From now on, my age will be year-based. I am now 25.
AI Just Made a 3-Minute Movie
Joanna Stern over at The Wall Street Journal has unveiled another mind-blowing development: with the help of WSJ staff, she created a three-minute movie in which nearly every scene, shot, and line of dialogue was generated using AI tools: specifically, Google Veo 2 and 3, ElevenLabs, and Runway.
As impressive as it is, you can tell you’re watching AI. There’s that familiar ultra-smooth feel, some shots of Joanna don’t quite look like her, and action sequences fall short of believability (for example, her karate kick barely connects but still shatters the target like something out of a low-budget CGI fest).
But it’s not hard to imagine how quickly this could evolve toward Hollywood-level production values. Probably the bigger shock is the price (or lack thereof): The entire short cost only about $1,000 to make. I’m no filmmaker, but I assume this is a fraction of what traditional filmmaking calls for.
We’ve come a long way since that bizarre, now-iconic Will Smith spaghetti video. It’s become an understatement to say AI is changing so fast. With these tools now available, what can we expect in another 3 years?
Last night I did a quick vide-coding session in Cursor for an iOS app that does currency exchange conversions. One of the key requirements in the PRD was to store the API key securely. First round of coding put the key in plain text right in the code.
So yeah, human oversight is definitely a thing.
Things are happening
Consider this: the CEO of an influential company being openly and brutally honest to his employees about the effects of AI on their jobs.
Am I referring to Micha Kaufman, CEO of Fiverr? Or maybe another tech company altogether like Shopify or Duolingo? Nope.
Instead it’s… Jim VandeHei, CEO of Axios? A media company?
Indeed, you read that right.
In a recent Axios column, V lays it all out, no sugarcoating. AI isn’t just some productivity boost, it’s a huge game changer that’s forcing Axios to rethink jobs, workflows, and the very future of journalism. VandeHei says that he gets that people are worried about what this means for their careers but says leadership is about facing those tough realities head on.
Sound familiar? It should. That’s because Fiverr’s CEO sent a similar message in an internal email (shared in a tweet by Greg Isenberg). Kaufman talks openly about the challenges AI brings, but also the opportunities, urging his team to adapt and innovate to stay ahead.
What I find particularly interesting here is that AI isn’t just shaking up tech startups, it’s pushing even traditional industries like media to make some tough calls right now.
Things are happening.
Sources:
AI Agent "Tutorials"
I’ve noticed a concerning trend: many “educational” videos and articles seem more focused on riding the AI hype wave than providing actual value. The creator’s understanding often seems surface-level at best. Especially on sites like YouTube, the videos seem more like “I’m making this video to demonstrate my barely-there understanding of this stuff” and less like “here’s a great step-by-step tutorial from someone who’s mastered it”.
After spending hours searching for quality resources, here’s what I’ve learned:
• Look for content creators who demonstrate real applications, not just theory.
• Quality tutorials include code samples and address common errors.
• The best resources explain WHY, not just HOW.
• The videos should conclude with a call to action you can take right now in your current role.
There’s so much buzz out there, people are dying to create content from it. So, here’s some resources I’ve found genuinely helpful:
• Tina Huang has addressed this dichotomy in a number of her videos (Agentic flow vs. actual AI agents)
• The Neuron Newsletter has also talked about this in their daily email.
Sources:
Homebrew with a Twist
Want a super simple and easy way to keep your apps and tooling up-to-date on your Mac?
No surprise here: use 🍺 Homebrew. Install all apps using brew and then use a script to upgrade both brew itself and your brew-installed apps.
However, there’s a certain category of apps that aren’t available via Homebrew. A large swath of these are only available in the Mac App Store. So, what can we do about these?
Luckily, there’s a brew-installable tool that allows you to work around this as well: mas (short for “Mac App Store”). mas allows you to script installing and upgrading these apps just like with brew
Once you install mas, you can now install and update all of your apps and tooling from your favorite terminal (mine is Warp).
Also, you can use shell commands to change the color of the text output to make it easier on the eyes. Below is a script I run occasionally in order to update all of my apps on my Mac and then give me a report of all apps I have installed via brew or mas.
Enjoy!
P.S. Upgrading apps using mas is a little trickier. You have to find the Mac App Store id and then feed that to mas in the terminal. However, I gotchu: the script above will give you all of those ids every time. mas link here: https://github.com/mas-cli/mas
Sam and Satya
Two interesting news morsels coming out of Redmond.
-
Microsoft is re-negotiating their partnership with OpenAI to clear a path for an OpenAI IPO.
-
Microsoft just announced layoffs across LinkedIn(!) and GitHub, with a focus on reducing management layers and streamlining operations. It amounts to about 3% of the workforce.
So, are these two news stories related in any way?
Microsoft has invested north of $13 billion in OpenAI. With OpenAI exploring an IPO, MS likely wants to ensure its continued access to bleeding edge AI technology.
Given that they reported record revenue ($62 billion) and profit in the latest quarter, the Microsoft layoffs appear targeted, because clearly there isn’t financial distress here.
Freeing up headcount and budget allows Microsoft to redirect resources to AI and other growth initiatives.
Revisiting my Vibe Coding Prediction.
About a year ago, I made a prediction: that the rise of LLMs like ChatGPT would lead to fewer programming languages over time. If AI agents can translate high-level instructions into code, then language complexity becomes an implementation detail rather than a developer concern.
Fast forward to today, with tools like Cursor and Windsurf, and we’re closer to that reality. For example, I can ask Cursor to “build a weather app in React with Redux” and get a runnable scaffold without touching the docs. In a matter of seconds.
So far, we haven’t seen any significant drop in the number of languages. But I think the original intuition holds: we’re moving toward a world where the choice of language matters less because the AI abstracts it away. Hence: vibe coding.
