In this week’s newsletter, I share a rough “Theory of Literacies” that has formed in my mind after getting many career questions from my AI workshop participants. I also recap what I wrote about Claude Dashboard, Nano Banana 2.1, reducing sloppiness in AI imaging, a16z’s latest lists of top consumer AI apps, unnecessary anti-AI hate, poor sales pitches, and a Dummies’ Guide to USB-C cables.
Weekly Recap
Claude’s new Dashboard artifact feature is very, very good. I used it to analyse Singapore’s Q2 2026 retrenchment figures.
Google also launched Nano Banana 2.1. I tested the imaging engine: It’s great but I still had to do a lot of manual editing to reduce its verbosity.
Here’s how to iterate with AI images to reduce the “sloppiness”.
Andreessen Horowitz (a16z) has published the latest lists of top consumer AI apps. This time, there’s an interesting new list.
A person vibe-coded a useful haze-monitoring app, only to face toxic anti-AI people on Reddit.
Use this article from Andrew Ng if someone picks a fight with you about the impact of AI datacentres.
Older people often react differently to AI because we lived through the age of crappy analogue tech.
A wonderful quote from Robert Updegraff about complaining.
Why you shouldn’t send a poorly-written (probably AI-generated) insurance pitch to me at 815am in the morning: I will actually reply.
My Dummies’ Guide to USB-C Cables. You just need two types of cables in your bag.
No AI was used in this photo. This image may confuse anyone who assumes that only AI or Photoshopping can create such visuals. On the other hand, this image used both AI and Canva photo-editing.
Sunday Essay
Recently, a lady in her early 60s asked me during my AI workshop: “How do I pivot my career?” She had held many different roles in her life, and her current job was a contract role where she trained AI models by giving feedback on their output. Essentially, she was teaching AI how to replace other workers. I told her she needed to get into a job where she provides value to other people, not the machine.
The more I teach people how to apply AI to their work while retaining human agency, the more I realise how many non-AI things they need to know beforehand. At this point, I see that there are several literacies that each worker needs to have, and I have come up with a rough Theory of Literacies, or a Literacy Stack:
AI Literacy – How AI works and how to manage it.
Work Literacy – Understanding your economic value and why you get paid.
Media Literacy – How to consume content critically.
Cybersecurity Literacy – How to keep data safe.
Data Literacy – How to handle and transform data sets.
Tech Literacy – How hardware and software works.
At this time, I think Tech Literacy is the most basic part of the Literacy Stack. It is as simple as knowing how to use Wi-Fi, and how to key in a URL into a browser. Within Tech Literacy, you can go much further, such as knowing how to change the settings on your computer, and how to assemble a computer. Yep, each literacy has different levels of proficiency.
Then, as we go up to Data Literacy, people have to deal with data, and learn how to organise and share it. Most workers know how to do the basic stuff such as saving and sending files. However, they are often not trained in advanced data skills such as creating pivot tables, merging datasets, comparing data and so on.
When we come to Cybersecurity Literacy, Media Literacy and Work Literacy, these are the areas where I’ve observed the biggest gaps.
For example, if a person does not know how to recognise scam methods (Media Literacy), and doesn’t know how to secure his bank account data (Cybersecurity Literacy), then he has a high chance of getting scammed.
If a person pursues his passion but only pursues jobs in sunset industries (Work Literacy), then he will struggle to be employed. The ignorance could also be due to poor Media Literacy where he has not gathered enough information through different communication platforms to assess his work prospects.
Then enter AI Literacy, which is a shock to everyone because it demands proficiency in every other Literacy; otherwise AI’s impact becomes difficult to grasp.
Here’s what happens when a person deficient in each Literacy encounters AI:
Tech Literacy: “Which AI app to use? What is cloud AI vs local AI? Why must I pay for AI? You mean my phone photos are processed with AI?”
Data Literacy: “Why does AI not give me fixed answers? Why is my chatbot not working properly? Why is AI unable to digest my massive Excel file or PDF?”
Cybersecurity Literacy: “Why is vibe coding risky? Why is my AI-generated app exposing our customer data? Why does Copilot have so many guardrails?”
Work Literacy: “Can AI actually replace my job? What kind of jobs are difficult to replace? Do we need Universal Basic Income?”
I’ve been conducting Gen AI training since early 2023, and back then, Gen AI was more of a technological curiosity and toy. Today, it is an existential threat on multiple levels to many people, and I believe it is because it exposes an individual’s lack of knowledge and proficiency in the multiple literacies.
AI Literacy is the point where all our other Literacies come together.
The more I teach AI, the more philosophical I get about its nature and its impact, and the more I think about each person’s stack of Literacies. Unfortunately, most people are too busy with their daily work and life responsibilities to see what’s going on, and then when the shock happens (eg. job loss, industry collapse, paradigm change), they may lack the mental frameworks to make sense of things.
What can you do about it? You’re probably not going to like the naggy answer: Go find out what you don’t know. Read new things, apply the knowledge and practice lifelong learning. Don’t waste your time consuming mindless content, it’s time to create your future with knowledge.