Dr. John Sweller is my instructional design hero.
I’ve given away a couple dozen of his books because cognitive load theory is essential to ID practice. For the first time I’m seeing him in an interview.
The interviewer is an academic with a math focus, but John’s ideas are applicable to everyone who cares about knowledge transfer.
It’s a thrill to share this!
Rudi Garcia:
Drag Ream wide or high then run Lukaku or a wide runner into the channel before Richards can recover. Belgium can break down Richards and attack the second ball.
Poch:
The US loves to press and the crowds encourage it. Emotional maturity required to press selectively, based on coordinated triggers like a bad Belgian first touch, a back pass, or an opportunity to trap on the sideline. The team's midfield shape has to prevent De Bruyne from turning on the ball and spraying it around.
McKennie and Tillman can reduce De Bruyne effectiveness by forcing him to defend. Tillman offers natural deception that I've never seen in a US side before.
Belgium has similar age and pace issues in the back. US needs to run quick switches, Puli isos, late midfield runners.
Match inside the match:
Doku v Freeman, one of the world's best v a breakthrough young talent:
Freeman has to survive without pulling Richards over. The right sided midfielder will have to slide back for help. Put pressure on Doku going the other way.
A heroic game from Freeman and the USA wins. If the game turns into a Jeremy Doku highlight reel, the USA shape will collapse.
This will be a fascinating battle.
A colleague and I were talking about cognitive load theory and how to explain it to SMEs. He described CLT as the best use of a limited mental budget.
I about fell off my stool. It's so simple and so fitting.
Please keep 'em coming. We need more wit and wisdom in these conversations! 😁
We brought in someone from the outside to optimize our video workflows.
I'm stunned at what an expert has been able to accomplish. We plowed through subtitling 310 video clips totaling more than 10 hours of content in less than two hours using scripts that this woman built for us, including the translations! They all tested out with a flaw.
It's amazing to see how fast files can be processed when they are freed from a GUI. We were told that Apple silicon has some superpower that make this possible. That was news to me, but the results were awesome.
It's got me looking for other workflow improvements. What have you done to accelerate your processes?
This is a cautionary tale.
I've posted before about peers that have been ousted by product managers. This is how I am dealing with the same issue (surprise!) this week in my own organization this week.
Like many companies, we get funded for organizational development and other departments allocate funds for our work. That means that product management can go outside for content development or do it themselves to save money.
This week I found out that an ambitious assistant product manager was developing videos using NoteBook LLM. His fully loaded cost is about $100/hour and we're paying for the Google service anyway, so the cost for this activity is minimal and he is enjoying the creative process and the attention he is getting.
How do you respond to a situation like this?
I'll share the denouement of this after a few responses from others.
The attached comment from a college professor is alarming. Are you considering the sorry state of college grad comprehension when you develop elearning?
Multiple focus group interviewees have said that their preferred "learning style" was video, because it's challenging to read a manual.
We've definitely moved towards short video (3-4 min), simplification of slides (4 lines, max), more activities, and conversational experiences.
Are you implementing similar adaptions for the focus challenged GenZ?
Are there any applications online that do credible 3D animation? We use 3D 3-4x per year for mechanism of action videos.
They are often beautiful journeys through body systems and usually expensive.
People ask me every day how they can get a job in ID, especially this time of the year when teachers are looking to bail out of public school systems. Here are two skills that differentiate you:
Taxonomy
Not a glamour skill, but indispensable. Understanding how to properly structure knowledge isn't simple — and it sits at the root of knowledge access and transfer. If AI is going to be used effectively, strong taxonomies are essential.
Agent Skill Development
Claude Skills and their equivalents teach non-humans how to perform a task to acceptable standards. It means sequencing the steps in a process, scaffolding the context, and defining success criteria. It's neither lightweight nor simple.
Both of these skills live in structured plain text. To differentiate yourself right away learn Markdown. It takes an afternoon, it's portable across every tool...but almost every candidate still can't produce a clean markdown file. Pick any Markdown editor — Obsidian, Typora, VS Code — and write everything in it for a week.
The IDs who can structure knowledge and define processes are in demand. They're the reason AI moves from chaos to productivity.
</rant>
Just got back from dinner with a friend, a VP of learning.
He was fired Monday. One of his internal clients, a product manager, complained to the CEO about the quality of the work and said, "I could do better myself." My friend did nothing to push back.
He said, "My big mistake is not having the data to prove the value of our work."
The exact same thing happened three weeks ago to another friend of ours. No proof. No job. It's extremely sad. These are talented people in their late 50s will have a hard time finding new work.
The only reason I am employed well past normal retirement age is because I became devoted to data years ago. When there are gaps in training we identify them quickly and get them fixed.
For every 100 learning messages we deploy, we typically clarify or adjust 10-12 before we go wide.
Data matters.
I paid for dinner.
A member of my family is in the business and doing well but he can always use more work.
Someone just offered him $16,500 for two relatively simple commercials using the "The Advanced Fee Scam."
The way this works (according to Claude):
As Stripe's fraud department describes: "the scammer makes a payment (generally in the thousands) for a project and pays more than they need so you can then transfer money direct to a third party (e.g. a consultant, graphic designer, etc)... the card used is stolen and the third party is the scammer. So, when the legitimate cardholder disputes the payment the business is left holding the bag on the money retrieved from their account and what they transferred to the scammer."
The "private project consultant" is the signature move. In every variant I'm seeing, the consultant is the one supposedly holding the script, logo, and content, and the consultant is who you eventually get asked to wire money to. In one well-documented case targeting a video production company, the scammer actually went all the way through signing an agreement and FedEx'ing a fake $9,400 check from a real company (BPX Energy) signed by a fictitious person, with instructions: "Once the payment of $9400 was delivered, you can proceed to have it deposited for clearance at your bank (this usually takes overnight or maximum 24 hours). you can send the Project consultant her money after taking out your own payment of 6000 from the total amount."
The reason it works is a banking quirk most people don't understand. "In some cases, the victim may even 'cash' the check with a local bank, and the local bank may accept the check for a time. This reinforces the freelancer victim's belief that they have already been paid... The scammer then convinces the victim to send some amount of money 'back' to the scammer. Rather than reversing the payment they have received, the victim is convinced to send or wire money back to the scammer from their own account or funds... The scammer's initial forgery is later discovered, by the individual or by the bank." Provisional credit is not cleared funds. The check can bounce weeks later and your wire to the "consultant" is gone.
The specific offer you provided follows the pattern precisely.
- - - - - - -
The bottom line: If it seems too good to be true it usually is.
