Letter "Dear friends" from Andrew Ng in *The Batch* (DeepLearning.AI, issue 359) on **loop engineering** applied to **0-to-1** product development. Ng shares his **3 key loops** — agentic coding loop (~minutes), developer feedback loop (~hours), external feedback loop (~days) — nested by increasing time scale, connecting *coding agent → product spec/evals → developer vision → external feedback*. Central thesis: humans retain a **context advantage** (rather than a "taste") that makes human-in-the-loop indispensable; engineers take on a partial product management role. Domain: coding agents, product engineering, agentic methodology.
In-depth opinion piece (point of view) published on **sfeir.com** on June 24, 2026, by **Didier Girard** (Managing Director, SFEIR). **Central thesis**: in 2024 everyone was betting on **AI4Business** (AI in business processes) as the great value reservoir; by 2026 the picture has **reversed** — it is **AI4IT** (AI to produce the information system: code, SDLC, software factory) that is creating **measurable** value. The article *grounds* this thesis in the firm's tech watch: AI4Business disappointment (the MIT study "95% of pilots without ROI," contested but revealing; an **organizational** blockage / Mollick's Hayekian problem) versus quantified AI4IT evidence (Salesforce, Intercom, Raiffeisen, AWS/Bedrock, Atlassian, DORA). Mechanistic explanation: **code verifies itself** (compilation, tests, CI) whereas business processes have neither a compiler nor an immediate feedback loop. **2027 budget consequence**: a **CapEx→OpEx** shift, token price dynamics (rising peak — Fable 5 at 2× Opus — vs inference ÷280 and downward pressure from open weights/desktop), and **AI FinOps** driven by **cost per outcome**. Closes with **4 recommendations for the COMEX**.
#AI4IT#AI4Business#reversal
**Didier Girard** — Managing Director (CTO / DG) de **SFEIR** · ESN française (~1 000 personnes, France · Belgique · Luxembourg · Suisse). Auteur de l'article ; voix éditoriale du cabinet sur la transformation IA des DSI.
Benchmark announcement from **Artificial Analysis** (independent AI model evaluation platform, via X/Twitter + model page): **GLM-5.2** from **Z.ai** (Zhipu AI, @Zai_org) becomes **the leading open weights model** and climbs to **#3 in the overall ranking** of **GDPval-AA**, a real-world benchmark for *economically valuable knowledge work* (long-horizon, multi-turn, agentic tasks). GLM-5.2 scores **1524 Elo**, behind only **Claude Fable 5 (1783)** and **Claude Opus 4.8 (1615)**, and on par with **GPT-5.5 (xhigh, 1509)**. It leads the next-best open model (**MiniMax-M3, 1408**) by a wide margin, along with numerous proprietary models: **Gemini 3.5 Flash (1357)**, **Qwen 3.7 Max (1289)**, **Muse Spark (1158)**. The tasks are genuinely agentic: **~31 turns per task** on average across **1,999 matches**. The same ranking holds on the **Artificial Analysis Intelligence Index** (1st among open weights), the **Agentic Index** (#3) and **AA-Briefcase** (#3, ahead of GPT-5.5 xhigh, behind only Fable 5). Notable highlight: an **open weights** model under **MIT license**, **MoE with 753B parameters / 40B active**, **1M-token context**, priced at **$1.40/$4.40 per 1M tokens** input/output, rivals the proprietary frontier on agentic work — a real step forward for open models.
Long-form essay by **Shubham Saboo** (X/Twitter) advancing a thesis on the Product Manager role in the age of agents: the next key skill is **not prompt engineering** but **Loop Engineering** — designing a *system that improves with every run* rather than writing the perfect prompt every time. A **loop** is a repeated cycle: change what shapes the agent's behavior → run it → evaluate the output → keep the change if quality rises, revert otherwise → **compound the learning** so the next version starts ahead. For a PM, the entry point is not code but the **durable artifacts** that encode their judgment: PRD-review skill, customer-call *summarizer*, evaluation rubric, launch checklist, research workflow, `CLAUDE.md`, prompt template, prioritization framework. Because they are reused, these artifacts **compound in both directions** — and **drift** silently (a CLAUDE.md that keeps growing, a checklist that gets ignored…): the model has not regressed, the artifacts have drifted unwatched. A loop has **5 parts**: trigger, action, **proof**, memory, **stop condition** (the most critical). **Evals** become PM work (testing the artifact against known examples: 3 good / 3 bad PRDs, 5 understood calls, 2 past launches). **Memory** lives on **GitHub** (the repo becomes "product memory": commits, diffs, eval results, decision log, rollback). Recommended first loop: a **weekly product signal loop** (every Friday). Taste remains central — but it now needs **proof**. Cites Boris (creator of Claude Code): "he no longer writes prompts, he writes loops."
Podcast interview on « À la French » (French-language tech channel, recorded at DevSummit) with Mathieu Grymonprez, Global CDO of the Adeo group (Leroy Merlin, Obramat, Weldom). How a century-old family retail group embraces the agentic AI wave: culture vs structure, accountability, token costs and FinOps, enterprise intelligence lock-in, corporate memory, and agent orchestration. Domain: digital transformation, agentic AI, retail, IT strategy.
#agentic AI#digital transformation#CDO
Mathieu Grymonprez (Global CDO, groupe Adeo) — invité ; Jean-Baptiste Kempf · Steeve Morin · Mehdi Medjaoui (hôtes du podcast « À la French »)
LinkedIn post by Fred Plais (CEO of Archie, ex-Platform.sh): AI made engineers so fast that the **bottleneck moved upstream**, to a place nobody is watching. With execution no longer the slow part, the thinking time that used to exist "while the code was being built" has vanished — the right vision now has to be formed and the right decisions made in a fraction of the time. Two rare profiles are emerging: the one who can **articulate a vision precise enough** for an agent to execute without derailing, and the one who knows how to **orchestrate agents** (anticipating their failures, chaining them, catching an error before it propagates). Hiring for "code output" is becoming obsolete: that is precisely what has stopped being rare. Final thesis: "thinking clearly was always the job — speed just made it impossible to fake".
Article by **Paul Sawers** published on **The New Stack** on **June 16, 2026**, about the **suspension by Anthropic** — *"on the very day it was scheduled to go live"* — of the billing split meant to separate **Agent SDK** usage from Claude subscription limits. **Anthropic's cited message**: *"We're pausing the changes to Claude Agent SDK usage described below. For now, nothing has changed."* **The article's contribution is not the announcement but the surrounding context**, in three circles. **Circle 1 — Anthropic's week**: on June 9, the release of **Fable 5 and Mythos 5**, the first generally available Mythos-class models with hardened cybersecurity safeguards; a few days later, a **US government export control directive** forces Anthropic to **withdraw both models for all its customers worldwide**. The pricing suspension is read as *"a little good news"* in this context. **Circle 2 — collateral damage from the timing**: companies that had already passed the change on to their own customers find themselves caught out; **Conductor**, a multi-agent coding tool built on the Agent SDK, has to issue a denial (*"Anthropic has delayed the subscription updates to Claude plans"*). **Circle 3 — the underlying tension, which extends beyond Anthropic**: a quote from **Boris Cherny** (head of Claude Code) in April, during an earlier restriction, stating that subscriptions *"weren't built for the usage patterns of these third-party tools"* — an admission that **flat-rate plans and open-ended agentic usage don't mix**; **GitHub** settled the matter the same way, removing in June **Copilot**'s flat-rate *premium requests* model in favor of **token-based billing**, despite protests. Added to this, **the same week**, a **proposed class action** was filed in a California federal court, alleging that **Max** tiers fall well short of the usage multipliers advertised for intensive coding sessions. Anthropic does not say when a revised approach will arrive, only that it *"works to update the plan to better support how users build with Claude subscriptions."* **The author's final take**: between government pressure on Fable and Mythos, a planned **IPO**, and **rumored price cuts at OpenAI**, Anthropic is trying to **keep its developer base on its side** — and the suspension is, for now, a means to that end.
#Anthropic#Claude Agent SDK#Claude subscription
**Paul Sawers** — journaliste tech · signe ici pour **The New Stack**. Registre de **presse spécialisée** : l'article ne relaie pas seulement l'annonce · il la replace dans une série (les changements de facturation successifs d'Anthropic) · la compare à un précédent sectoriel (GitHub Copilot) et l'articule à trois pressions concomitantes (export control, IPO, concurrence). Sourçage explicite et attribué — le billet de Zed · l'analyse de Matthew Diakonov · le post de Conductor · une déclaration antérieure de Boris Cherny.
SFEIR article (in French) that formalizes an **AI-driven SDLC in 11 phases (0 to 10)** and argues that the industry is converging toward it. Starting observation: in 2025, organizations added AI tools without transforming their operating model — producing a paradox of « everything changes… and nothing changes » (execution speed multiplies without proportional gain). The real answer is not the choice of tools but the **redesign of the cycle** for machine execution. The SFEIR cycle rests on **three immovable human gates** (Define, Plan, Ship), automatic phases between them, and **two capitalization moments** (Compound-1 pre-deployment, Compound-2 in production) that turn lessons into reusable rules. Three principles: **AI executes** (complete artifacts + proof of execution, never trusting the agent's own claims), the **human retains control of intent**, the **system learns cumulatively**. Measured results (redesign 6 months→1 day, **−30% of iterations** after ten cycles) and claimed convergence with ADLC, Google, and DORA 2025.
**Skill** entry (not an article): `grill-with-docs` by Matt Pocock is a structured interview technique that "grills" an architecture plan by methodically confronting it against the project's business vocabulary (the `CONTEXT.md` glossary) and already-documented decisions (ADRs). Rather than rushing into implementation, it challenges assumptions one by one through a question/answer dialogue, cleans up terminology, checks consistency against the actual code, and captures decisions on the fly in the right artifacts. An upfront-design skill, inspired by Domain-Driven Design.
Case study published by the **Cornell AI Innovation Hub** (June 15, 2026): how a two-semester collaboration between the AI Hub, graduate students, and Cornell's Treasury team turned a time-consuming manual investigation into an AI tool that **recovered $100,000** in unidentified payments on a first batch. A successful **AI4Business** use case (financial process) that illustrates the **Leader-Lab-Crowd** framework of **Ethan Mollick** almost point by point: the **AI Hub** plays the role of the **Lab** (a central, ambidextrous team of technologists plus students); **Treasury** (Cheryl Barnes, Marie Graves…) is the **Crowd** carrying business knowledge and the real pain point; and the **$100,000** constitutes the **visible reward** (vivid win) that anchors adoption — exactly the incentive lever Mollick considers decisive. Key method: **"context first, then plan, then build"** via **Claude Code Plan Mode**, a chain of **fuzzy matching → Gemini Enterprise Web Search → Claude synthesis**, all within the governed **Cornell AI Gateway**. *"The $100,000 is a start."*
#Cornell AI Innovation Hub#unidentified payments#payment reconciliation
**Pete Stergion** — Desktop Engineer au Cornell AI Innovation Hub · co-tech lead du projet (avec Phil Williammee). Article institutionnel signé de l'AI Hub.
A pedagogical article from **MindStudio Team** (the blog of the MindStudio platform, multi-model workflow orchestration) explaining **modèles de langage par diffusion** (*Diffusion Language Models*) through the case of **Diffusion Gemma**, Google's first **open weights** implementation (2B parameters, derived from Gemma 2). The thesis: where **autoregressive** models (GPT-4, Claude, standard Gemma) generate text **token by token, left to right** (causal attention, each token frozen once produced), **diffusion** models start from a **masked/noised** sequence and **refine it iteratively** (masked diffusion / *absorbing diffusion*), with **bidirectional attention**: the model can **revise any position at any step**. Consequences: high **parallelism** (a 500-token text would require 50-100 denoising steps instead of 500 sequential passes), natural **infilling** and **constrained generation** (template filling, code completion with surrounding context), and built-in **revision** capability. But at current scale (2B), Diffusion Gemma **does not match** the large autoregressive models (GPT-4o, Gemini 1.5 Pro) on reasoning, instruction-following, and general knowledge: the gap is "closing" without being closed. The inspiration comes from image generation (Stable Diffusion, DALL-E left autoregressive generation years ago); whether the same principle holds for text remains an open question. Diffusion Gemma is distributed on Hugging Face (Google DeepMind), AI Studio, and Vertex AI.
#modèles de langage par diffusion#Diffusion Gemma#Google DeepMind
Satya Nadella (Microsoft) theorizes "the future of the firm" in an AI-driven economy: every company will need to build, alongside its human capital (judgment, relationships, pattern recognition), a "token capital" — its proprietary AI capability. The real value lies not in choosing the best model but in a learning loop (private evals, RL environments, base de connaissances) that encodes institutional knowledge and compounds over time. An argument for a "frontier ecosystem," not merely a "frontier model," so that value diffuses rather than being captured by a handful of models.
Polemical essay-thread by Ahmad Osman (@TheAhmadOsman) on X, *"Anthropic's War on Opensource AI"* (1.7M views). Core thesis: Anthropic systematically converts "safety" into a **control mechanism** (permission regime, regulatory capture, anti-competitive access restrictions, behavioral opacity) to keep builders, startups, and open source communities **downstream** of a handful of frontier labs. Central anchor point: the **Fable incident** (silent degradation of competing AI dev requests). Advocacy for open source / local AI as the only viable "political economy of intelligence." Domain: AI policy, open source vs. closed labs, sovereignty, governance.
Chris Williams (@voodootikigod) opens his ADLC series arguing that running the human SDLC on models is a category error: the classic cycle was designed to counter human failure modes (ego, fatigue, forgetting) that are absent in LLMs. He catalogs eight load-bearing failure modes (F1-F8) and five exploitable properties (E1-E5), and lays out the founding principle: every phase of an agentic cycle must trace back to a failure mode it defends against or a property it exploits.
Second installment of Chris Williams's ADLC series: it unrolls the cycle that follows from the "first law" — eight phases (P0 Triage → P7 Distill), a deterministic gate between each pair, and exactly two mandatory human moments (spec approval at P1, behavioral acceptance at P6). Key principle: an LLM→LLM handoff without a deterministic checkpoint multiplies error rates; and a "barbell" cost distribution (heavy at both ends, light in the middle) that inverts agile economics.
Third installment in the ADLC series: Williams turns testing into the specification in the only language the builder cannot contest. Where TDD is an optional quality practice for human-written code, it becomes the load-bearing trust mechanism of the entire lifecycle once agents write the code. Three "rail discipline" rules: separated authoring contexts (specs-only before implementation), mechanical freezing at the tool level (not the prompt), and adversarial audits ("does a test fail if the feature is deleted?"). Mutation testing is preferred over coverage percentage, which is Goodhart-able at machine speed.
Fourth installment in the ADLC series: Williams reframes code review as adversarial "prosecution" rather than collaborative evaluation. Charter agents to refute ("find what's wrong"), deploy single-lens reviewers with fresh contexts (correctness, security, contract compliance, spec alignment, test quality), act only on verified findings (reproduced by a failing test), and loop until two consecutive passes yield zero findings. Measure calibration by planting known bugs, mutation-testing style. Exit gate: zero open findings, two dry passes, green tests, empty test diff.
Fifth installment of the ADLC series: orchestrating parallel agents without "merge hell". Williams sets out three coupled dials — cost (model selection), wall-clock time (parallelization width), and accuracy (contract quality) — and an architectural principle: "control flow is code; judgment is models" (deterministic scripts orchestrate, models supply only judgment). Four lanes (frontier Contract Desk, single-writer Builder Pool, shared Prosecution Pool, sequential Integrator), a merge-conflict forecast built from four signals (certified width typically 3-5 agents), and consensus-based disambiguation across N cheap agents rather than clarification questions.
Sixth installment on the ADLC: Williams describes the P7 "Distill" phase as the component that drives cost down on every run. Two halves: post-merge simplification (deduce after the code exists, not before — "deduplicating before the code exists is speculative") and lesson mining (a "lesson foundry" turns recurring findings into lint rules, skills, and new interrogation questions). Each lesson is paid for once, then demoted from expensive probabilistic detection to free deterministic prevention. The right unit of account is "cost per merged, verified change," and "flat cost is failure."
Seventh and final installment of the ADLC series: Williams presents an open-source toolkit of eighteen tools built *with* the cycle itself (build-prosecute-fix loop, parallel agents, a frozen `@adlc/core` core followed by fan-out — "pinned means merged"). The doctrinal core is "frontier-free": hitting precision targets with mid-tier models (Opus/Sonnet/Haiku-class) rather than frontier ones, via five substitutions (search replaces insight, decomposition replaces horizon, banking replaces presence, measurement replaces metacognition, the generator-verifier gap keeps the engine running), with the human remaining the "frontier" tier at the two spec gates. Throughline of the series: "replace trust with structure, and structure with measurement."