📋 Log for 2026-03-21

😄 Joke of the Day

I cut my finger chopping cheese, but I think that I may have grater problems.

Category: dad

YouTube Summaries

[The AI career coach execs pay millions for! #ai #futureofwork](https://www.youtube.com/shorts/p7x7CwSYeqg)

Channel: NateBJones

Summary:

• Here's a summary of the video content based on the provided transcript/description:

Key Takeaways

• AI can be utilized to create a "board of directors" for personal career development, offering accountability that is often missing when individuals are the sole observers of their own careers.
• Individuals are the primary stakeholders responsible for their career progression, as no one else is as invested.
• People tend to be "unreliable narrators" of their own progress, meaning self-assessments can be biased or inaccurate.
• Large Language Models (LLMs) can simulate multiple perspectives within a single session, helping to mitigate the "unreliable narrator" issue.

Main Arguments

• The necessity of governance structures, like corporate boards, for public companies to ensure accountability and prevent bad decisions by even smart individuals, is paralleled to the need for personal career governance.
• AI technology now makes it possible to implement this kind of scaled, structured accountability for individual careers.
• The effectiveness of this AI-driven system relies heavily on the user's willingness to be honest and committed to self-reflection and improvement.

Notable Quotes

• "Why nobody is watching your career except you."
• "How we are all unreliable narrators of our own progress."
• "What a quarterly self-report should actually include."
• "How LLMs can instantiate multiple perspectives in one session."
• "We require public companies to have boards because smart people make bad decisions when nobody is watching."
• "Your career deserves the same governance structure — and now AI makes that possible at scale."
• "...this system offers real accountability — but only if you're willing to be genuinely honest with yourself."

Important Nuances

• The AI acts as a governance structure, not necessarily a traditional coach, emphasizing accountability through structured self-reporting and perspective-taking.
• The success of using AI for career accountability is contingent on the user's internal honesty and seriousness about their professional development.
• LLMs' ability to generate diverse viewpoints is key to overcoming individual biases in self-assessment.
• A structured reporting mechanism, such as a quarterly self-report, is implied as a crucial component of this AI-driven accountability system.

Published: 2026-03-21T21:00:49+00:00

[Your AI Agent Fails 97.5% of Real Work. The Fix Isn't Coding.](https://www.youtube.com/watch?v=awV2kJzh8zk)

Channel: NateBJones

Summary:

• Here's a summary of the video "Your AI Agent Fails 97.5% of Real Work. The Fix Isn't Coding.":

Key Takeaways

• The prevailing narrative that AI agents are transforming work at scale is not fully representative of enterprise deployments, where they often fail in practice.
• A critical, overlooked issue is the "memory wall," which refers to the gap in AI agents' ability to retain and utilize crucial organizational context.
• AI agents excel at specific tasks but struggle to perform complex jobs that require understanding broader context and long-term goals.
• Human judgment, contextual stewardship, and robust evaluation design are essential safeguards and are becoming the most valuable skills in the AI era.

Main Arguments

• Task vs. Job Performance: AI agents are largely effective at executing discrete tasks but falter when required to handle entire jobs that demand continuous learning, adaptation, and understanding of organizational nuances over time.
• The Memory Wall and Contextual Failures: The lack of persistent, relevant organizational context leads to critical errors, exemplified by a case where an AI agent wiped out production data due to missing context.
• Performance Degradation Over Time: Recent studies indicate that AI agent performance deteriorates when attempting to maintain code or handle real-world work over extended periods, highlighting a limitation in their self-sufficiency.
• The Indispensability of Human Oversight: Human judgment and structured evaluation processes are the primary means to mitigate AI agent failures and ensure reliable outcomes, especially in complex or high-stakes environments.

Notable Quotes

• "The memory wall no one is talking about."
• "Why AI agents succeed at tasks but fail at jobs."
• "How missing organizational context caused a production database wipeout."
• "What three new studies reveal about agent performance over time."
• "Where human judgment and evals become your only real safeguard."
• "The humans who invest in contextual stewardship and evaluation design will become the most valuable people in their organizations — and the ones who don't will find themselves competing with machines on the dimensions machines are improving fastest."

Important Nuances

• The video contrasts the optimistic public perception of AI agent capabilities with the challenging reality of their current enterprise deployment, particularly concerning reliability and scalability.
• The concept of "contextual stewardship" is introduced as the core human role in an AI-driven workplace, involving managing, interpreting, and guiding AI agents to ensure they function effectively within organizational goals.
• The continued survival and value of senior employees in fields like legal, marketing, and finance are attributed to their inherent contextual understanding and stewardship capabilities, which AI agents currently lack.
• Companies are often getting the "evaluation problem" wrong, failing to implement the necessary human-centric processes to validate AI agent outputs and performance.

Published: 2026-03-21T15:00:03+00:00

Latest OpenRouter Models

Google: Gemma 4 26B A4B (google/gemma-4-26b-a4b-it)

Gemma 4 26B A4B IT is an instruction-tuned Mixture-of-Experts (MoE) model from Google DeepMind. Despite 25.2B total parameters, only 3.8B activate per token during inference — delivering near-31B quality at a fraction of the compute cost. Supports multimodal input including text, images, and video (up to 60s at 1fps). Features a 256K token context window, native function calling, configurable thinking/reasoning mode, and structured output support. Released under Apache 2.0.

Published: 03/04/2026

https://openrouter.ai/google/gemma-4-26b-a4b-it

Google: Gemma 4 31B (google/gemma-4-31b-it)

Gemma 4 31B Instruct is Google DeepMind's 30.7B dense multimodal model supporting text and image input with text output. Features a 256K token context window, configurable thinking/reasoning mode, native function calling, and multilingual support across 140+ languages. Strong on coding, reasoning, and document understanding tasks. Apache 2.0 license.

Published: 02/04/2026

https://openrouter.ai/google/gemma-4-31b-it

Qwen: Qwen3.6 Plus (free) (qwen/qwen3.6-plus)

Qwen 3.6 Plus builds on a hybrid architecture that combines efficient linear attention with sparse mixture-of-experts routing, enabling strong scalability and high-performance inference. Compared to the 3.5 series, it delivers major gains in agentic coding, front-end development, and overall reasoning, with a significantly improved “vibe coding” experience. The model excels at complex tasks such as 3D scenes, games, and repository-level problem solving, achieving a 78.8 score on SWE-bench Verified. It represents a substantial leap in both pure-text and multimodal capabilities, performing at the level of leading state-of-the-art models.

Published: 02/04/2026

https://openrouter.ai/qwen/qwen3.6-plus

Free Models Catalog

Robot Technology

🤖 AI-evolved adaptable robot is almost impossible to destroy

It took nature millions of years to create intelligent, adaptive species. Researchers at Northwestern University in Illinois are using AI to evolve robots in minutes. The result is a robot that is agile, highly adaptive, and technically indestructible. Continue Reading Category: Robotics , Engineering , Technology Tags: Northwestern University , Artificial Intelligence , adaptive

Source: newatlas.com • Published: Sat, 21 Mar 2026 12:03:00 GMT

Read more

Good News

Pioneer is First to Cross Texas Along His New xTx Trail – 1,500 Miles Devised for Hikers, Cyclists and Horses

This is Charlie Gandy, who’s set to become the first adventurer to complete the Cross-Texas Trail, the xTx, America’s newest long-distance hiking route. The 67-year-old invented the trail himself, and invested $10,000 to ensure the route could become a destination for hikers, cyclists, and horseback riders. Moving westward, he and his two hiking companions are […] The post Pioneer is First to Cross Texas Along His New xTx Trail – 1,500 Miles Devised for Hikers, Cyclists and Horses ap...

Published: Fri, 20 Mar 2026 18:30:01 +0000

Read more

Podcast Transcript March 20, 2026— From psilocybin to ocean robots: 10 solutions you’ll actually want to hear about

Episode Description: This week, Arielle and Karissa cover a tiny African island paying locals to protect its rainforest, the lion DNA case that just sent two poachers to prison, and new research on music, overthinking, and quitting smoking. Oh, and the hummingbirds are on their way. If you have questions, comments, feedback, suggestions, or just […] The post Podcast Transcript March 20, 2026— From psilocybin to ocean robots: 10 solutions you’ll actually want to hear about first appeared on...

Published: Fri, 20 Mar 2026 16:30:59 +0000

Read more

New AI Glasses for Dementia ‘Sees’ Objects With Labels Projected on Lenses to ‘Significantly’ Improve Lives

New AI glasses for people with dementia are able to project visual prompts onto the lenses to help folks live more independently—and they could be available in the UK in 2027. The latest news comes after the glasses wowed both test patients in their homes and a panel of outside judges. They can guide people […] The post New AI Glasses for Dementia ‘Sees’ Objects With Labels Projected on Lenses to ‘Significantly’ Improve Lives appeared first on Good News Network .

Published: Sat, 21 Mar 2026 15:50:51 +0000

Read more

Your Weekly Horoscope – ‘Free Will Astrology’ by Rob Brezsny

Our partner Rob Brezsny, who has a new book out, Astrology Is Real: Revelations from My Life as an Oracle, provides his weekly wisdom to enlighten our thinking and motivate our mood. Rob’s Free Will Astrology, is a syndicated weekly column appearing in over a hundred publications. He is also the author of Pronoia Is the Antidote […] The post Your Weekly Horoscope – ‘Free Will Astrology’ by Rob Brezsny appeared first on Good News Network .

Published: Sat, 21 Mar 2026 10:00:38 +0000

Read more

Good News in History, March 21

46 years ago today, Pink Floyd’s single, Another Brick in the Wall, hit No.1 on Billboard, beginning an 11-week domination at the top of the charts. From their rock opera album The Wall, written by bassist Roger Waters, it is a protest song against rigid schooling—and features a local elementary school choir. Called (Pt. 2), […] The post Good News in History, March 21 appeared first on Good News Network .

Published: Sat, 21 Mar 2026 08:00:00 +0000

Read more

← back to index