Tech@Youtube on MacWorks

2026-03-21

Mon, 01 Jan 0001 00:00:00 +0000

Tech Videos — 2026-03-21#

Watch First#

If you only have time for one, watch the Google Cloud Tech multi-agent tutorial. It is the only video in this batch that provides concrete code, architectural diagrams, and a live deployment of Model Context Protocol (MCP) servers and agent routing patterns.

Video Rundown#

Build a multi-agent system | Hands On AI (Part 1) · Google Cloud Tech This hands-on lab provides a detailed code walkthrough for building multi-agent systems utilizing Google’s Agent Development Kit (ADK) and the Model Context Protocol (MCP). The presenters demonstrate how to write and configure MCP servers to expose external APIs, mathematical functions, and Cloud SQL databases, followed by deploying them to Cloud Run. Technically, the most valuable segment is the practical implementation of sequential, parallel, and loop routing patterns to orchestrate sub-agents. Verdict: Worth watching in full (or skimming the codebase) if you are actively building agent orchestrations or evaluating MCP on GCP.

2026-03-22

Mon, 01 Jan 0001 00:00:00 +0000

Sources

Tech Videos — 2026-03-22#

Watch First#

Nicole Forsgren’s talk at the Pragmatic Summit is the standout today. She accurately identifies that while AI tools have hyper-accelerated the coding inner-loop, they are absolutely crushing downstream CI/CD, review, and deployment pipelines.

2026-03-23

Mon, 01 Jan 0001 00:00:00 +0000

Sources

Tech Videos — 2026-03-23#

Watch First#

Watch Hung-yi Lee’s 加快語言模型生成速度 (2/2)：KV Cache to deeply understand the brutal VRAM bottlenecks of LLM inference. If you want to know why we are inventing architectures like Multi-Head Latent Attention and Group Query Attention, this breaks down the math and memory constraints better than any high-level blog post.

2026-03-24

Mon, 01 Jan 0001 00:00:00 +0000

Sources

Tech Videos — 2026-03-25#

Watch First#

Unlock Predictable, High-Performance Serverless with AWS Lambda Managed Instances: This is the single video most worth your time because it introduces a tangible architectural shift—attaching dedicated EC2 compute to Lambda functions—that demonstrably cuts steady-state serverless workload costs by up to 80% without losing managed infrastructure benefits.

2026-03-25

Mon, 01 Jan 0001 00:00:00 +0000

Sources

Tech Videos — 2026-03-25#

Watch First#

Watch the NVIDIA GTC 2026 Open Models Panel for a high-signal discussion on the shift from monolithic LLMs to orchestrated “compound agents” and why open-source models are strictly necessary for enterprise data control and air-gapped infrastructure.

2026-03-26

Mon, 01 Jan 0001 00:00:00 +0000

Sources

Tech Videos — 2026-03-26#

Watch First#

Building LiveKit Agents with Gemini Live API offers a highly practical, code-level tutorial on building a local voice agent using Google’s new native audio model (Gemini 3.1 Flash), providing actual solutions for real engineering headaches like speaker drift and tool-calling latency.

2026-03-27

Mon, 01 Jan 0001 00:00:00 +0000

Sources

Tech Videos — 2026-03-27#

Watch First#

The limits of AI scaling laws - NVIDIA CEO explains is the single most valuable watch today because it lays out the architectural roadmap for hardware moving forward, explaining exactly why inference (test-time scaling) is fundamentally harder and more compute-intensive than pre-training.

2026-03-28

Mon, 01 Jan 0001 00:00:00 +0000

Sources

Tech Videos — 2026-03-28#

Watch First#

How to Build a production-ready RAG AI agent from Google Cloud Tech is the most actionable watch today. It provides a credible, hands-on walkthrough of moving from batch LLM processing to a low-latency, real-time RAG pipeline using Cloud SQL, pgvector (HNSW indexing), and Apache Beam for scalable stream processing.

2026-03-29

Mon, 01 Jan 0001 00:00:00 +0000

Sources

Tech Videos — 2026-03-29#

Watch First#

From skeptic to true believer: How OpenClaw changed my life | Claire Vo This is a rare, pragmatic deep-dive into orchestrating local open-source AI agents that cuts through the peak hype cycle to show actual utility and system limitations,. It offers credible, technical advice on sandboxing, managing context windows, and avoiding prompt injection risks when giving models access to your local machine,,,.

2026-03-30

Mon, 01 Jan 0001 00:00:00 +0000

Sources

Tech Videos — 2026-03-30#

Watch First#

NVIDIA Nemotron Unpacked: Build, Fine-Tune, and Deploy Open Models From NVIDIA is an outstanding, hype-free technical deep dive into NVIDIA’s new hybrid Mamba-2 Transformer architecture and their bleeding-edge 4-bit pretraining methods.

2026-03-31

Mon, 01 Jan 0001 00:00:00 +0000

Sources

Tech Videos — 2026-03-31#

Watch First#

CUDA: New Features and Beyond | NVIDIA GTC NVIDIA details “Green Contexts” for dynamic GPU SM partitioning and the new CUDA Tile compiler stack, fundamentally changing how engineers can optimize concurrent prefill and decode inference workloads without the overhead of multi-process solutions.

2026-04-01

Mon, 01 Jan 0001 00:00:00 +0000

Sources

Tech Videos — 2026-04-01#

Watch First#

Scaling Uber with Thuan Pham (Uber’s first CTO) is an incredibly substantive, hype-free breakdown of surviving engineering hypergrowth. It provides a rare inside look at how Uber was forced into a microservices architecture not out of design purity, but purely to survive scale before their NodeJS dispatch monolith hit a hard computational brick wall.

2026-04-02

Mon, 01 Jan 0001 00:00:00 +0000

Sources

Tech Videos — 2026-04-02#

Watch First#

An AI state of the union: We’ve passed the inflection point & dark factories are coming Simon Willison delivers an incredibly pragmatic breakdown of “agentic engineering,” explaining how to safely leverage AI coding agents using test-driven development, and offering a sobering technical explanation of why prompt injection (the “lethal trifecta”) remains an unsolved security flaw.

2026-04-03

Mon, 01 Jan 0001 00:00:00 +0000

Sources

Tech Videos — 2026-04-03#

Watch First#

37,000 Lines of Slop A vital, pragmatic teardown of AI-generated code hype that demonstrates why blindly shipping 37,000 lines of LLM output a day results in catastrophic, unreviewed production payloads.

2026-04-04

Mon, 01 Jan 0001 00:00:00 +0000

Sources

Tech Videos — 2026-04-04#

Watch First#

Skip the heavy vendor demos today and watch Lenny’s Podcast’s brief clip on “The cognitive cost of AI coding”. It provides a much-needed reality check on the cognitive burnout engineers face when managing multiple AI coding agents simultaneously, separating the hype of parallelized agent output from the actual biological limits of developer context-switching.

2026-04-05

Mon, 01 Jan 0001 00:00:00 +0000

Sources

Tech Videos — 2026-04-05#

Watch First#

Anthropic’s $1B to $19B growth run: how Claude became the fastest-growing AI product in history from Lenny’s Podcast offers a rare, operationally dense look at how a company scaled its ARR by 19x in 14 months by augmenting engineers with AI and actively eliminating traditional PM overhead.

2026-04-06

Mon, 01 Jan 0001 00:00:00 +0000

Sources

Tech Videos — 2026-04-06#

Watch First#

Insights from NVIDIA Research | NVIDIA GTC is the standout watch today, offering a dense, highly credible look into how GPU hardware architectures are physically evolving to support high-throughput LLM inference alongside novel reinforcement learning pre-training techniques.

2026-04-07

Mon, 01 Jan 0001 00:00:00 +0000

Sources

Tech Videos — 2026-04-07#

Watch First#

Agentic Engineering: Working With AI, Not Just Using It — Brendan O’Leary A highly pragmatic talk on moving from “AI as autocomplete” to “AI as collaborator,” outlining a concrete “Research, Plan, Implement” workflow that prevents coding agents from hallucinating or mutating your architecture blindly.

2026-04-08

Mon, 01 Jan 0001 00:00:00 +0000

Sources

Tech Videos — 2026-04-08#

Watch First#

Why, and how you need to sandbox AI-Generated Code? — Harshil Agrawal, Cloudflare from the AI Engineer channel is the most critical watch of the day. It strips away the AI hype to state a fundamental truth: if your agent executes generated code, you are running untrusted code from the internet in production. It delivers a strict, pragmatic capability-based security framework for deciding when to use V8 Isolates versus full Linux containers to prevent credential leaks and compute exhaustion.

Youtube Tech Channels

Mon, 01 Jan 0001 00:00:00 +0000

Tech Videos — 2026-03-19#

Watch First#

Google just changed the future of UI/UX design… Google’s new “Stitch” tool introduces a “design markdown file” export that bridges the gap between generated UI components and LLM coding agents, which could actually shift frontend workflows if it works as demoed.

Video Rundown#

Google just changed the future of UI/UX design… · Fireship Fireship covers Google “Stitch”, an infinite canvas UI/UX generator that creates interactive prototypes, responsive components, and full user flows from text prompts or reference URLs. The most technically notable feature is its ability to export a “design markdown file” containing the design system, which can be fed directly into models like Claude or OpenAI Codex for consistent UI generation across projects. Verdict: Worth watching in full to see if this design markdown paradigm is a viable replacement for traditional CSS/Tailwind workflows.

Youtube Tech Channels

Mon, 01 Jan 0001 00:00:00 +0000

Tech Videos — 2026-03-20#

Watch First#

Andrej Karpathy’s interview provides a highly credible look into the cutting-edge reality of “AI psychosis,” where human engineers are now the primary bottleneck in software development,. It is an essential, paradigm-shifting watch on how to stop writing code manually and start orchestrating asynchronous, autonomous agent loops using markdown specifications,.

Video Rundown#

Google Cloud Live: Supercharge your AI agents: Inside the new ADK integrations ecosystem · [Google Cloud Tech] This video introduces Google Cloud’s new ADK integrations ecosystem designed to enhance AI agent capabilities. It outlines the technical patterns for integrating agents directly into Google Cloud infrastructure. No detailed transcript was provided for deep technical analysis. Verdict: Skim for specific API integration patterns if you are currently building agents on GCP.