Engineer

Joined @SecRubio in Kolkata for a visit to the Missionaries of Charity. Moments like these remind us the U.S.-India partnership rests not only on strong policies, but also on shared values and the spirit of selfless service that transcends borders.

Mother Teresa left a tremendous legacy of compassion and service. I was honored to visit the Missionaries of Charity today to pay homage to her legacy and to see the living example of the Catholic faith in action.

The Madras High Court has called Parental Alienation a menace to children and observed that it is inhuman to turn a child against a parent. #MadrasHighCourt #SupremeCourtOfIndia #ParentalAlienation #ChildCustody #SharedParenting

NEW: Students go nuts after donor announces during his commencement speech that he is paying off all of their senior year debts. Anil Kochhar and his wife decided to give the gift to all ~200 graduates in N.C. State's family. Kochhar is the son of Prakash Chand Kochhar, an immigrant from India who studied textile manufacturing in Raleigh. "My father found not just an education, but an opportunity that allowed him to build a life, support his family, and begin a legacy that continues today. And it will never stop, never," Kochhar said.

"The rights of man come not from the generosity of the state, but from the hand of God." —John F. Kennedy

When asked what her 14-year-old son should study to become a car engineer, Elon Musk’s answer was simple: physics. Physics teaches you to think from the ground up, to break problems apart and rebuild them using logic instead of assumptions.

@philz1337x Is there an updated guide for 2026 ? would like to know improved models in use.

🧑‍💻 How to Get a Job as an AI Image Developer or Build Your Own AI Image Apps I quit my $100k engineering job to make it as a generative AI developer. There are many ways, but here is how I have been doing it for the last 12 months: 1. Create AI images on your computer with Stable Diffusion. A1111 is a good tool to begin with: github.com/AUTOMATIC1111/… 2. Extend your A1111 skills by learning how to use img2img and ControlNet. @SebastianKamph has great videos: @sebastiankamph" target="_blank" rel="nofollow noopener">youtube.com/@sebastiankamph 3. If you are new to coding, try ComfyUI; it allows you to build complex workflows without coding: github.com/comfyanonymous… Now it’s time for coding, but you can also avoid coding and go directly to step 8. 4. Create an image with code only. Diffusers is a great Python library because it’s quite simple, but you can do very complex stuff. Install VS Code and diffusers with "pip install diffusers." Run these 5 lines of code: from diffusers import AutoPipelineForText2Image import torch pipeline = AutoPipelineForText2Image.from_pretrained( "runwayml/stable-diffusion-v1-5", torch_dtype=torch.float16, variant="fp16" ).to("cuda") image = pipeline( "photo of a dog" ).images[0] image huggingface.co/docs/diffusers… 5. Extend your diffusers skills: Replace the base model with a custom checkpoint like Realistic Vision. Use img2img. Check the diffusers docs and try something; it’s well documented. huggingface.co/docs/diffusers… AI code for AI apps needs to handle multiple requests at a time, so they need to run on scalable GPUs. @replicate auto scales AI models on hundreds of GPUs. 6. Push your first replicate model: Install cog #install" target="_blank" rel="nofollow noopener">github.com/replicate/cog#…, clone the hello world example github.com/replicate/cog-…, create a new model on replicate and push your model with "cog login" and "cog push r8.im/" replicate.com/docs/guides/pu… 7. Now the final step is to bring your custom AI model with diffusers to replicate. For this step, you need a GPU, so go to cloud.lambdalabs.com/instances and launch an A10 GPU for example. Install cog, copy your hello world example, and push to replicate. replicate.com/docs/guides/ge… extend the example with your diffusers code to create images from text. Try your code with "cog predict" and push to replicate. replicate.com/docs/guides/pu… 8. If you like ComfyUI, you can also build an API for an AI app with ComfyUI on replicate. Try this tutorial from @fofr: replicate.com/guides/comfyui… You did it! You are able to create basic workflows with A1111, ComfyUI, and the diffusers library, and create an auto-scalable AI model on replicate which runs on GPUs. Now you need a project to extend your knowledge by getting challenged, apply for a job as an AI developer, or build AI models for your own AI app like I did for my upscaler, ClarityAI.co

KING CHARLES III: “The Christian faith is a firm anchor and daily inspiration that guides us not only personally…but together as members of our community.”

🚨Wow! Scans just uncovered tunnels and possible support beams beneath the massive boat formation long claimed as Noah’s Ark in Turkey. Decades after early explorers flagged it — cutting-edge tech now backing the biblical site.

@Sam_Badawi what are best ETFs for AI infra ?

The network layer is starting to find its footing in the AI infrastructure stack as inference scales, helping explain the continued strength in names like $ANET, $ALAB, and $CRDO. As hyperscalers push distributed training and inference workloads tied to the $100B deal between $AMZN and Anthropic, high-speed interconnect and custom silicon from $MRVL and $AVGO are becoming critical parts of the scaling equation. That shift is turning networking into a structural tailwind rather than a supporting layer as AI clusters expand beyond single-node compute into full-fabric infrastructure. Credit to @SergeyCYW for the image

@jun_song would like to know the optimized command flags to run mlx and llama.cpp on mac silicon.

로컬LLM 가장 많이 물어보는 질문 : Mac Mini 16gb에서 어떤걸 사용하나요? >SuperGemma4-e4b-mlx-4bit 제 VRAM에는 어떤 모델을 사용할 수 있나요? > 대략 모델 사이즈를 2로 나누면 필요한 VRAM (35b -> 16gb, e4b(8b) -> 4gb) 어떤 포맷이 좋은가요? > 맥 = mlx / 엔비디아 = nvfp4

If you’ve been paying even minimal attention to American life, you know our boys and men are in crisis. For years, much of elite opinion responded with a shrug: “Boys just need to try harder” or “It’s women’s turn.” But now, both parties are finally noticing that millions of disconnected young men are a serious concern yet also represent political opportunity. Whatever the motives, we welcome the attention. Reps. Nathaniel Moran (R-Texas) and Derrick Van Orden (R-Wis.) introduced the White House Council on Fathers and Sons Act of 2026 last month. After nearly two decades of advocacy by scholars and activists led by @drwarrenfarrell, the idea has arrived at the perfect moment. Check out the latest from @mattklewis and I over on S to learn more: open.substack.com/pub/lisabritto…

@siddharthabvc Where to get similar printed books with SC/HC family law judgements. please share links to buy. thanks in advance.

NEW: Astronaut Reid Wiseman, who says he is not religious, says he broke down in tears when he saw the Cross after getting back to Earth. Wiseman: "I'm not really a religious person, but there was just no other avenue for me to explain anything..." Victor Glover: "The only thing I would add is I am a religious person, but everything else is the same."

There are no words.

With the resignation of Justice Yashwant Verma, here’s a replug of my zero hour intervention on accountability in judiciary

@VladArtym would love to try.

lemme know if I should open source this: vladartym.com/interfaces/alw…

Okay I think I finally found a good use for @_chenglou 's Pretext. A resume builder that always fits on one page with realtime rendering.

@mfranz_on which model is your pick to detect fall and other gestures ? on edge, low power devices for a closed source commercial use.

Today I want to talk about Nipotino. Nipotino means grandson in Italian, and this is the project where I'm distilling all my various interests, from computer vision to embedded systems. An intelligent device to serve our elderly, frail, and lonely loved ones. An ESP32 that listens and responds, connected to an LLM for advanced conversations. The LLM will be local or cloud-based; if local, for greater privacy, a dedicated PC will be required for hosting. The device will have a small camera to detect falls or to help with forgetfulness, such as the refrigerator open, the light on, the window open, etc. The device will be able to make emergency calls to predefined numbers in the event of trigger events or upon specific requests. In some ways, it may resemble an existing home assistant, but that's not the case. The idea is to focus the application on specific needs and a personality suited to the purpose. Therefore, it won't be a generalist AI, but rather a dedicated AI for personal support. I would like this project to be completely open, so that as many people as possible can make it happen and contribute to its development. I think this can have a social impact so let's try to make it for real.

@MactelecomN @Ubiquiti cool. Ubiquiti released good products in the past year.

Found a @ubiquiti UTR . Gonna give it away like this post to enter will give announce the winner March 31 ( no this isn’t to gain followers but to give back to people. ) I’ll pay the shipping whoever wins