# This Week in All Things AI - Week 21-2026

*Sunday 17th May 2026 to Saturday 23rd May 2026*

By [This Week in All Things AI](https://paragraph.com/@twiata) · 2026-05-24

---

This Week in All Things AI covers key developments in models, agents, tools, infrastructure, and policy curated from discussions in the [All Things AI Telegram group.](https://t.me/+PnCnwhgH8V4yMTFl)  
  
If you follow AI for work, research, investing, or just to understand where the technology is heading, this [weekly brief](https://paragraph.com/@twiata) is a concise way to scan the most important launches, risks, and resources in a few focused minutes.

Across the week, we saw advances in agentic coding and web tooling (Cursor’s Composer 2.5, Google’s Gemini 3.5 Flash, Modern Web Guidance), major moves in AI infrastructure and pricing (Anthropic’s expanded Colossus 2 deal with SpaceX, DeepSeek’s permanent discount on V4‑Pro), a wave of creative tools (Krea 2 and several AI‑assisted video editors), and notable policy and geo developments from Pope Leo XIV’s forthcoming AI encyclical to Singapore’s latest AI initiatives.  
  
The sections that follow walk through these items day by day, with short context and links so you can dive deeper into the pieces most relevant to your work or interests.

* * *

Sunday 17th May 2026
--------------------

Monday 18th May 2026
--------------------

Tuesday 19th May 2026
---------------------

Cursor’s biggest model upgrade yet. Strong gains in long-running agentic coding tasks, reliability with complex instructions, and up to **10x better cost efficiency**. It’s built on Moonshot’s Kimi K2.5 and reportedly performs on par with top frontier models in some real-world evaluations. Cursor is also training a much larger model from scratch with SpaceXAI on a massive Colossus 2 cluster.

[![User Avatar](https://storage.googleapis.com/papyrus_images/9036a3c92d415f41257c8e035d81b161c8c2bb45207ddeeba39395e5059e8b2c.jpg)](https://twitter.com/cursor_ai)

[Cursor](https://twitter.com/cursor_ai)

[@cursor\_ai](https://twitter.com/cursor_ai)

[](https://twitter.com/cursor_ai/status/2056415413077233983)

Introducing Composer 2.5, our most powerful model yet.  
  
It's more intelligent, better at sustained work on long-running tasks, and more reliable at following complex instructions.  
  
For the next week, we’re doubling the included usage of the model.

![](https://storage.googleapis.com/papyrus_images/da611f2fa30be754e8970adeb78f0de9151291f54ed673afb1ea64548fad383d.png)

[13.2K](https://twitter.com/cursor_ai/status/2056415413077233983)[

4:43 PM • May 18, 2026

](https://twitter.com/cursor_ai/status/2056415413077233983)

[

Introducing Composer 2.5 · Cursor
---------------------------------

A substantial improvement in intelligence and behavior over Composer 2, particularly on long-horizon agentic tasks.

https://cursor.com

![Introducing Composer 2.5 · Cursor](https://storage.googleapis.com/papyrus_images/faef26b987f84e7ad05bbc83f741803f7837f8c888c91bc296a95f34d5e31762.png)

](https://cursor.com/blog/composer-2-5)

The Vatican announced that Pope Leo XIV’s first major encyclical on AI will be released on May 25. It will focus on human dignity and “the protection of the human person in the age of AI.” It is being prepared in collaboration with an Anthropic co-founder.

[

Pope Leo to issue text on human dignity and AI with Anthropic co-founder
------------------------------------------------------------------------

The pope's encyclical will address 'the protection of the human person in the age of AI', the Vatican says

https://www.theguardian.com

![Pope Leo to issue text on human dignity and AI with Anthropic co-founder](https://storage.googleapis.com/papyrus_images/f56b0ee43049141dfd3eaa7354d8a5fbbaa099e0624b2c99465f5cf5bdf3e05a.jpg)

](https://www.theguardian.com/world/2026/may/18/pope-leo-encyclical-human-dignity-ai-anthropic)

[

Pope Leo and co-founder of Anthropic to launch pontiff's AI encyclical on May 25
--------------------------------------------------------------------------------

Pope Leo XIV and the co-founder of artificial intelligence company Anthropic will launch the pontiff's first encyclical on May 25, a document on the care of human dignity in the era of AI, the Vatican said Monday.

https://www.chicagotribune.com

![Pope Leo and co-founder of Anthropic to launch pontiff's AI encyclical on May 25](https://storage.googleapis.com/papyrus_images/630bb0adab52897c953ff3444a2d9d3d0bc3de06ee1d3145cc60400f2ec2f8f3.jpg)

](https://www.chicagotribune.com/2026/05/18/pope-leo-anthropic-ai-encyclical/)

Cloudflare tests Mythos against 50+ repositories, highlights its ability to chain bugs into a single exploit, and details a vulnerability discovery harness

[

Project Glasswing: what Mythos showed us
----------------------------------------

In recent weeks, we pointed Mythos and other security-focused LLMs at live code across critical parts of our infrastructure. We share what we observed, the models' strengths and weaknesses, and what the work around them needs to look like before any of it can scale.

https://blog.cloudflare.com

![Project Glasswing: what Mythos showed us](https://storage.googleapis.com/papyrus_images/30d5c7e51a83c4ca6b22eae930030745a202e7b2b908672d244bd9696fc1fffc.png)

](https://blog.cloudflare.com/cyber-frontier-models/)

[Krea](https://www.krea.ai/refer/EJQQP9FJ) 2 (also called K2 or Krea 2 Large) is [Krea](https://www.krea.ai/refer/EJQQP9FJ) AI’s first proprietary foundation image model

It was built entirely from scratch  with a deliberate focus on aesthetics, style transfer, creative control, and artistic exploration rather than literal prompt transcription or maximum photorealism

It forms part of the broader [Krea.ai](https://www.krea.ai/refer/EJQQP9FJ) platform—an all-in-one creative AI suite that aggregates 100+ models  for images, video, 3D, editing, upscaling, and fine-tuning under one subscription

They are offering unlimited [Krea 2](https://www.krea.ai/refer/EJQQP9FJ) generations to all their subscribers for the entire week.

[![User Avatar](https://storage.googleapis.com/papyrus_images/18f53ae6908d6ab83680deda44b1d8c769a37aa8c3c99afe4414141870f76cec.jpg)](https://twitter.com/krea_ai)

[Krea](https://twitter.com/krea_ai)

[@krea\_ai](https://twitter.com/krea_ai)

[](https://twitter.com/krea_ai/status/2054207481421975829)

this is Krea 2.  
  
our first foundation model, built completely from scratch for aesthetic diversity and stylistic control.  
  
learn more and get early access ![👇](https://abs-0.twimg.com/emoji/v2/72x72/1f447.png)

![](https://pbs.twimg.com/amplify_video_thumb/2054177498464948224/img/6Cj1OGC6dg1a8TM5.jpg)

[2,148](https://twitter.com/krea_ai/status/2054207481421975829)[

2:30 PM • May 12, 2026

](https://twitter.com/krea_ai/status/2054207481421975829)

[![User Avatar](https://storage.googleapis.com/papyrus_images/43c9317cc3219b064944eb4ee86f8d6dd26d624d533cc2510faf56c7b2b15e0e.jpg)](https://twitter.com/HBCoop_)

[Heather Cooper](https://twitter.com/HBCoop_)

[@HBCoop\_](https://twitter.com/HBCoop_)

[](https://twitter.com/HBCoop_/status/2054213883590553836)

Had the pleasure of testing this out over the past week.  
  
Very, very good model. Quality aesthetics, fast generations along with Moodboards and Style transfer.  
  
Excited to explore more with Krea 2 ↓

![](https://storage.googleapis.com/papyrus_images/5c5cd160bf450af07853c3181121ae4d3a3eed6249ae4afc67382c0b35078815.jpg)

[![User Avatar](https://storage.googleapis.com/papyrus_images/18f53ae6908d6ab83680deda44b1d8c769a37aa8c3c99afe4414141870f76cec.jpg)](https://twitter.com/krea_ai)

[Krea](https://twitter.com/krea_ai)

[@krea\_ai](https://twitter.com/krea_ai)

[](https://twitter.com/krea_ai/status/2054207481421975829)

this is Krea 2.  
  
our first foundation model, built completely from scratch for aesthetic diversity and stylistic control.  
  
learn more and get early access ![👇](https://abs-0.twimg.com/emoji/v2/72x72/1f447.png)

![](https://pbs.twimg.com/amplify_video_thumb/2054177498464948224/img/6Cj1OGC6dg1a8TM5.jpg)

[35](https://twitter.com/HBCoop_/status/2054213883590553836)[

2:55 PM • May 12, 2026

](https://twitter.com/HBCoop_/status/2054213883590553836)

[

Krea 2: AI Image Foundation Model & Style Control
-------------------------------------------------

Create expressive AI images with Krea 2, Krea AI's in-house foundation model for aesthetic diversity, style control, moodboards, and creative workflows.

https://www.krea.ai

![Krea 2: AI Image Foundation Model & Style Control](https://storage.googleapis.com/papyrus_images/3178c87d77b786fdb12b55b1fa04ab201012aca9b0d0594281640ebd1b1c1f1d.webp)

](https://www.krea.ai/krea-2)

Wednesday 20th May 2026
-----------------------

via [Robby Yung](https://t.me/robyung)

Gemini Spark is Google's new cloud-based 24/7 personal AI agent announced at I/O 2026, designed to autonomously handle tasks in the background even when devices are off while requiring user approval for major actions.

Initial rollout targets trusted testers this week and U.S. Google AI Ultra subscribers next week in beta, with enterprise availability planned through Google Workspace and Gemini Enterprise.

Summer roadmap includes MCP integrations, macOS support, email/text interaction, Chrome browser control, and subagent creation to expand proactive capabilities.

[![User Avatar](https://storage.googleapis.com/papyrus_images/6ed01cb5f40ffd1310bef8f822fd5d46338638adf5d3c06c1d1908d2b6149b27.png)](https://twitter.com/GeminiApp)

[Google Gemini](https://twitter.com/GeminiApp)

[@GeminiApp](https://twitter.com/GeminiApp)

[](https://twitter.com/GeminiApp/status/2056792333132460322)

Gemini Spark is your new 24/7 personal AI agent.  
  
Give it a task and it works autonomously in the background, even if your phone and laptop are turned off. You choose to turn it on and it's designed to check with you before taking major actions. [#GoogleIO](https://twitter.com/hashtag/GoogleIO)

[2,633](https://twitter.com/GeminiApp/status/2056792333132460322)[

5:41 PM • May 19, 2026

](https://twitter.com/GeminiApp/status/2056792333132460322)

Google Unveils Gemini 3.5 Flash at I/O Conference

Google launched Gemini 3.5 Flash, generally available now and touted as its strongest model for agentic workflows like multi-step tool use and terminal coding. Benchmarks show it topping prior versions on key tests, holding ground against rivals like Claude 3.5 Sonnet, though independent reviews note mixed results on cost and performance. 

The event also introduced Gemini Omni for video creation, Spark as a 24/7 personal agent, developer tools like Antigravity 2.0, plus consumer updates to Search and the Gemini app—ushering in what execs call the agentic era.

[![User Avatar](https://storage.googleapis.com/papyrus_images/eade4fc97d80786e97854a2783e9d17b73944abe73f9d2a531a92b12f4d9e2d9.jpg)](https://twitter.com/JeffDean)

[Jeff Dean](https://twitter.com/JeffDean)

[@JeffDean](https://twitter.com/JeffDean)

[](https://twitter.com/JeffDean/status/2056793419033588091)

1/ Today at [#GoogleIO](https://twitter.com/hashtag/GoogleIO), we’re releasing Gemini 3.5, our latest family of models combining frontier intelligence with action.  
  
We’re starting by releasing 3.5 Flash, which is built to help you execute complex, long-horizon agentic workflows.  
  
Gemini 3.5 Flash is our strongest model

![](https://storage.googleapis.com/papyrus_images/a6b43a92391410472178965cc12f7078bd4f5fb8506d6f7c5f9b8cbac4172088.jpg)

[1,484](https://twitter.com/JeffDean/status/2056793419033588091)[

5:45 PM • May 19, 2026

](https://twitter.com/JeffDean/status/2056793419033588091)

[![User Avatar](https://storage.googleapis.com/papyrus_images/ed64843b03ee9810a155441de9f720c6bafa9cb1c25583de38975180d59aec37.jpg)](https://twitter.com/GoogleAIStudio)

[Google AI Studio](https://twitter.com/GoogleAIStudio)

[@GoogleAIStudio](https://twitter.com/GoogleAIStudio)

[](https://twitter.com/GoogleAIStudio/status/2056797490553540756)

[x.com/i/article/2056…](https://t.co/TDJrxQJM37)

[1,301](https://twitter.com/GoogleAIStudio/status/2056797490553540756)[

6:01 PM • May 19, 2026

](https://twitter.com/GoogleAIStudio/status/2056797490553540756)

[

Gemini 3.5: frontier intelligence with action
---------------------------------------------

At Google I/O we released Gemini 3.5, our latest series of models combining frontier intelligence with action.

https://blog.google

![Gemini 3.5: frontier intelligence with action](https://storage.googleapis.com/papyrus_images/c29a6691936d35ebdb2ac6c6e7c7a1a8e2cee6d983158bc12e08b4ceac0be27d.png)

](https://blog.google/innovation-and-ai/models-and-research/gemini-models/gemini-3-5/)

[

Gemini 3.5 Flash might be fast enough for gen AI to make sense
--------------------------------------------------------------

Google says its more efficient Gemini 3.5 Flash is the key to your agentic AI future.

https://arstechnica.com

![Gemini 3.5 Flash might be fast enough for gen AI to make sense](https://storage.googleapis.com/papyrus_images/02edaea580357e415f4245086c6d20852c5a0ce236598f734981a757ea982168.jpg)

](https://arstechnica.com/google/2026/05/google-announces-agent-optimized-gemini-3-5-flash-and-a-do-anything-model-called-omni/)

Google Launches Modern Web Guidance for AI Coding Agents

At Google I/O 2026, the company unveiled Modern Web Guidance, an early preview toolset that equips AI agents like Claude Code and GitHub Copilot with skills for the latest browser APIs, layouts, and performance techniques. It tackles legacy patterns by injecting concise prompts drawn from Chrome's Baseline initiative, boosting agent success on modern tasks from 52% to 86% in early tests. Backed by Chrome and Microsoft Edge teams, plus experts like Lea Verou, it offers easy installation via a simple npx command and covers real-world uses from CSS animations to WebAuthn logins.

[![User Avatar](https://storage.googleapis.com/papyrus_images/71232ace28aa738afad62b342acdf2432ee5ee4f1f64d644af02ed6694953df4.jpg)](https://twitter.com/sarah_edo)

[Sarah Drasner](https://twitter.com/sarah_edo)

[@sarah\_edo](https://twitter.com/sarah_edo)

[](https://twitter.com/sarah_edo/status/2056846557924712664)

![💥](https://abs-0.twimg.com/emoji/v2/72x72/1f4a5.png) Game changer for Web Development announced at GoogleIO- Modern Web Guidance! It’s expert-vetted skills for web development based on best practices of latest specs and APIs.  
  
It ensures your agent/coding harness doesn’t default to older and out of date patterns to build sites.

![](https://pbs.twimg.com/ext_tw_video_thumb/2056846526433923072/pu/img/O7k8Og-GSFrpWze6.jpg)

[751](https://twitter.com/sarah_edo/status/2056846557924712664)[

9:16 PM • May 19, 2026

](https://twitter.com/sarah_edo/status/2056846557924712664)

[

Modern Web Guidance | Chrome for Developers
-------------------------------------------

Guidance on how to build for the modern web.

https://developer.chrome.com

![Modern Web Guidance | Chrome for Developers](https://storage.googleapis.com/papyrus_images/7689530dd12af274be0e6701f2f549667f925d73d69c893cadac352101b98df0.png)

](https://developer.chrome.com/docs/modern-web-guidance)

Q&A with Harvey CEO Winston Weinberg on launching the legal AI startup in 2022, how AI could shake up law firms' business models, legal AI competition, and more

[https://www.ft.com/content/47eafdcb-5a0a-4515-a07a-70e89ad2fa78?syn-25a6b1a6=1](https://www.ft.com/content/47eafdcb-5a0a-4515-a07a-70e89ad2fa78?syn-25a6b1a6=1)

Thursday 21st May 2026
----------------------

Anthropic is paying SpaceX $1.25B/mo. until May 2029 under their compute deal; Anthropic says it's expanding the deal to include Colossus 2 capacity

[![User Avatar](https://storage.googleapis.com/papyrus_images/4638dc0961533e43bad6554f462bbb5fcd4f12a30cc6622f5befa99810f0607c.jpg)](https://twitter.com/nottombrown)

[Tom Brown](https://twitter.com/nottombrown)

[@nottombrown](https://twitter.com/nottombrown)

[](https://twitter.com/nottombrown/status/2057194829986300375)

We’re expanding our partnership with [@SpaceX](https://twitter.com/SpaceX), and will be scaling up on GB200 capacity in Colossus 2 throughout June.  
  
Appreciate [@elonmusk](https://twitter.com/elonmusk) and the team helping us find good homes for the Claudes.

[![User Avatar](https://storage.googleapis.com/papyrus_images/4638dc0961533e43bad6554f462bbb5fcd4f12a30cc6622f5befa99810f0607c.jpg)](https://twitter.com/nottombrown)

[Tom Brown](https://twitter.com/nottombrown)

[@nottombrown](https://twitter.com/nottombrown)

[](https://twitter.com/nottombrown/status/2052062566126649448)

In the next few days we'll be ramping up Claude inference on Colossus.  
  
Grateful to be partnering with SpaceX here. We are going to need to move a lot of atoms in order to keep up with AI demand, and there's nobody better at quickly moving atoms (on or off planet Earth)

[7,469](https://twitter.com/nottombrown/status/2057194829986300375)[

8:20 PM • May 20, 2026

](https://twitter.com/nottombrown/status/2057194829986300375)

[![User Avatar](https://storage.googleapis.com/papyrus_images/5848bccca58719b811c7fb02b5f45961d7b1ebf8c07f3b15d4288a08097885c5.jpg)](https://twitter.com/elonmusk)

[Elon Musk](https://twitter.com/elonmusk)

[@elonmusk](https://twitter.com/elonmusk)

[](https://twitter.com/elonmusk/status/2057228707606196434)

As the recently expanded partnership with [@AnthropicAI](https://twitter.com/AnthropicAI) demonstrates, [@SpaceX](https://twitter.com/SpaceX) is offering AI compute as a service at significant scale.  
  
We are in discussions with other companies to do the same.  
  
Over time, especially with orbital data centers, we expect to serve AI at extremely

[72.7K](https://twitter.com/elonmusk/status/2057228707606196434)[

10:35 PM • May 20, 2026

](https://twitter.com/elonmusk/status/2057228707606196434)

FYI,  in case some in this group find the below useful for their personal and professional workflows.  Personally I consider video-editing as something akin to 'magic of the dark arts' so huge respect to practioners of this craft. 

The below are a links to websites of AI assisted video editing tools I've come across (in no particular order),  pretty sure these are but a small fraction of tools out there.  

Within this list I'm not mentioning the usual suspects such as Premiere,  Final Cut Pro, Davinci Resolv, Capcut

[

Create videos for Socials - keyframes.studio
--------------------------------------------

Create videos for Socials in 5 minutes. Don't waste time looking up aspect ratios or learning heavy video editors - Keyframes Studio takes care of everything.

https://keyframes.studio

![Create videos for Socials - keyframes.studio](https://storage.googleapis.com/papyrus_images/4a1ebeed9031ad42393e650cb2e6264285a900802aecaad777d3628a512ec012.png)

](https://keyframes.studio/)

[

Revid AI - Ideate, Publish, Go Viral | #1 AI Video Generator
------------------------------------------------------------

Create viral short videos for TikTok, Instagram, and YouTube. Turn ideas into content instantly. Ideate, create, edit, publish - all in one. No skills needed.

https://www.revid.ai

![Revid AI - Ideate, Publish, Go Viral | #1 AI Video Generator](https://storage.googleapis.com/papyrus_images/29099dec75399b249233d417d60addc6828774a6faf92fa252508fc5da450af2.png)

](https://www.revid.ai/)

[

Topaz Labs | Professional-grade photo and video editing powered by AI.
----------------------------------------------------------------------

Photo and video enhancement software powered by deep learning gets you the best image quality available for noise reduction, sharpening, upscaling, and more.

https://www.topazlabs.com

![Topaz Labs | Professional-grade photo and video editing powered by AI.](https://storage.googleapis.com/papyrus_images/32c8cb05aea0554e8606af6318272cdc64eee870e0cf6e322f7f5d486e400ff7.jpg)

](https://www.topazlabs.com/)

[

Capsule
-------

Meet Capsule-the AI-powered video editor made for content and marketing teams. Create stunning videos 10x faster, collaborate seamlessly, and stay on brand.

https://capsule.video

![Capsule](https://storage.googleapis.com/papyrus_images/dc3861409cee29bebb514943eaa4a10b1f98ca086bec958e518476570ee11bdf.png)

](https://capsule.video/)

[

Descript - AI Video & Podcast Editor | Free, Online
---------------------------------------------------

Descript makes editing video and audio as easy as editing text. Record, transcribe, edit, and publish in one tool. Try for free, with powerful upgrades for creators & teams.

https://www.descript.com

![Descript - AI Video & Podcast Editor | Free, Online](https://storage.googleapis.com/papyrus_images/bb178aa31c47150679cdaedf8ec6cd003d18a961fe5375a0aa989479a16b9341.jpg)

](https://www.descript.com/)

[

Overlap - AI Video Clipping for Media, Podcasts and Creators
------------------------------------------------------------

Use Overlap to create AI clipping workflows to turn long videos into clips, shorts, and social posts automatically. Scale short form content creation with Agentic workflows.

https://overlap.ai

![Overlap - AI Video Clipping for Media, Podcasts and Creators](https://storage.googleapis.com/papyrus_images/3fb7bd73f918a5d7e57bea85fdaf676e4180bcc72d667486340c8e8f8958144d.png)

](https://www.joinoverlap.com/)

[

Cardboard | Agentic Video Editor
--------------------------------

A fast, collaborative, and agentic video editor that runs in your browser. Create, cut, and compose videos with AI - no installs, no downloads.

https://www.usecardboard.com

![Cardboard | Agentic Video Editor](https://storage.googleapis.com/papyrus_images/e7e17c6b749a474c9315b33fa1713d95fd740567ceebb6227c7036dd567b9dbf.png)

](https://www.usecardboard.com/)

[

Eddie AI | The Assistant Video Editor for Pros
----------------------------------------------

Eddie is the AI assistant video editor trusted by 40,000+ video professionals. Import interviews and chat with Eddie to get rough cuts and edits in seconds. Then export to MP4, Adobe, FCP, and DaVinci Resolve.

https://www.heyeddie.ai

![Eddie AI | The Assistant Video Editor for Pros](https://storage.googleapis.com/papyrus_images/d3006cea9f5f21bab3750f6ee67d1ffda1590241267ac2d50b71302b252a3d91.png)

](https://www.heyeddie.ai/)

[

Ponder
------

Transform your video editing workflow with Ponder.AI's revolutionary AI-powered platform. Sign up free or download now to experience the future of video creation.

https://ponder.ai

![Ponder](https://storage.googleapis.com/papyrus_images/f089c74090caf1961d370c7ad3e282385dcd564173176c867b728ff6b4577535.png)

](https://ponder.ai/)

[

OpusClip: #1 AI video clipping tool to create viral shorts
----------------------------------------------------------

OpusClip turns long videos into high-quality viral clips, and publishes them to all social platforms in one click. We help 10M+ creators create and grow faster.

https://www.opus.pro

![OpusClip: #1 AI video clipping tool to create viral shorts](https://storage.googleapis.com/papyrus_images/70845a9a487e47fb6695bbb079a6fe9e07124285c60282eeddefc147e9c7d16c.jpg)

](https://www.opus.pro/)

If any of these do end up being used by you/your team,  do share your experiences with this group so that others can also benefit from your learnings

via [Alex](https://t.me/zk_alex)

SpaceX S1 fillings is a good read for SciFi fans like me. Emphasizing Fiction here 😂  
 Claiming capabilities to launch 100GW of compute to LEO when hyperscalers are barely able to do 1GW on earth. Claiming 220B daily token usage when the actual xAI usage numbers seem to be more like 15B tokens daily. Claiming the TAM for xAI is over 20T$.

[https://www.sec.gov/Archives/edgar/data/1181412/000162828026036936/spaceexplorationtechnologi.htm](https://www.sec.gov/Archives/edgar/data/1181412/000162828026036936/spaceexplorationtechnologi.htm)

via [Alba Chung of Alibaba Cloud](https://www.linkedin.com/in/albachung )

**Qwen3.7-Max** — now available viaAlibaba Cloud Model Studio:

\- frontier coding agent: from frontend prototyping to complex software engineering

\- office productivity and workflow automation via MCP and multi-agent orchestration

\- sustained autonomous execution across long-horizon tasks

\-cross-scaffold generalization across diverse agent frameworks

My LinkedIn: https://www.linkedin.com/in/albachung 

Feel free to PM me for more details.🙌🏼

Friday 22nd May 2026
--------------------

Singapore firing on all cylinders

• OpenAI first AI lab outside of the US is Singapore 

• Anthropic’s lead investor for their Series G is GIC (Singapore’s sovereign wealth fund)

• NVIDIA chose Singapore for their research hub

• Google Deepmind is strengthening its partnership with Singapore

[

Introducing OpenAI for Singapore
--------------------------------

OpenAI for Singapore launches a multi-year AI partnership to expand deployment, build local talent, and support businesses and public services with AI.

https://openai.com

![Introducing OpenAI for Singapore](https://storage.googleapis.com/papyrus_images/6d021df1ad100282fbe86f0e7f8e609159a79b1fff687d2edaf27789896e424e.png)

](https://openai.com/index/introducing-openai-for-singapore/)

[https://www.cnbc.com/2026/05/20/nvidia-to-launch-singapore-research-hub-as-city-state-boosts-ai-plans.html](https://www.cnbc.com/2026/05/20/nvidia-to-launch-singapore-research-hub-as-city-state-boosts-ai-plans.html)

[

Google DeepMind & Singapore: National AI partnership
----------------------------------------------------

Google DeepMind and Singapore partner to apply frontier AI to address challenges across health, education, sustainability and more through the National Partnerships for AI initiative.

https://deepmind.google

![Google DeepMind & Singapore: National AI partnership](https://storage.googleapis.com/papyrus_images/8169bf1732e5a4ee8af07b44413e2d8b04ca65ac6a61e329480f18724f5245e3.webp)

](https://deepmind.google/blog/strengthening-singapores-ai-future-a-new-national-partnership/)

Saturday 23rd May 2026
----------------------

Zero to Claude Code Course Draws 18,000 Beginners Worldwide

The Zero to Claude Code course guides users from terminal basics to advanced AI techniques like prompt engineering and subagents, with hands-on projects including a real-time game capstone. Launched recently at [zero2claude.dev](https://zero2claude.dev/) it has enrolled over 18,000 students globally, running on a high-performance site that Itay Shmool, VP @ WIX  created solo with Claude Code

[![User Avatar](https://storage.googleapis.com/papyrus_images/ae3f8755b3cc095a64fddae67a361538c938f969356daa6b3d765957a3edaa59.jpg)](https://twitter.com/IShmool)

[itay shmool](https://twitter.com/IShmool)

[@IShmool](https://twitter.com/IShmool)

[](https://twitter.com/IShmool/status/2057872918348976191)

I built "zero2claude", a free course that takes people from zero terminal experience to shipping with Claude Code.  
  
The curriculum goes from absolute zero → software basics → Claude Code fundamentals → advanced usage. No shortcuts, no assumptions.  
  
17,000+ students. 7

![](https://storage.googleapis.com/papyrus_images/f25fd53066b6d86145ed72858dba8e2ba1942ae2ce052be84580ebd4708d1f48.jpg)

[617](https://twitter.com/IShmool/status/2057872918348976191)[

5:15 PM • May 22, 2026

](https://twitter.com/IShmool/status/2057872918348976191)

[![User Avatar](https://storage.googleapis.com/papyrus_images/d82a61804cf26d786a900ebe6d5022ebc306b47253bb1f62ca4f211ae1b80a14.jpg)](https://twitter.com/bcherny)

[Boris Cherny](https://twitter.com/bcherny)

[@bcherny](https://twitter.com/bcherny)

[](https://twitter.com/bcherny/status/2057876668501434603)

Big fan of teaching more people the basics of using Claude Code in an accessible way.  
  
So much of the world has not yet used agents. There's a lot of opportunity to level the playing field and expand access.

[![User Avatar](https://storage.googleapis.com/papyrus_images/ae3f8755b3cc095a64fddae67a361538c938f969356daa6b3d765957a3edaa59.jpg)](https://twitter.com/IShmool)

[itay shmool](https://twitter.com/IShmool)

[@IShmool](https://twitter.com/IShmool)

[](https://twitter.com/IShmool/status/2057872918348976191)

I built "zero2claude", a free course that takes people from zero terminal experience to shipping with Claude Code.  
  
The curriculum goes from absolute zero → software basics → Claude Code fundamentals → advanced usage. No shortcuts, no assumptions.  
  
17,000+ students. 7

![](https://storage.googleapis.com/papyrus_images/f25fd53066b6d86145ed72858dba8e2ba1942ae2ce052be84580ebd4708d1f48.jpg)

[1,683](https://twitter.com/bcherny/status/2057876668501434603)[

5:30 PM • May 22, 2026

](https://twitter.com/bcherny/status/2057876668501434603)

In the world of increasing prices for models,  Deepseek says that they have made their 75% discount originally ending May 31st 2026 for V4-Pro **permanent** if one uses DeepSeek's inference endpoint.     
Recommend group members do their own research to interpret the terms of service/privacy policy under which Deepseek's own  inference endpoint is served   Deepseek v4-Pro is an open-source/open weight model so there are many inference providers which serve this model and each inference provider serves it under their  own ToS/privacy policy

[![User Avatar](https://storage.googleapis.com/papyrus_images/5e1311e1bfa0da893f7934d2831628d79c254fbc3e39a50634635be0803c483e.jpg)](https://twitter.com/deepseek_ai)

[DeepSeek](https://twitter.com/deepseek_ai)

[@deepseek\_ai](https://twitter.com/deepseek_ai)

[](https://twitter.com/deepseek_ai/status/2057854261699195173)

We are making our discount permanent! ![🎉](https://abs-0.twimg.com/emoji/v2/72x72/1f389.png)  
  
Enjoy building with DeepSeek-V4-Pro and bring your innovative ideas to life! ![🚀](https://abs-0.twimg.com/emoji/v2/72x72/1f680.png)

![](https://storage.googleapis.com/papyrus_images/7fe65b75080bb6bcb3f41444e5cfe8f7e0801d242a58b7b92f9ac0829670b9ab.jpg)

[![User Avatar](https://storage.googleapis.com/papyrus_images/5e1311e1bfa0da893f7934d2831628d79c254fbc3e39a50634635be0803c483e.jpg)](https://twitter.com/deepseek_ai)

[DeepSeek](https://twitter.com/deepseek_ai)

[@deepseek\_ai](https://twitter.com/deepseek_ai)

[](https://twitter.com/deepseek_ai/status/2049312932014813344)

The DeepSeek-V4-Pro discount has been extended until May 31, 2026, 15:59 UTC!

![](https://storage.googleapis.com/papyrus_images/0f863d780e2a4e0b8805554fbe7d7170349922e58c9ed3dbcf4567bdb10335dc.jpg)

[13.6K](https://twitter.com/deepseek_ai/status/2057854261699195173)[

4:01 PM • May 22, 2026

](https://twitter.com/deepseek_ai/status/2057854261699195173)

via [Robby Yung](https://t.me/robyung)

an AI demo where Gemini Omni transforms a Google Maps screenshot with a drawn route into a realistic first-person taxi driving video through city streets, complete with accurate buildings, traffic, and dashboard details.

[![User Avatar](https://storage.googleapis.com/papyrus_images/4cad1cd0a7636b42e817b40128adc55b89457e4a6b9e4718e14525415785d515.jpg)](https://twitter.com/giffmana)

[Lucas Beyer (bl16)](https://twitter.com/giffmana)

[@giffmana](https://twitter.com/giffmana)

[](https://twitter.com/giffmana/status/2057891707140645234)

Can't believe we're getting this before GTA 6

[![User Avatar](https://storage.googleapis.com/papyrus_images/884aaec7e196123ae533a7fefcb80938b7395870e6ef2e9711b089ed45bcee9d.jpg)](https://twitter.com/chrisfirst)

[CHRIS FIRST](https://twitter.com/chrisfirst)

[@chrisfirst](https://twitter.com/chrisfirst)

[](https://twitter.com/chrisfirst/status/2057612088651252062)

I uploaded a screenshot of Google Maps to Gemini Omni with a route drawn on it.  
  
Then I prompted it to create a first person view of someone driving a taxi cab along the route in the reference image.  
  
Pretty close to the real thing.

![](https://pbs.twimg.com/amplify_video_thumb/2057604891603058688/img/EN2vYBZchMJqgQ9I.jpg)

[508](https://twitter.com/giffmana/status/2057891707140645234)[

6:29 PM • May 22, 2026

](https://twitter.com/giffmana/status/2057891707140645234)

* * *

Below is my personal website which aggregates links to many of my socials as well as the various content and community that I curate. Feel free to share this link to others who you think may find this content/community useful to them

[**https://linktr.ee/goolamabbas**](https://linktr.ee/goolamabbas)

The cover image of this newsletter via generated via the **OpenAI's ChatGPT Images 2.0** model within the [**Freepik**](https://www.freepik.com/) tool via the following prompt

![](https://paragraph.com/editor/callout/information-icon.png)

A fascinating view through a natural rock formation by the sea, which serves as a frame for the scene. The dark, jagged rocks are visible in the foreground, lapped by gentle, turquoise waters. A luxurious yacht can be seen through the opening in the rock face, lying calmly on the clear blue water. Behind it stretches a hilly landscape that glows in soft colors under bright daylight. The contrasting composition between the rough, dark rocks and the elegant, modern yacht creates a harmonious balance of natural beauty and luxurious lifestyle. (Nature and luxury, rich in contrast, maritime scene, turquoise water)

---

*Originally published on [This Week in All Things AI](https://paragraph.com/@twiata/this-week-in-all-things-ai-week-21-2026)*
