HM Godoy

Video by OpenCV via YouTube

Join our Patreon to support the show: https://patreon.com/opencv

We welcome back the team behind the VAND anomaly-detection challenge, a staple of recent CVPR conferences. VAND brings together cutting-edge research on detecting what doesn’t belong in visual data—spanning anomaly, novelty, and out-of-distribution detection. Building on three successful editions, VAND 4.0 unites supervised, semi-, and unsupervised approaches, including few-, one-, and zero-shot learning, with a strong focus on real-world impact.

Official site: https://sites.google.com/view/vand4-cvpr2026

Info for CVPR attendees: June 4th (1pm-6pm), 2026 in Denver, CO, USA (In Person) + Zoom (Virtual), Half Day
Room: 601, Posters: Exhibit Hall A

OpenCV is a 501(c)(3) registered non-profit in the United States. See how you can support open source CV & AI: http://opencv.org/support/

Watch along for your chance to win during our live trivia segment, and participate in the live Q&A session with questions from you in the audience.

Become a paid member of the channel to help us make more episodes https://www.youtube.com/channel/UCkrcW82Y2kbgU-U9RaYfgxw/join

Got a cool project of your own? Send it to us and you may be featured https://www.jotform.com/form/233105358823151

Source

Video by PyTorch via YouTube

Why WideEP Inference Needs Data-Parallel-Aware Scheduling – Maroon Ayoub, IBM; Tyler Michael Smith, Red Hat

WideEP—wide expert parallelism fails not because experts are expensive, but because routing ignores where state already lives. In PyTorch LLM serving with vLLM, WideEP fans tokens across many experts while KV caches accumulate unevenly across data-parallel replicas. When routing is unaware of KV placement and per-replica load, requests land on replicas that cannot reuse cache or make progress efficiently and latency spikes as expert fan-out grows.

The fix is not reshaping expert parallelism, but making routing data-parallel aware using signals vLLM already exposes. In this talk, we show how llm-d extends its router to leverage KV-cache locality and load awareness when routing WideEP flows. Rather than treating replicas as interchangeable, the router prefers replicas with warm KV state and available capacity, aligning routing decisions with vLLM’s execution reality and reducing cache fragmentation.

This session walks through how KV-aware, data-parallel routing changes WideEP inference in practice: which signals matter, how routing behavior evolves, and where the gains come from. Attendees leave with a clear mental model for when KV- and load-aware routing unlocks higher throughput.

Source

Video by NetApp Instaclustr via YouTube

While LLMs produce fantastic results, the costs can be steep. A small language model is a great option for limited budgets and resources while still providing solid, fast outputs.

Learn how to build a SLM from scratch with Sr. AI developer advocate David VonThenen.

Demo: https://www.youtube.com/watch?v=VUuVro-Dv7c

Be sure to subscribe for all things AI!

Timestamps:
00:46: What is a small language model?
01:55: Setting up a nanoGPT-style casual LM
09:26: Building our model
13:38: Results
14:59: Fine-tuning the model
21:06: The types of LMs

Source

Video by FOSSASIA via YouTube

Welcome to the livestream of FOSSASIA 2026 taking place in Bangkok, Thailand as part of the FOSSASIA Summit. Community Day brings together open source contributors, developers, students, maintainers, and technology leaders from across Asia and around the world.

The program features talks, discussions, and community sessions covering:

Artificial Intelligence and Machine Learning

Cloud Infrastructure and DevOps

Cybersecurity and Digital Safety

Web and Mobile Development

Open Hardware and Embedded Systems

Databases and Data Engineering

Open Source community collaboration

Speakers include engineers, researchers, and maintainers from global technology companies and major open source projects. Participants share real world experience, technical insights, and community initiatives shaping the future of open technology.

FOSSASIA is a non profit organization supporting open technologies and developer communities across Asia. Through conferences, hackathons, mentoring programs such as Google Summer of Code, and open source projects, FOSSASIA connects contributors and organizations working on impactful technology.

Location: Bangkok, Thailand
Event: FOSSASIA Summit 2026
Track: Community Day

Learn more about the event and upcoming activities:
https://fossasia.org

Join the community:
https://github.com/fossasia

Follow FOSSASIA for updates on open source events, projects, and collaborations across Asia.

#FOSSASIA #OpenSource #AI #Cloud #Developers #Bangkok #FOSSASIASummit

Source

Video by CNCF [Cloud Native Computing Foundation] via YouTube

Don’t miss out! Join us at our next KubeCon + CloudNativeCon events in Mumbai, India (18-19 June, 2026), Yokohama, Japan (29-30 July, 2026), and Shanghai, China (8-9 September, 2026). Connect with our current graduated, incubating, and sandbox projects as the community gathers to further the education and advancement of cloud native computing. Learn more at https://kubecon.io

Project Lightning Talk: A Curator’s Guide to the CNCF Landscape – Katherine Druckman and Lori Lorusso

Step into the studio with two of the Netherlands’ most famous masters as we explore the CNCF Landscape like a gallery of modern technical masterpieces. Don your berets and join Rembrandt and Van Gogh as we paint a clearer picture of the CNCF ecosystem.

With more than 190 projects across the landscape, discovering what each one does can feel a bit like wandering through an enormous museum without a guide. “Just go to the website” isn’t always enough. Sometimes you need a curator to point out the highlights, explain the movements, and help you see how the pieces fit together.
In this lightning session, we’ll tour a selection of CNCF projects the way art historians might walk through a gallery, highlighting the themes, techniques, and innovations that make them stand out. By the end, you’ll have a clearer mental map of the landscape and be ready to navigate KubeCon like a seasoned collector.

Source

Video by Meta Open Source via YouTube

Meta is partnering with the Oscar Mike Foundation and using Aria Gen 2, our next-generation research glasses, to make advancements in accessibility like providing memory assistance and improving spatial awareness for people with disabilities.

Aria Gen 2 is continuing to partner with researchers, organizations, and people with disabilities as we bring the human perspective to AI.

Learn more about Oscar Mike: https://oscarmike.org/pages/mission
Read the blog post: https://about.fb.com/news/2026/02/helping-people-with-memory-loss-with-ai-glasses/
Partner with Project Aria: https://www.projectaria.com/partners/
Follow Project Aria on X: https://x.com/meta_aria

Source

Video by Software Engineering Daily via YouTube

#SubseaCables #CyberSecurity #InternetInfrastructure #geopolitics

Subsea cables are high-capacity fiber-optic lines laid along the ocean floor to enable global communication by transmitting data between continents. Spanning thousands of miles, they carry an estimated 95% of international internet, phone, and data transmissions.

Critically, these cables are vulnerable to sabotage by state actors, as they form critical infrastructure for global communication and economic stability. Indeed, Russia and China have been implicated in activities targeting subsea cables as recently as November 2024, and experts warn that these networks are likely to be focal points in future conflicts, heightening geopolitical tensions.

Josh Dzieza is a reporter for The Verge and has covered the subsea cable industry and the strategic importance of subsea cables. He joins the podcast alongside Gregor Vand to discuss this invisible, and increasingly important, network infrastructure.

⌚Timestamps:

00:00 Introduction and Background
01:25 Interview

👍🏼 Give your thumbs-up and subscribe to the channel!

🎙️ Check out more of our podcasts
http://softwareengineeringdaily.com/

Twitter http://twitter.com/software_daily

LinkedIn http://www.linkedin.com/company/software-engineering-daily

Source

Video by Reality 2.0 Podcast via YouTube

Is goodwill enough to sustain billion-dollar ecosystems? Join Katherine Druckman and Lori Lorusso as they dive into the complexities of open source sustainability and why everyone needs to pay attention. #OpenSource #techtalk

Full episode: https://www.reality2cast.com/159

Source

Video by DataStax via YouTube

When quantum meets AI, everything changes. ⚡️🤖

On the next Context Window Podcast, Ed Anuff and Anant Jhingran are joined by Alan Ho, CEO & Co-Founder of Qolab, to explore how quantum could quietly reshape the future of AI, from drug discovery to… well, maybe digital apocalypse 🧟 .

We’ll dig into:
👉 How quantum may first show up inside AI workflows
👉 Why small performance gains can mean massive advantage
👉 Where classical, quantum, and AI start working together
👉 What this shift means for builders and product leaders

Tune in for a conversation about where the next edge in AI might come from.

If quantum + AI sounds interesting… but still a little fuzzy 🤪, this one’s worth checking out.

🚨 Also, we’re moving the Context Window Podcast to our new YT page @ContextWindowPodcast so follow along as we roll things out for the upcoming episodes!

Source

Video by Changelog via YouTube

Adam and Jerod get into the news, Jerod officially retires from the pod (and Changelog), plus a bonus for our [Changelog++ subs](https://changelog.com/++)!

SUBSCRIBE:

RSS: https://changelog.am/rss
Apple: https://changelog.am/apple
Spotify: https://changelog.am/spotify
Android: https://changelog.am/android
Overcast: https://changelog.am/overcast
Email: https://changelog.am/email
X: https://x.com/changelog
Bluesky: https://bsky.app/profile/changelog.com
Mastodon: https://changelog.social/@changelog

SPONSORS:

Augment Code – Adam loves "Auggie" — Augment Code’s CLI that brings Augment’s context engine and powerful AI reasoning anywhere your code goes. From building alongside you in the terminal to any part of your development workflow. https://www.augmentcode.com/

Squarespace – Turn your expertise into a business with the all-in-one platform for websites, services, and getting paid. Use code `CHANGELOG` to save 10% on your first website purchase. https://www.squarespace.com/changelog

Notion – Custom Agents that automate the busywork so your team can focus on real work. Try them free at notion.com/changelog http://notion.com/changelog

LINKS:

– https://ladybird.org/posts/adopting-rust/
– https://oxc.rs
– https://oxc.rs/sponsor
– https://x.com/boshen_c
– https://www.reddit.com/r/devops/comments/1rcc7en/am_i_the_only_one_who_genuinely_prefers_onprem/
– https://boristane.com/blog/the-software-development-lifecycle-is-dead/
– https://www.citriniresearch.com/p/2028gic
– https://x.com/illscience/status/2025775386760732845
– https://wesmckinney.com/blog/mythical-agent-month/
– https://nanoclaw.dev
– https://zed.dev/acp

CHAPTERS:

00:00 Friends Forever
00:03 Finale & Friends
00:40 We have history
01:50 Back in the 2012 days
04:48 It’s been a journey!
06:15 It’s "Changelog" (sweating the details)
06:56 Back to the dirt
07:49 Ladybird cracks and is now using Rust
15:21 Rust vs Go
16:07 C++ style Rust
20:50 Contributing to Ladybird?
22:26 The cost of Agents
26:32 Industrial-scale distillation attacks
31:05 Who hasn’t done a little distillation?
32:56 Bonus Rust mentions
33:19 Rust tooling for JavaScript from Oxc
37:36 The future is seft-hosted and on-prem
39:24 Jerod goes on-prem with Mac mini
41:52 We’ll do it on-prem!
42:49 On-prem over Cloud
44:37 Self-host it first
48:13 100+ tailnet devices?
50:52 turk.run is coming soon
51:21 SS – Incus website
55:53 github.com/dnshole
57:52 RAM and storage costs are UP!
59:01 NanoClaw!!
1:05:09 Maintaining OSS in the era of agents
1:05:52 Code review is dead?
1:10:57 The SDLC Is dead
1:17:15 Queues are biggest source of waste in development
1:20:48 Only so many keystrokes left
1:21:26 When the agent is confidently !wrong
1:22:19 Slingn’ Swift code
1:24:10 That’s ReMarkable!
1:25:16 Explore new worlds! Now.
1:29:29 Gosh babe!!
1:30:59 I hate goodbyes
1:32:54 ++ BONUS??!
1:33:35 Enjoy some extended good beats

Source