Ed Fenton
banner
edfenton.bsky.social
Ed Fenton
@edfenton.bsky.social
AI executive | Making legacy industries smarter | B2B tech, supply chain, & Salesforce obsessive | Believer in practical AI, not just the hype | Views my own
🔥 Agile hot take #2: Scrum is as bad as SAFe for killing agility
Ceremonies, certifications, and bureaucracy. That’s not agility; it’s sticky notes in a suit.

#Agile #Scrum #Kanban #SAFe
September 9, 2025 at 3:40 PM
🔥 Agile hot take #1: Story points are the worst thing ever invented
They’re just hours in disguise. Endless debates over 5 vs 8 waste time. Drop points. Ship value.

#Agile #Scrum #ProductManagement
September 8, 2025 at 9:40 PM
Open source as foreign policy?
The U.S. and China both released AI action plans. One talks “global standards,” the other wants to own them.
Cooperation or competition? The stakes are rising.
🎧 youtu.be/Xj9C1wlSprw?...
#AI #Geopolitics #China #OpenSource #AIDailyBrief
Is Global AI Cooperation Even Possible?
We explore the escalating AI competition between the US and China following the release of both nations' AI action plans. We analyze China's response to America's AI strategy, including their…
youtu.be
September 4, 2025 at 2:04 PM
Deepfakes are easier than you think. With just 27 images + 2 minutes of audio, I generated a convincing fake of myself.

Inspired by @computerphile.bsky.social (@seanski44.bsky.social, Dr. @mikepound.bsky.social, Lewis Stuart, @uniofnottingham.bsky.social).

🎥 Watch here:
The dangers of deepfake video production (and why you should care)
Deepfake technology is moving fast. With just a handful of photos, a few seconds of audio, and some open-source tools, anyone can generate convincing synthetic videos. This video isn’t a tutorial.…
youtu.be
September 3, 2025 at 2:02 PM
Can you use Google Gemini's Nano Banana for more than just inserting Grimace into historical photos? Yes, but why would you want to?
September 2, 2025 at 2:01 PM
Ada Developers Academy = tuition-free software training, 94% placement.
Ready to hire. Ready to grow.
adadevelopersacademy.org
#WomenInTech #TechCareers #EngineeringTalent
Home - Ada Developers Academy
Ada Developers Academy is a non-profit organization committed to closing the tech skills gap through free programs that teach coding fundamentals and intensive full-stack software development. We…
adadevelopersacademy.org
September 1, 2025 at 9:40 PM
Walmart just skipped to agent orchestration.
4 super agents. 900K users. Billions in time saved.
Agentic retail isn’t a concept... it’s already here.
🎧 youtu.be/CaH9FAVrjUk?...
#AI #Agents #Walmart #RetailTech #AIDailyBrief
Walmart Blasts Past Agent Experimentation
Walmart has evolved from individual AI agents to "agent orchestration" - a unified system of four "super agents" that coordinate specialized sub-agents across their entire operation. This shift…
youtu.be
August 29, 2025 at 4:07 PM
GPT-5 isn’t maxed out by default.
Push it deeper: think harder, plan better, self-improve.
🎥 youtu.be/4U19EQ2DS0A?...
#GPT5 #PromptEngineering #AItools #DeepPrompting
Prompting GPT-5: Deep Prompting Techniques
Stop treating auto mode like a mind reader. GPT-5 rewards deliberate prompts. In this video I push past basics and show six techniques that force deeper reasoning, sharper outputs, and faster…
youtu.be
August 28, 2025 at 2:02 PM
Stop settling for surface-level AI answers.
Role play. Interview. Tree of thought.
Better prompts → better decisions.
🎥 youtu.be/7E-qW2VaYQU?...
#AI #PromptEngineering #GPT5 #AdvancedPrompting
Advanced Prompting Techniques
Basic prompt engineering will only get you so far. If you want to unlock the full power of GPT-5, you need advanced techniques that push your prompts beyond surface-level answers. In this video, I…
youtu.be
August 27, 2025 at 2:02 PM
TACO > lazy prompts 🌮
Task. Actor. Context. Output.
Small effort, big results with GPT-5.
🎥 youtu.be/nIyqeEUyD2U?...
#PromptEngineering #GPT5 #AI #TACOmethod
Prompt Engineering with the TACO Method
We’ve all gotten lazy with prompts. GPT-5 changed the game—its instruction-following makes it insanely steerable if you know how to guide it. In this video, I break down the TACO method (Task, Actor,…
youtu.be
August 26, 2025 at 2:02 PM
@outintech.bsky.social = 40K+ pros building stronger teams and giving back.
Mentor. Network. Contribute.
outintech.com
#TechForGood #LGBTinTech #Mentorship
Out In Tech
Visit the post for more.
outintech.com
August 25, 2025 at 9:40 PM
Finished Co-Intelligence by @emollick.bsky.social. A practical guide to treating AI as a partner, not just a tool. His 4 principles are simple but powerful. Especially “treat AI like a person (but tell it what kind of person it is).” Highly recommend.
a.co/d/2d6ZBDE
August 25, 2025 at 2:00 PM
“We’re reinventing computing for the first time in 60 years.” — Jensen Huang
From twin factories to trillion-dollar AI infra, Nvidia’s CEO sees everyone becoming a creator.
🎧 youtu.be/LaefsjgJkZs?...
#AI #Nvidia #JensenHuang #AIDailyBrief #FutureOfWork
9 AI Predictions from Jensen Huang
Nvidia CEO Jensen Huang lays out 9 bold predictions for the future of AI in a sweeping conversation that moves beyond near-term policy debates to the tectonic shifts reshaping global economics,…
youtu.be
August 22, 2025 at 3:40 PM
GPT-5 nailed my Wordle AI challenge — fast, polished, optimized, and the only one to give me a ZIP download. Grade: A. youtu.be/dazyYVA0s7M?...

#AI #GPT5 #Wordle #LLM
GPT-5’s Wordle AI: I Was Wrong — This Model is a Beast
I’ll admit it — last week’s GPT-5 performance left me unimpressed. This week? Completely different story. The challenge: build a fully playable Wordle clone with an entropy-based AI solver, all from…
youtu.be
August 21, 2025 at 2:04 PM
Claude made the best-looking Wordle AI yet — but slow debugging and high cost kept it to a C+. youtu.be/2JbQSdrjxcU?...

#AI #Claude #Wordle #LLM
Claude’s Wordle AI: Gorgeous UI, Slow Debugging, and a Price Tag to Match
Claude is my daily driver, so I went into this one-shot coding challenge with high expectations. The task: build a fully playable Wordle clone with an entropy-based AI solver — all from a single…
youtu.be
August 20, 2025 at 2:03 PM
Gemini finished my Wordle AI challenge with time to spare, great UI, and few bugs. Solid B. youtu.be/d8gctJ2y5Ik?...

#AI #Gemini #Wordle #LLM
Gemini’s Wordle AI: Fast, Polished, and Surprisingly Good
I went into this four-way AI coding challenge expecting Gemini to be solid — but it surprised me with just how capable it was. The task: build a fully playable Wordle clone with an entropy-based AI…
youtu.be
August 19, 2025 at 2:06 PM
i.c.stars delivers results: 90%+ job placement, 3x income growth.
Skilled devs + future leaders in Chicago & Milwaukee.
Get involved: icstars.org
#TechCareers #ChicagoTech #Mentorship
i.c. stars | Non-profit Tech Training Program
Offering intensive training and a two-year practicum; Through project-based learning, our stars work in teams to solve a range of real-world problems.
icstars.org
August 18, 2025 at 9:41 PM
LLMs can predict orbits but not gravity.
Harvard’s new paper questions if transformers can form true world models.
Are we scaling the wrong thing on the road to AGI?
🎧 youtu.be/eE6zdv15suc?...
#AI #AGI #LLMs #WorldModels #AIDailyBrief
Are World Models the Key to AGI?
A groundbreaking Harvard study trained AI on 10 million solar systems and found it perfectly predicted orbits but completely failed to understand gravity, raising questions about whether LLMs can…
youtu.be
August 15, 2025 at 9:40 PM
@anthropic.com Claude Opus beat OpenAI GPT-5 at my polygon physics challenge. Watch the win: youtu.be/O9Mjt5tPJVQ?...
#AI #ClaudeOpus #VibeCoding
If you enjoy the video, please like and subscribe. It would really help me out.
Claude Opus Nails the Physics Sim GPT-5 Couldn’t Build
After two failed attempts with GPT-5, one barely passable and one a total disaster, I handed the same spinning polygon physics challenge to Claude Opus 4.1. Same environment, same rules, same 45…
youtu.be
August 14, 2025 at 3:10 PM
Gave OpenAI GPT-5 every advantage for a simple physics sim. It still crashed and burned. Watch the fail: youtu.be/dXmop6a6j5k?...
#AI #GPT5 #CodingFails
If you enjoy the video, please like and subscribe. It would really help me out.
GPT-5 Completely Falls Apart Trying to Code a Simple Physics Sim
I gave GPT-5 every advantage this time: React, Matter.js, Web Audio API, Canvas API, all the docs loaded, clear rules, explicit step-by-step guidance. It still couldn’t pull off a basic polygon ball…
youtu.be
August 13, 2025 at 4:06 PM
Told OpenAI GPT-5 to build a polygon physics sim in 30 minutes. It finished… sort of. Watch the chaos: youtu.be/JODdAYarjRE?...
#AI #GPT5 #VibeCoding
If you enjoy the video, please like and subscribe. It would really help me out.
I Asked GPT-5 to Code a Physics Simulation and This Happened
I challenged GPT-5 to code my go-to developer test: a rotating polygon with a ball bouncing inside under real physics. No special rules, no extra prompts, no tools. Just GPT-5, a blank project, and…
youtu.be
August 12, 2025 at 3:40 PM
We took 60,000 employee engagement comments and built a full modular AI pipeline to:
- Import + clean
- Analyze sentiment
- Flag AI-solvable issues
- Deduplicate with embeddings
- Enrich + export ideas to CSV

Watch “Engagement Insights: Enrich and Export” (15 min):
youtu.be/PRMuGaVHS5U?...
Engagement Insights: Enrich and Export
Optimizing AI and Automation: Handling Employee Engagement Comments with Cursor In this episode, we take you through the final steps of our project to generate AI and automation ideas from 60,000…
youtu.be
August 9, 2025 at 12:45 AM
🔥🔥🔥 Full pipeline debug. 🔥🔥🔥

My AI-solvable feedback engine hit a wall: sentiment analysis skipped 300+ rows, solvability checks ran on the wrong data, clustering was misfiring.

Watch “Engagement Insights: Debugging” (7 min):
youtu.be/RBuk3UD2NQE?...
Engagement Insights: Debugging
Debugging and Optimizing Our Employee Engagement Pipeline In this episode, we dive into a detailed debugging session to fix issues in our pipeline that processes 60,000 employee engagement survey…
youtu.be
August 7, 2025 at 3:08 PM