🥔 OpenAI's 'Spud' reshuffles the AI frontier

^**[Read Online](https://www.therundown.ai/p/openai-spud-dethrones-claude-on-the-frontier?_bhlid=4eaa234680798ad3bdb45514b5705f029891f88c&jwt_token=eyJ0eXAiOiJKV1QiLCJhbGciOiJIUzI1NiJ9.eyJzdWJzY3JpYmVyX2lkIjoiYjBhNzU2NWUtZDJhMS00MGE2LTg5ZmMtNGUxMWJkYTQ5MWE1IiwicHVibGljYXRpb25faWQiOiI0ZDAzMzkwZC0yNDgxLTQyOTktYjk0OS1mZmQ4YjM4YjRjMzgiLCJhY2Nlc3NfdHlwZSI6InJlYWQtb25seSIsImV4cCI6MTc3NzE5ODAyOCwiaXNzIjoiaHR0cHM6Ly9hcHAuYmVlaGlpdi5jb20iLCJpYXQiOjE3NzcwMjUyMjh9.aUGTixuce6kf1UIAWpYa_b5TKI3sGexc57hfmmrTLJA)**^^ | ^^**[Sign Up](https://www.therundown.ai/subscribe)**^^ | ^^**[Advertise](https://therundownai.typeform.com/to/kraZ1TSO)**^

----------
View image: (https://media.beehiiv.com/cdn-cgi/image/fit=scale-down,format=auto,onerror=redirect,quality=80/uploads/asset/file/986504a9-f6e2-41c0-a7b7-3f322080a797/orkes-header.jpg?t=1776963966)
Follow image link: (https://orkes.io/webinars/the-4-layer-production-stack-every-ai-agent-needs ai&utm_source=Newsletter&utm_medium=email)
Caption:

----------

----------
**Good morning, AI enthusiasts.** The AI frontier doesn't stay settled for long — and this week, the pendulum is swinging back toward OpenAI.

The company’s new GPT 5.5 'Spud' model just capped a big week of releases with a jump up the leaderboards and a strong vibe shift… Landing at the exact moment Anthropic absorbs its roughest week of rate-limit and quality complaints in months.

_P.S. We love hearing how our readers are implementing AI into their daily lives. _[_Share your community workflows here_](https://www.videoask.com/ffohxg3vc)_ (with a new audio submission option and interview flow) for a chance to be featured in an upcoming newsletter. _

———————————————————————————

**In today’s AI rundown:**

* OpenAI retakes the frontier with GPT 5.5

* U.S. flags Chinese labs’ 'industrial-scale' AI theft

* Get a newspaper brief every morning with Claude

* AI's biggest productivity winners are also most worried

* 4 new AI tools, community workflows, and more

----------

----------
**LATEST DEVELOPMENTS**

----------

----------
###### OPENAI

#### 🥔 **[_OpenAI retakes the frontier with GPT 5.5_](https://openai.com/index/introducing-gpt-5-5/)**

View image: (https://media.beehiiv.com/cdn-cgi/image/fit=scale-down,format=auto,onerror=redirect,quality=80/uploads/asset/file/0dabd845-e3c7-41f3-8836-90af8c16272f/samapotato.jpg?t=1776980910)
Caption: Image source: Images 2.0 / The Rundown

**The Rundown: **OpenAI just [launched](https://openai.com/index/introducing-gpt-5-5/) GPT-5.5 (codenamed 'Spud'), the company’s long-awaited upgrade, pitched as a 'new class of intelligence' — topping benchmark scores across the industry and overtaking Anthropic on the AI model frontier.

**The details: **

* 5.5 sets highs across a series of reasoning, agentic, computer use, and coding tests for public models, with several scores [comparable](https://x.com/scaling01/status/2047378968412598688?s=20) to Claude Mythos.

* The model keeps the same speed as 5.4 with added efficiency, with OAI saying it used Codex and 5.5 to rewrite its own GPU code to improve infrastructure.

* GPT-5.5 lands at $5/$30 per million input/output tokens for API pricing, with OAI pitching it as 'half the cost of competitive frontier coding models.”

* 5.5 is rolling out across ChatGPT plans and in Codex with Thinking and Pro variants, with OAI continuing to highlight ‘generous usage’ for its new releases.

**Why it matters: **After months of Anthropic dominance, the vibe is shifting once again — with OpenAI rapidly shipping powerful new upgrades and rekindling the magic that felt a bit lost on previous releases. With Anthropic now wading through rate limit and quality degradation complaints, it’s a big week for Sama and co. on the sentiment front.

----------

----------
###### TOGETHER WITH ORKES

#### **🛠️**** **[_**Build durable agents with Agentspan**_](https://orkes.io/webinars/the-4-layer-production-stack-every-ai-agent-needs ai&utm_source=Newsletter&utm_medium=email)

View image: (https://media.beehiiv.com/cdn-cgi/image/fit=scale-down,format=auto,onerror=redirect,quality=80/uploads/asset/file/53c635bf-fec3-4a39-aa5e-2297ec6eb8a6/image1__36_.jpg?t=1776964022)
Follow image link: (https://orkes.io/webinars/the-4-layer-production-stack-every-ai-agent-needs ai&utm_source=Newsletter&utm_medium=email)
Caption:

**The Rundown:** Agentspan is an open-source framework and runtime designed for building, running, and observing durable agents. Join the upcoming webinar to see how modern engineering teams are leveraging Agentspan to build resilient agents that hold up in the real world.

**The April 30 session will cover:**

* The 4-layer production stack every AI agent needs for durability at scale

* How to make existing frameworks durable, including LangGraph, OpenAI Agents SDK, and Google ADK, using Agentspan

* Real-world patterns for keeping agents alive when processes fail

_[Register now](https://orkes.io/webinars/the-4-layer-production-stack-every-ai-agent-needs ai&utm_source=Newsletter&utm_medium=email)_.

----------

----------
###### AI & GEOPOLITICS

#### ⚖️ [_U.S. flags Chinese labs’ 'industrial-scale' AI theft_](https://www.whitehouse.gov/wp-content/uploads/2026/04/NSTM-4.pdf)

View image: (https://media.beehiiv.com/cdn-cgi/image/fit=scale-down,format=auto,onerror=redirect,quality=80/uploads/asset/file/003cd17d-6f4e-4de5-bde4-9ad612ac29ae/16a577c8-dbcd-4d65-98f6-21c04e0bc36a.png?t=1776983135)
Caption: Image source: Images 2.0 / The Rundown

**The Rundown: **The White House [published](https://www.whitehouse.gov/wp-content/uploads/2026/04/NSTM-4.pdf) a memo accusing Chinese firms of 'industrial-scale' distillation campaigns against U.S.-based frontier AI labs — coming weeks before Trump's scheduled Beijing summit with Xi Jinping.

**The details:**

* Distillation is training smaller AI systems on frontier model outputs, with Kratsios saying China runs it via thousands of fake API accounts and jailbreaks.

* Anthropic accused DeepSeek, Moonshot, and MiniMax of distillation in February, with this memo upgrading the private complaint to federal policy.

* The Chinese embassy dismissed the accusations and called them 'pure slander’, coming ahead of Trump and XI’s meeting in Beijing on May 14-15.

* A House Foreign Affairs bill that cleared its first vote this week would push the administration to add distillation offenders to the U.S. export blacklist.

**Why it matters: **Dario Amodei recently framed open-source and China as '6-12 months behind' frontier labs. The Kratsios memo reframes that gap, arguing the gains come from scraping tactics, not architecture work. Whether that holds depends on how much of the DeepSeek/Kimi trajectory truly traces to distillation vs. new research.

----------

----------
###### AI TRAINING

#### 📰 **_[Get a newspaper brief every Morning with Claude](https://app.therundown.ai/guides/get-a-personal-newspaper-written-by-claude-every-morning)_**

View image: (https://media.beehiiv.com/cdn-cgi/image/fit=scale-down,format=auto,onerror=redirect,quality=80/uploads/asset/file/da5d93ca-d90d-46a2-aaba-b1dfd1e00204/CleanShot_2026-04-23_at_16.28.39_2x.png?t=1776980419)
Follow image link: (https://app.therundown.ai/guides/get-a-personal-newspaper-written-by-claude-every-morning)
Caption:

**The Rundown:** In this guide, you will learn how to turn your daily updates into a personalized newspaper with Claude. It's useful because it gives you a one-ranked brief with the stories, action items, and calendar prep that matter.

**Step-by-step:**

1. In Claude or Claude Cowork, make sure Slack, Notion, Gmail, and Calendar are connected

2. Prompt: “Create a static Morning Edition from my Slack, Notion, Gmail, and calendar updates from the last 24 hours. Rank what matters most and format it like a newspaper with top stories, action items, and schedule prep”

3. Review the first draft and give feedback. Tighten the order, layout, and emphasis. Then tell Claude to turn the workflow into a skill

4. In Claude Cowork, create a recurring task that runs that skill each morning

**Pro tip:** Have other agents gather outside news and drop them into a Notion database. Then use this skill as the editor to turn everything into a daily newspaper.

----------

----------
###### PRESENTED BY OPTIMIZELY

#### **🎓**** **[_**Build 3 AI agents in 5 days for free**_](https://www.optimizely.com/ai-marketing-certificate/)

View image: (https://media.beehiiv.com/cdn-cgi/image/fit=scale-down,format=auto,onerror=redirect,quality=80/uploads/asset/file/5ac4833c-9dc5-4843-9174-31b2b6244477/image1_-_2026-04-23T120740.134.png?t=1776964079)
Follow image link: (https://www.optimizely.com/ai-marketing-certificate/)
Caption:

**The Rundown:** It’s time to put AI to work inside your marketing; embedded in your workflows and eliminating the drudge work. Consider this your official invite to Opal U: AI Marketing University from the team at Optimizely Opal.

**Here’s what you’ll get:**

* Live workshop (free) with 50 senior marketers - 1 hour a day (Mon-Fri)

* 3 working AI agents you build and take back to your team

* 5 hours invested to save 10+ hours every week

_[Apply for the next Opal U session starting Monday](https://www.optimizely.com/ai-marketing-certificate/)_.

----------

----------
###### AI RESEARCH

####  📊 **[_AI's biggest productivity winners are also most worried_](https://cdn.sanity.io/files/4zrzovbb/website/3a8d990bc90098038eabd77b0d12ff636ed58d50.pdf)**

View image: (https://media.beehiiv.com/cdn-cgi/image/fit=scale-down,format=auto,onerror=redirect,quality=80/uploads/asset/file/1557e4f5-89a1-437f-b6d1-a623bbbbb8bc/anthropicindexfear.jpg?t=1776970159)
Caption: Image source: Anthropic

**The Rundown: **Anthropic [published](https://cdn.sanity.io/files/4zrzovbb/website/3a8d990bc90098038eabd77b0d12ff636ed58d50.pdf) the economic-focused follow-up to its [81K-user Claude survey](https://www.therundown.ai/p/how-81k-people-really-feel-about-ai), finding that the people getting the biggest productivity lift from AI are also the most worried about losing their jobs to it, especially early-career workers.

**The details: **

* The survey ties Anthropic's Economic Index usage data (which jobs lean on Claude most) to 80,508 workers' takes on how AI is reshaping their roles.

* Workers whose jobs use Claude most voiced AI displacement fears, 3x more than those whose jobs use it least, with engineers leading the anxiety.

* Most respondents said AI's gains land on themselves via faster tasks and free time, but also lead to expanded scope and more work.

* Early-career respondents voiced the loudest displacement fears, backing Anthropic's earlier signal of a hiring slowdown for recent grads in the U.S.

**Why it matters: **The conventional view is that AI panic would come from lower-level adopters, but these results flip that, with anxiety coming from those getting the most out of the tools. Despite the productivity boosts, AI’s sentiment has never been lower — and there doesn’t seem to be many solutions in sight for easing the tensions.

----------

----------
**QUICK HITS**

----------

----------
### 🛠️ _**[Trending AI Tools](https://www.rundown.ai/tools)**_

* 🥔 [GPT 5.5](https://www.rundown.ai/tools/gpt-5-5) - OpenAI’s new top-rated flagship AI model

* ⚙️ [Ultrareview](https://code.claude.com/docs/en/ultrareview) - Claude Code command to run multi-agent code reviews

* 🏥 [ChatGPT for Clinicians](https://www.rundown.ai/tools/chatgpt-for-clinicians) - Free clinical version for verified U.S. doctors

* ⚙️ [Qwen3.6-27B](http://www.rundown.ai/tools/qwen3-6-27b) - Alibaba's new 27B model that out-codes its predecessor

----------

----------
### 📰 **_[Everything else in AI today](https://www.rundown.ai/)_**

**Band** launched the missing infrastructure layer for multi-agent systems. Any agent. Any environment. _[See it in action](https://www.band.ai/product)_.*

**Anthropic** [published](https://www.anthropic.com/engineering/april-23-postmortem) a post-mortem tracing Claude Code quality complaints to three separate bugs, resetting usage limits for subscribers due to the issues.

**OpenAI** [introduced](https://openai.com/index/making-chatgpt-better-for-clinicians/) ChatGPT for Clinicians, a free tool for verified U.S. health workers, with GPT-5.4 scoring 59.0 on HealthBench Pro, topping physicians and Opus 4.7.

**Meta **[sent](https://www.businessinsider.com/meta-plans-layoff-10-of-its-entire-staff-may-2026) an internal memo to employees informing them that the company is laying off 10% of its workforce in May, citing AI efficiency and other investments.

**Elon Musk’s SpaceX** is reportedly [in talks](https://www.businessinsider.com/elon-musk-xai-explored-collaborating-with-mistral-cursor-2026-4) with French AI startup Mistral on a three-way partnership alongside its recent deal with coding startup Cursor.

**Tencent **[open-sourced](https://hy.tencent.com/hy3-preview) Hy3 preview, its first model from a rebuilt training stack with competitive agentic coding and search-agent scores to top open models.

_*Sponsored Listing_

----------

----------
**COMMUNITY**

----------

----------
### 🤝 _**[Community AI workflows](https://www.videoask.com/ffohxg3vc)**_

Every newsletter, we showcase how a reader is using AI to work smarter, save time, or make life easier.

Today’s workflow comes from reader **Mark M.** in Virginia:

"I used AI to turn a tax-season headache into a streamlined system. Every year, the same scene plays out: I drop off bags of donations, stash the receipt, and then scramble at tax time trying to decipher my own list and determine each item's worth.

This year, I photographed my handwritten donation lists and uploaded them to Claude AI. Claude transcribed my lists, asked clarifying questions, verified charities with the IRS, grouped items into logical categories, applied condition-appropriate FMV, and produced professional PDF summaries — all conversationally, in minutes.

Multiple donations and nearly 400 items in fully documented and defensible deductions. For anyone who donates, this is the workflow you didn't know you needed."

How do you use AI? Tell us [here](https://www.videoask.com/ffohxg3vc).

----------

----------
### **🎓 **_**[Highlights: News, Guides & Events](https://app.therundown.ai/)**_

* Read our last AI newsletter: [Anthropic’s locked-down Mythos leaks](https://www.rundown.ai/articles/anthropic-locked-down-mythos-leaks)

* Read our last Tech newsletter: [Apple gets a new boss](https://www.rundown.ai/articles/apple-gets-a-new-boss)

* Read our last Robotics newsletter: [Sony’s new robot has a killer backhand](https://www.rundown.ai/articles/sony-new-robot-has-a-killer-backhand)

* Today’s AI tool guide: [Get a morning newspaper brief](https://app.therundown.ai/guides/get-a-personal-newspaper-written-by-claude-every-morning) [with Claude](https://app.therundown.ai/guides/get-a-personal-newspaper-written-by-claude-every-morning)

* RSVP to workshop April 30 @ 2PM EST: [Codex for non-technical operators](https://app.therundown.ai/calendar-events/69e77469e7e908f6d4ebdf72)

----------

----------
See you soon,

_Rowan, Joey, Zach, Shubham, and Jennifer—the humans behind The Rundown_

View image: (https://media.beehiiv.com/cdn-cgi/image/fit=scale-down,format=auto,onerror=redirect,quality=80/uploads/asset/file/b105d547-252f-4914-8821-3eb7236b1257/TheRundown_EditorialTeam.png?t=1755849086)
Caption:

----------

———

You are reading a plain text version of this post. For the best experience, copy and paste this link in your browser to view the post online:
https://www.therundown.ai/p/openai-spud-dethrones-claude-on-the-frontier

**LATEST DEVELOPMENTS**

###### OPENAI

#### 🥔 **OpenAI retakes the frontier with GPT 5.5**

**The Rundown:** OpenAI just [launched](https://openai.com/index/introducing-gpt-5-5/) GPT-5.5 (codenamed 'Spud'), the company's long-awaited upgrade, pitched as a 'new class of intelligence' — topping benchmark scores across the industry and overtaking Anthropic on the AI model frontier.

**The details:**

* 5.5 sets highs across a series of reasoning, agentic, computer use, and coding tests for public models, with several scores comparable to Claude Mythos.

* The model keeps the same speed as 5.4 with added efficiency, with OAI saying it used Codex and 5.5 to rewrite its own GPU code to improve infrastructure.

* GPT-5.5 lands at $5/$30 per million input/output tokens for API pricing, with OAI pitching it as 'half the cost of competitive frontier coding models."

* 5.5 is rolling out across ChatGPT plans and in Codex with Thinking and Pro variants, with OAI continuing to highlight 'generous usage' for its new releases.

**Why it matters:** After months of Anthropic dominance, the vibe is shifting once again — with OpenAI rapidly shipping powerful new upgrades and rekindling the magic that felt a bit lost on previous releases. With Anthropic now wading through rate limit and quality degradation complaints, it's a big week for Sama and co. on the sentiment front.

----------

###### AI & GEOPOLITICS

#### ⚖️ U.S. flags Chinese labs' 'industrial-scale' AI theft

**The Rundown:** The White House [published](https://www.whitehouse.gov/wp-content/uploads/2026/04/NSTM-4.pdf) a memo accusing Chinese firms of 'industrial-scale' distillation campaigns against U.S.-based frontier AI labs — coming weeks before Trump's scheduled Beijing summit with Xi Jinping.

**The details:**

* Distillation is training smaller AI systems on frontier model outputs, with Kratsios saying China runs it via thousands of fake API accounts and jailbreaks.

* Anthropic accused DeepSeek, Moonshot, and MiniMax of distillation in February, with this memo upgrading the private complaint to federal policy.

* The Chinese embassy dismissed the accusations and called them 'pure slander', coming ahead of Trump and XI's meeting in Beijing on May 14-15.

* A House Foreign Affairs bill that cleared its first vote this week would push the administration to add distillation offenders to the U.S. export blacklist.

**Why it matters:** Dario Amodei recently framed open-source and China as '6-12 months behind' frontier labs. The Kratsios memo reframes that gap, arguing the gains come from scraping tactics, not architecture work. Whether that holds depends on how much of the DeepSeek/Kimi trajectory truly traces to distillation vs. new research.

----------

###### AI TRAINING

#### 📰 Get a newspaper brief every Morning with Claude

**The Rundown:** In this guide, you will learn how to turn your daily updates into a personalized newspaper with Claude. It's useful because it gives you a one-ranked brief with the stories, action items, and calendar prep that matter.

**Step-by-step:**

1. In Claude or Claude Cowork, make sure Slack, Notion, Gmail, and Calendar are connected

2. Prompt: "Create a static Morning Edition from my Slack, Notion, Gmail, and calendar updates from the last 24 hours. Rank what matters most and format it like a newspaper with top stories, action items, and schedule prep"

3. Review the first draft and give feedback. Tighten the order, layout, and emphasis. Then tell Claude to turn the workflow into a skill

4. In Claude Cowork, create a recurring task that runs that skill each morning

**Pro tip:** Have other agents gather outside news and drop them into a Notion database. Then use this skill as the editor to turn everything into a daily newspaper.

----------

###### AI RESEARCH

####  📊 AI's biggest productivity winners are also most worried

**The Rundown:** Anthropic [published](https://cdn.sanity.io/files/4zrzovbb/website/3a8d990bc90098038eabd77b0d12ff636ed58d50.pdf) the economic-focused follow-up to its 81K-user Claude survey, finding that the people getting the biggest productivity lift from AI are also the most worried about losing their jobs to it, especially early-career workers.

**The details:**

* The survey ties Anthropic's Economic Index usage data (which jobs lean on Claude most) to 80,508 workers' takes on how AI is reshaping their roles.

* Workers whose jobs use Claude most voiced AI displacement fears, 3x more than those whose jobs use it least, with engineers leading the anxiety.

* Most respondents said AI's gains land on themselves via faster tasks and free time, but also lead to expanded scope and more work.

* Early-career respondents voiced the loudest displacement fears, backing Anthropic's earlier signal of a hiring slowdown for recent grads in the U.S.

**Why it matters:** The conventional view is that AI panic would come from lower-level adopters, but these results flip that, with anxiety coming from those getting the most out of the tools. Despite the productivity boosts, AI's sentiment has never been lower — and there doesn't seem to be many solutions in sight for easing the tensions.

----------

**QUICK HITS**

### 🛠️ Trending AI Tools

* 🥔 [GPT 5.5](https://www.rundown.ai/tools/gpt-5-5) - OpenAI's new top-rated flagship AI model

* ⚙️ [Ultrareview](https://code.claude.com/docs/en/ultrareview) - Claude Code command to run multi-agent code reviews

* 🏥 [ChatGPT for Clinicians](https://www.rundown.ai/tools/chatgpt-for-clinicians) - Free clinical version for verified U.S. doctors

* ⚙️ [Qwen3.6-27B](http://www.rundown.ai/tools/qwen3-6-27b) - Alibaba's new 27B model that out-codes its predecessor

### 📰 Everything else in AI today

**Anthropic** [published](https://www.anthropic.com/engineering/april-23-postmortem) a post-mortem tracing Claude Code quality complaints to three separate bugs, resetting usage limits for subscribers due to the issues.

**OpenAI** [introduced](https://openai.com/index/making-chatgpt-better-for-clinicians/) ChatGPT for Clinicians, a free tool for verified U.S. health workers, with GPT-5.4 scoring 59.0 on HealthBench Pro, topping physicians and Opus 4.7.

**Meta** [sent](https://www.businessinsider.com/meta-plans-layoff-10-of-its-entire-staff-may-2026) an internal memo to employees informing them that the company is laying off 10% of its workforce in May, citing AI efficiency and other investments.

**Elon Musk's SpaceX** is reportedly [in talks](https://www.businessinsider.com/elon-musk-xai-explored-collaborating-with-mistral-cursor-2026-4) with French AI startup Mistral on a three-way partnership alongside its recent deal with coding startup Cursor.

**Tencent** [open-sourced](https://hy.tencent.com/hy3-preview) Hy3 preview, its first model from a rebuilt training stack with competitive agentic coding and search-agent scores to top open models.

----------

**COMMUNITY**

### 🤝 Community AI workflows

Every newsletter, we showcase how a reader is using AI to work smarter, save time, or make life easier.

Today's workflow comes from reader **Mark M.** in Virginia:

"I used AI to turn a tax-season headache into a streamlined system. Every year, the same scene plays out: I drop off bags of donations, stash the receipt, and then scramble at tax time trying to decipher my own list and determine each item's worth.

This year, I photographed my handwritten donation lists and uploaded them to Claude AI. Claude transcribed my lists, asked clarifying questions, verified charities with the IRS, grouped items into logical categories, applied condition-appropriate FMV, and produced professional PDF summaries — all conversationally, in minutes.

Multiple donations and nearly 400 items in fully documented and defensible deductions. For anyone who donates, this is the workflow you didn't know you needed."

Tu es l'extracteur de contenu de Breviat. On te fournit le contenu Markdown nettoyé d'une newsletter.

Ta mission : produire une version PROPRE du contenu en supprimant tout ce qui n'est pas de l'information utile au lecteur. Tu es un FILTRE, pas un résumeur.

À RETIRER :
- Publicités, encarts sponsors, mentions "sponsorisé par X", "ad", "présenté par"
- Intros vides : formules de bienvenue, météo de l'humeur de l'auteur, anecdotes personnelles non liées au contenu
- Appels à l'action marketing : s'abonner à la newsletter, parrainer un ami, "follow us on Twitter", "join our Discord"
- Signatures, mentions légales, adresses postales, "view in browser", "unsubscribe"
- Boutons / CTAs / "cliquez ici" / "lire la suite" sans contenu derrière
- Promotions d'autres produits / événements / formations payantes de l'auteur ou de tiers
- Encarts récurrents type "Read of the day" ou "Quote of the day" sans valeur informationnelle propre

À CONSERVER (intégralement, sans résumer ni reformuler) :
- Toutes les annonces, news, analyses, commentaires factuels
- Les chiffres, dates, noms d'entreprises, citations
- Les explications techniques
- Les liens vers des sources réelles (annonces officielles, papers, articles cités)
- La structure (titres, sous-titres, listes)

RÈGLES :
- Ne reformule pas. Garde la formulation d'origine.
- Ne résume pas, ne condense pas. Si une section fait 200 mots et est utile, garde 200 mots.
- N'ajoute aucun contenu (pas de titres ni de transitions de ton cru).
- Ne fabrique aucune URL. Garde celles d'origine, ou retire-les.
- Si la newsletter entière est de la pub / promo / contenu inutile, sors un Markdown vide (rien d'autre).

Sortie : UNIQUEMENT le Markdown nettoyé, sans préambule ni commentaire sur ton travail.