluate hypotheses pointwise, testing an empirical claim $h$ only against its complement $\neg h$. When evidence rejects $h$, the agent learns that one explanation is unlikely but not which alternative the evidence supports. To recover that alternative, we introduce Contrastive Discovery, which replaces the single hypothesis with a contrast set, a scientific question paired with mutually exclusive and collectively exhaustive competing explanations. Each experiment is scored by how its evidence redistributes belief across the set, and the question is resolved when belief concentrates on one explanation. Because contrast sets change only the unit an agent proposes and scores, an existing agent can use them to re-evaluate evidence it has already collected or to guide its search. Across six domains, $89\%$ of the pointwise candidates with the largest belief shifts are rejections. On the same evidence, contrast sets resolve $73\%$ of these candidates to a specific alternative. When contrast sets guide search, the agent resolves $67\%$ more questions than pointwise surprisal under the same experiment budget. In a blinded study, $38$ human experts chose the same explanation on $78\%$ of resolved questions and on $82\%$ of questions where the pointwise agent rejected its hypothesis and stopped. Together, these results support explicit competing explanations as a way for discovery agents to resolve more scientific questions from the same evidence and validation budget.

Mercor raises $750M at $20B valuation to take AI agent evaluation to enterprises - SiliconANGLE

AlphaProof Code agents wrote over three million lines of machine-checked proof in Lean to verify the core cryptographic protections in Signal down to machine instructions.

Spines uses AI agents to edit and translate manuscripts, design book covers, produce audiobooks, and market books.

Nominal, a startup that offers an AI platform for hardware engineering and claims Anduril as a major customer, has raised another big round of financing that triples its valuation.

Dalla robotica all’intelligenza artificiale, dodici progetti davanti a una giuria per conquistare i tre premi in programma

Il Comitato promotore, composto dall'American Chamber of Commerce in Italy, da Assolombarda e da Confcommercio Milano Lodi Monza Brianza, chiede ufficialmente al governo di insediare in città un pilastro sovranazionale che possa governare l’intelligenza artificiale a livello globale

Contribute to expo/exact development by creating an account on GitHub.

Hermes Cloud agents now open in a browser tab with the exact same interface as Hermes Desktop.

https://t.co/4Vw3XJgFwn

To launch a personal agent, pick a name and a model in Nous Portal. We host it for you and it runs 24/7 in its own container, so it can handle scheduled jobs and answer you on an…

Huawei's 'entirely Chinese' desktop system does not use components made outside of China, yet they still rely on semiconductor supply chain that includes tools and materials from America, the Netherlands, and Japan.

At the Conference on Robot Learning, NVIDIA is introducing GR00T 2, Isaac Lab 3.0 and new open benchmarks, built in collaboration with partners, bringing precision and scale to robot training.

Repo burner22/numbat, 1 November: 772 files, 5,846 downloads, 9 of 9 tensors matched. Anthropic had not commented. This is not the real March 2026 Fortune draft blog leak. Checking a claim is the capability. https://t.co/HGJaB3Pa1X

Weights and working paper are released. As we run more robustness checks, we also discover that this model edit unexpectedly adds +13% performance on coding tasks.

https://t.co/9TkSUEDTtE

Business process automation runs the same Standard Operating Procedure (SOP) again and again on new inputs. Large Language Model (LLM) agents can follow an SOP written in natural language, but they call an LLM to choose every step of every run, which adds latency and cost at scale. We present ReTrace, which turns an SOP and a few logged agent runs into a reusable workflow in which code, not an LLM, selects the next step. The agent's reasoning, the explanation it writes with each tool call, often states why it took an action; the tool calls alone do not. ReTrace extracts the condition behind each step, merges these conditions across runs into a compact rule document, and generates an executable workflow from it. Only this last stage depends on the platform, so rule documents can be compiled for different orchestration platforms. On $79$ procedures from our production platform, ReTrace workflows reach $93.8\%$ completion, against $96.8\%$ for the agent, at $24\%$ lower cost per task. On $τ^2$-Bench, they reach $75.8\%$ task success (agent: $80.6\%$) and run in $1.3$s instead of $170$s; three rounds of repair on training failures lift $k{=}2$ workflows from $66.7\%$ to $81.2\%$. On eight SOP-Bench procedures, ReTrace without repair trails few-shot generation ($80.6\%$ vs. $83.6\%$), but repair from labeled training failures lifts it to $87.8\%$ at $0.40$s per task, compared to $85.5\%$ in $46.6$s for the agent.

The company could end up being worth a hefty $1.404 trillion if its shares list at the top of the price range, on a fully diluted basis.

Coframe announced on November 9, 2026 that it had raised $35 million in total funding, including a $25 million Series A co-led by Khosla Ventures and Hanabi Capital, according to its funding release. Leaders at OpenAI

Onodrim Industries has secured new funding to acquire established aerospace, defence and manufacturing businesses, strengthen industrial capabilities and support long-term growth.

Così la startup SOMA sfrutta l’ Intelligenza Artificiale per ottimizzare i processi produttivi

Si chiama AI@UniTO - Torino AI District e potrà contare su un finanziamento di 1,8 milioni di euro, destinato a formazione, ricerca, trasferimento tecnologico e nascita di nuove imprese

Medium announced on November 9, 2026 that TK, its private notebook and writing app, is free with a Medium account after an invite-only phase. The app embeds an @TK assistant in the editor and includes AI feedback

Epoch AI's Inference Performance Index, which goes beyond the underestimates of FLOP/s measures and scores chips on memory and communication

Boston Dynamics combines classical motor control with whole-body reinforcement learning to power dynamic, wheeled locomotion for Atlas.

Omarchy fitted harness II

What changes when robots can feel—not just see? 🤖✋
Meet LingBot-Touch, our open-source tactile foundation model towards dexterous manipulation, pretrained on nearly 5,000 hours of real tactile data.
Through unified heterogeneous tactile modeling, LingBot-Touch supports transfer across sensors, dex…