Tech Videos — 2026-05-24#

Watch First#

The AI paradox: More automation, more humans, more work | Dan Shipper from Lenny’s Podcast offers the most pragmatic signal today, arguing that AI automation is actually creating more demand for engineering review and pushing IDEs to become the primary operating system for all knowledge work. Instead of replacing engineers, models like GPT-5.5 require heavy oversight, turning software development into a process of managing agents and reviewing AI-generated code.

Highlights by Theme#

Developer Tools & Platforms#

In the developer space, Dan Shipper predicts the “CLI era” is effectively dead, arguing that AI-integrated desktop apps like Codex are replacing the terminal by embedding browsers directly into the IDE to observe developer workflows alongside them. On the historical design front, Anders Hejlsberg: C# was designed by 6 people via The Pragmatic Engineer serves as a quick reminder of software fundamentals, noting that C# was designed by a tight-knit group of six or seven people meeting for two hours three times a week to rigorously shoot down each other’s ideas. Other platform videos hitting the feed include GitHub’s How to find your first open source project and ThePrimeagen’s commentary in Microsoft just doesn’t get it.

AI & Machine Learning#

For applied AI tooling, NVIDIA CEO Jensen Huang at Meet-a-Claw in Taipei introduces “Open Claw”, an open-source harness that essentially acts as an operating system wrapped around large language models for building agents. Dan Shipper shares a notable benchmark on Lenny’s Podcast, revealing that GPT-5.5 scored a 60/100 on his custom “VibeCoded” evaluation compared to a human senior engineer’s 90/100, though he notes it is the first model with enough autonomous agency to successfully rip out old code and rewrite from first principles rather than just patching edges. Additional deep dives into agent architecture include Scaling the Next Paradigm of Heterogeneous Intelligence, Let’s Talk About FOMAT: Fear of Missing Agent Time, and How Google DeepMind Runs Agents at Scale from the AI Engineer channel, alongside Hung-yi Lee’s AI 要跨越盧比孔河了嗎？自我成長的 AI 離我們多遠 (下集).

Hardware & Infrastructure#

In cloud infrastructure, the All-In Podcast unpacks a massive $15B Anthropic compute deal in SpaceX is About to Explode: $15B from Anthropic is just the beginning. The core technical signal here is SpaceX’s ruthless optimization of data center construction for “Elon Web Services”—cutting their facility build times sequentially from 122 days, to 91 days, down to 66 days to rapidly deploy GPUs.

Everything Else#

On the historical and cultural front, All-In covers SpaceX’s Unknown Origin Story: Elon Wanted to Back Up Earth, noting Musk’s original vision was to send geodesic domes into space to back up the Earth’s biosphere. The remaining general interest videos today are Dwarkesh Patel’s Following the Yamnaya Trail into India - David Reich and a brief clip of NVIDIA CEO Jensen Huang at Taipei’s Raohe St. Night Market.