Another AI era?š¦
The tibicos obsession continues
š Hi friend, howās it going?
The tibicos era is getting serious. I got a proper chopping board this week, made mango tibicos and have been making everyone around me try it - friends, my team, no one is safe. Reactions have been good. š¤® I'm taking that as a sign to keep going. š„
Also, for everyone who celebrated - happy Lunar New Year! š§§ Hope you got to rest, eat well and spend time with the people you love.
š» Interesting Things (I think)
A range of things that I find interesting!
šļø Learning
š¤ AI
Building with OpenClaw š¦
Iāve been using OpenClaw and it feels like when AI coding agents first came out. Itās clunky and you could potentially ruin your whole life by leaking all your private data and credit cards, but wow it feels like the beginning of something. š®
Thereās a whole category of things I just couldnāt do before. Automations that required too many moving parts, workflows that needed something to persist across tools, tasks that were too multi-step to reliably hand off. Agentic systems are what make all of that actually possible.
Looking at the architecture (see above) your message comes in, gets normalized, routed, context gets managed and then the model runs in a loop: does this need a tool call? execute. does it need another? execute again. Until itās done. It keeps going until the job is finished rather than waiting for you to prompt it forward.
That loop is the thing that changes everything. Itās the difference between AI that reacts and AI that actually executes. š
š ļø How Coding Has Changed
Anthropic dropped a free Claude Code in Action course and it got me thinking, coding in 2026 just feels different.
Itās less about syntax and more about understanding systems and architecture design. Knowing why things are structured the way they are, how systems talk to each other, what good context looks like. The people getting the most out of AI tools arenāt the ones trying all the new tools, theyāre the ones who understand whatās happening underneath.
The course covers things like context management, MCP servers, hooks and GitHub integration, which reflects what ācodingā looks like now.
Iām thinking about making a video on how Iād learn to code in 2026 from scratch. Would you guys be interested? Reply and let me know š
š Anime: Shinsekai Yori
Been watching Shinsekai Yori and itās set in a future where humans develop psychic powers. Instead of that being a good thing, society builds itself around one central fear: what do you do when your own children might destroy you?
The adults respond the way scared institutions always do. They monitor, suppress, remove the ones who donāt fit. What gets me is how reasonable everyone sounds while doing deeply unsettling things. More power doesnāt mean more stability, it just means more elaborate ways to maintain control. š«
š Not Sure Where to Start With AI?
This is genuinely one of the most common things I hear : I want to learn AI but I donāt know where to begin. I get it, the overwhelm is real. Thereās so much out there.
So our team at Lonely Octopus put together a free 28-Day AI Sprint Roadmap - pick your goal (leveling up at work, building AI products, or offering AI as a service) and it gives you a clear, day-by-day path forward. No more staring at a list of 50 things wondering where to start.
If youāve been putting it off, this is a low-commitment way to just begin. š Get it here for free š
š„ This Weekās Livestream
š„ This Weekās Video
š Join the waitlist for Lonely Octopus if youāre interested in learning how to build AI agents!
P.S. All constructive feedback is greatly appreciated!
I can relate to your fantastic fermenting obsession! I been fermeting every single meals since 2012. Healthwise it has paid off as I been getting stronger, healthier and younger since. Thank you for sharing.
Fermentation as a coping mechanism? Hmmmmm... NAH! I prefer pizza and porn.
"Itās less about syntax and more about understanding systems and architecture design."
We call it "software engineering" where I come from. I lived through the whole "Modular Programming", "Structured Programming", "Data-Drive Programming", "Structured Design" years of the 1980's. Ahem, before your time... And before management decided "time to market" was more important and invented "Agile Programming."
Except it's not really "engineering" - it's more like arts and crafts. Unfortunately. Which is why most software is unusable, unreliable and insecure.
Which, if you've been following the news, is why a recent study shows open source software now has twice as many vulnerabilities than previously - due to vibe coding.
And the entire cybersecurity industry is losing it over this agent fiasco called OpenClaw. There are MANY people calling for improved agent monitoring, control, human-in-the-loop, and vulnerability analysis for deploying these things in corporate environments.
Which is why the current brew-ha-ha over the US DOD threatening Anthropic for refusing to provide its tech for military use - at the exact same time Anthropic has dropped its "safety" principle from its mission statement - is causing people like Gary Marcus to lose it.
See his latest:
America, and probably the world, stands on a precipice.
https://garymarcus.substack.com/p/america-and-probably-the-world-stands
Plus another study from Anthropic (if anything they say can be believed) shows their programmers are forgetting how to code - and worse, forgetting how to debug - which skill is exactly what is needed to fix AI-generated code. It's being called "cognitive debt" as opposed to "legacy debt". Instead of old code no one knows how to fix, it's people forgetting how to analyze bad code and fix it. This is going to have real consequences in the future if not addressed.
Can you say "future software apocalypse"? I knew you could.
By the way, you should do more with your Substack. I know it's time-consuming to write long form. I'm so bad at it I haven't posted anything on my Substack since I can't even remember when. But I intend to. It's an important outreach tool, as Dan Koe will tell you.
Looking forward to your next livestream. Remember my suggestion to have an AI cybersecurity expert weigh in on OpenClaw in a livestream. Before this thing goes completely amuck and destroys the world.
As a final aside, an AI security expert actually had OpenClaw almost delete all her emails. See here:
Top AI Safety Exec LOSES CONTROL Of AI Bot
https://www.youtube.com/watch?v=j0da1ZftUIo
So in the immortal words of the Honorable Henry J. LePetomane, governor of the territory of Kansas, in the movie "Blazing Saddles": "You watch your ass."