♻️ The Transposed Organization ()
What if you transpose your operational flow and use AI and automation to enable people to own the full product loop rather than one piece in a chain?

🤔 Controversial Career Advice ()
This is not your typical career advice.

🐮 5 Types of BS Job ()
AI will come for the BS first. Here are five kinds of jobs most at risk. Taskmasters beware.

⚒️ Against the Luddites ()
Even Marx and Engels saw through the elitist Luddite movement.

⛓️‍💥 Disappearing Childhood Freedom ()
The supervision era has ended the freedom to roam the neighborhood.

⚾ Why I Love Baseball ()
I am not a baseball fan, but there are reasons to love baseball.

🥛 Soy Right Ascendant ()
The desperate need for approval and respect combined with a refusal of agency and responsibility is a special combination.

🌅 Was There a House In New Orleans? ()
How the Animals started singing about a fictional house in NOLA.

The Look

has dropped the Consumer Electronics Design Trends 2026 report. Electronics are twisting into different forms and hiding in plain sight.

Subscribe now

Did you enjoy reading this post? Hit the ♥ button above or below because it helps more people discover great Substacks like this one and it helps train your algorithm to get you more posts you like. Please share here or in your networks to help us grow!

🦞 What is OpenClaw? (More or Less Podcast)
OpenClaw has been the Main Character in the AI world this week. It's an AI agent that runs on your computer and it's reshaping who controls the AI stack.

🏃‍♂️ The Great Claw Forward ()
While we debate security, China is mass-mobilizing.

💻 How to Use Perplexity Computer ()
Perplexity Computer is the new agentic tool that comes with your subscription. It is not a search engine but your own agentic worker.

🪖 The Art of War in the 21st Century ()
Warfare is different now. Electrically-powered, AI-enabled drones make asymmetric, omnipresent warfare a reality.

🦅 The Eagle and The Lions ()
America and Israel wanted a sprint, Iran is getting the marathon it wants.

🛢️Funding The Enemy ()
Waiving sanctions on Iranian oil is just rewarding Iran for closing the Straits.

🏝️ The Kharg Island Fantasy ()
Can you invade an island in the Persian Gulf without any ships in the Persian Gulf?

🎈 Don’t Buy the Helium Hype ()
Helium production has been curtailed, but how much does it matter?

🌍 Modeling the Messy World ()
What if you could tokenize and understand the real world like current LLMs work with language? World Models learn by seeing rather than reading and they are going to get very interesting soon.

🥊 Fight the Fight 3 Times ()
Figuring out when you should give up is hard.

🤓 Storytelling Mistakes by Technical Leaders ()
Technical leaders tend to give too much detail, overthink tactics, overshare back story and go on too long to be effective storytellers. Here are some ways to work on those tendencies.

🗺️ Non-Technical User’s Guide to Claude Code ()
In case you haven’t taken the plunge, this is a handy guide to getting going with Claude Code for the not-so-technically inclined.

💿 Cop Killer ()
The song Cop Killer struck a nerve in 1992. The band still rocks on.

The Look

Today’s reminder that oil and gas are in everything.

Subscribe now

Did you enjoy reading this post? Hit the ♥ button above or below because it helps more people discover great Substacks like this one and it helps train your algorithm to get you more posts you like. Please share here or in your networks to help us grow!

🔮 Amy Webb (FTSG)
Amy Webb accepts the creative destruction of the Trend Report. Her Convergence Outlook sees a world with machines at the wheel and all-seeing eyes.

🏭 Dylan Patel on AI Compute Bottlenecks ()
There are three physical chokepoints in compute: logic, memory and power. They will determine how fast we can scale.

📱 The iPhone vs the Bank Teller ()
The ATM did not reduce bank teller jobs. It only replaced a single task. The iPhone replaced many other reasons you would need to visit a bank, and the tellers disappeared as branches closed.

🚲 Drones like Bicycles (Phenomenal World)
Iran’s Shahed-136 drone has changed warfare. The cheap cruise missile has changed the air defense equation, often being far cheaper than the interceptor vehicles sent to take it down. How does it come together? What does it cost?

🙋‍♂️ How to Start a One Person Business in 2026 ()
How you can build a $1m, one-person business today.

🔨 Atoms Are Cheap, Process is Pricey ()
How SpaceX rethinks from first principles and implements process to make amazing launch vehicles.

🪜 AI Wrecks the Corporate Career Ladder ()
Tools like Claude Code reduce the need for assistance with execution, breaking the breadth-first model many people have trained in. Taking a depth-first approach where juniors are given whole projects with AI assistants scales better in the new world.

🤖 How Can A Regular Person use an AI Agent ()
AI agents are a lot simpler than you think.

🎹 The Piano Bar is Musical Truth ()
Today I learned everyone requests “Pink Pony Club” at the piano bar when they are not singing “Teenage Dirtbag.”

📻 The Iconic Boom Box ()
Boom boxes were a cultural icon, delivering hip hop and the mixtape to America.

The Look

points out how much water farmland uses compared to ChatGPT:

Subscribe now

Did you enjoy reading this post? Hit the ♥ button above or below because it helps more people discover great Substacks like this one and it helps train your algorithm to get you more posts you like. Please share here or in your networks to help us grow!

🙈 The Invisible Information War ()
Information warfare might be the most important kind of warfare. Here are six observations on how it is playing out in Iran today.

🌍 Is the Intelligence Monopoly Over? ()
You can reconstruct ongoing military operations from public sources.

🏖️ Special Update from Abu Dhabi ()
Stephen McBride lives in Abu Dhabi. He gives you a sense of what life is like at one facet of the front. He would remind you that drone warfare is important.

🔥The Ultimate LNG Report ()
Everything you need to know about liquid natural gas in the Gulf.

🤖 Why You Only Hear About AI Doom (Sahil Bloom)
It is far easier to monetize the negative doom loop than to sell positive stories.

📰 AI Will Reshape Public Opinion ()
Social media democratized public opinion, fueling the rise of populists. LLMs might be capable of pulling things in the opposite direction should they overcome their current shortcomings.

💿 Rick Rubin is the Future of Work ()
Rick Rubin has taste and can express what he feels. He does not know how to play an instrument or work a sound board. In a world of AI you do not need to know how.

💊 The Peptide Wave ()
Peptides are the drug of 2025. Here is a great rundown on a few popular models and the underlying science.

🪖 The 100 Year War ()
We have been fighting the war on drugs for a lot longer than we remember.

☀️ 64% of Americans Want to Stop Changing Clocks ()
We all just want to stop changing clocks. We don’t care which clock wins. Just pick one.

📚 Gen Z Lives in the Archive ()
Gen Z is actually reacting to everything that ever happened in quasi-real time.

🪨 A History of Rock, Paper, Scissors ()
Using hand shapes to settle disputes is a very old human tradition.

The Look

Beware of fake satellite images, via Flowing Data.

image

Subscribe now

Did you enjoy reading this post? Hit the ♥ button above or below because it helps more people discover great Substacks like this one and it helps train your algorithm to get you more posts you like. Please share here or in your networks to help us grow!

🥳 The 2028 Global Intelligence Boom ()
An alternative vision of 2028. What if the jump in agentic capabilities created an abundance boom rather than a job destroying economic death spiral?

📓 Citrini is Great and Deeply Flawed ()
A detailed examination of the essay and a multi-sourced look into all of its flaws.

🌍 10 Charts that Capture How the World Is Changing ()
Another great collection of graphs showing how the world is changing underneath our feet.

🔍 Large Scale Deanonymization with LLMs ()
Anthropic might have been right to be concerned. It might be easy to identify people based on large-scale anonymous data.

🎢 Hold on to Your Hardware (mrusme)
The best time to upgrade your hardware was yesterday. The next best time is today.

🥸 Teamwork is a Scam ()
There are certain advantages to solopreneurship in the modern era.

🧑‍🔬 The Engineerification of Everything ()
Engineering has escaped the lab and permeates every function today.

👮 Crime is Why America Feels Worse ()
Crime and public disorder make the US unlike every other rich country.

🚋 Why the DC Streetcar Deserved to Die ()
A third of a streetcar line does not create a viable transit option.

🎁 The Spring Product List ()
A great list of products to satisfy your spring retail therapy requirements.

The Look

NVIDIA shows how it is very profitable via .

Subscribe now

Did you enjoy reading this post? Hit the ♥ button above or below because it helps more people discover great Substacks like this one and it helps train your algorithm to get you more posts you like. Please share here or in your networks to help us grow!

Timeline

If you have not been tracking the conflict, here is a quick refresher on some meaningful events that materially shape the current engagement.

• 2017-2018: the Trump administration first renews sanctions then withdraws from the JCPOA nuclear weapons deal. Maximum pressure initiated.

• 2019: conflict heats up with Iran’s Islamic Revolutionary Guard Corps (IRGC). Tanker attacks in the Straits of Hormuz lead to a visit from a US Navy task force.

• 2020: A US drone attack kills IRGC leader Qasem Soleimani in response for strikes on US bases in Iraq.

• 2023-4: Iran backed Houthi rebels attack Red Sea shipping. US and UK forces launch strikes on Houthi missile and drone infrastructure, culminating in large-scale strike on 1,000 targets in March 2025.

• September 2024: Israel decapitates Hezbollah in a daring strike using booby trapped pagers distributed through a years-long supply chain attack. Hezbollah is rendered ineffective as a fighting force to this day.

• June 2025: Israel attacks Iran in the Twelve Day War. The US jumps in near the end of the fight, attacking Iranian air defense and nuclear capabilities directly. Proved US and Israeli assets were capable of penetrating Iranian airspace at will.

• January 2026: Nationwide protests erupt and are bloodily repressed. Thousands killed, many more detained during a brutal repression led by the IRGC. This wave of protests was much larger scale and more generalized than previous waves.

By February 2026, Russia is fairly consumed with the ongoing Ukraine War, unable to send material help. Iran is on her own, proxies like Hezbollah are diminished, rebuilding her air defenses and nuclear program. If there was a time to strike, it was now.

The Territory and the Lead-Up

Here are some curated background sources that should help make sense of the headlines.

🗺️ Land war is unrealistic. See ’s exploration of Iran to get a sense of how daunting the terrain is.

🚢 Can Iran close the Strait of Hormuz, a narrow waterway carrying a fifth of the world’s oil trade? Keeping it closed would be harder than it looks according to the .

🔮 For a wider view of this conflict, explained the military and political situation effectively. Conflict was inevitable, there was no offramp between Trump and how boxed in Iran had become.

Staying Updated

For a raw but detailed and digestible feed keep an eye on the Institute for the Study of War’s Middle East desk for daily updates. This is the best way to keep up with the conflict on a detailed level if that is your goal.

The Bismarck Cables, mentioned above, are a succinct way to keep abreast of international affairs of all sorts.

From an investment point of view, the most important question is “what is happening in the strait?” Real-time shipping information is a good place to start. Check out this map on Marine Traffic to start. Traffic is very light going through the Strait at this time. The thing to look at is if we see significant mining or other activity and maritime insurance rates effectively close the channel. It is probably too late to buy Brent Crude Oil futures.

Let’s hope for a fast and peaceful resolution to the fighting. Hopefully involving the end of a horribly repressive regime.

🌐 Tariff Ruling I: What Changes? ()
ChinaTalk sat down with scholar to help understand the effects of the ruling on trade and ongoing negotiations.

⚖️ Tariff Ruling II: Supreme Court Dynamics ()
What did this ruling say about the court and the dynamics between the justices? If you like legal drama, don’t miss the following deep dive.

👎 DOGE’s Final Failure ()
Saying you are taking a chainsaw to government is one thing, effectively wielding it in America’s multistakeholder system is another.

🏛️ How a Congressional Office Actually Works ()
Laws are made by people working in small offices with funny rules. Here is how they operate.

📉 Wall Street’s Memo from The Future ()
Investment research firm games out what happens when intelligence becomes a commodity and it breaks our economic assumptions.

📉 Don’t Short SaaS ()
The market is pricing in extreme AI disruption of the software business. Should it?

👨‍💻 Thin is In (Stratechery)
Thick clients won the last war, but thin clients are winning this one.

💻 Is the Future “AWS for Everything?" ()
AWS is flexible automation at scale. Will AI create conditions allowing that dynamic to appear in other businesses?

🤷 How Will OpenAI Compete (Benedict Evans)
OpenAI lacks any durable strategic advantage other than a large, largely no-paying customer base. Can they figure out how to navigate that?

🔨 Effective Claude ()
Claude is amazing but intimidating — it is really five big, hairy, complicated tools in one. Here is a great guide to get started.

🔌The Wonderful World of AI Plugins ()
AI is more than just your chat window, here is some advice on how to pick integrations.

🍸 How to Do Cocktails for 100 ()
Throwing a cocktail party for 100 in a DC row house is not logistically easy.

🗿 Concrete Tech (Wallpaper*)
Concrete electronics to match your brutalist style.

The Look

Steak prices are almost worse than memory prices.

Subscribe now

Did you enjoy reading this post? Hit the ♥ button above or below because it helps more people discover great Substacks like this one and it helps train your algorithm to get you more posts you like. Please share here or in your networks to help us grow!

🏪 Brand Lessons of the Superbowl ()
The Superbowl is more a marketing event than a sporting event.

💪 The Ascension of Clavicular ()
Looksmaxxing is so popular young Los Angelinos are sharing peptide dealers.

🏰 NVIDIA’s Moat is Cracking ()
The economics of mass inference are poking holes in CUDA’s armor.

🤕 AI’s Pandemic Moment ()
The vibes of AI seem a bit off, take 1: the world stares into the abyss.

🔥 Something Big is Happening (Matt Shumer)
The vibes of AI seem a bit off, take 2: the agents have arrived.

🕵️ The Agents Are Here ()
Real improvement in real time is hard to notice.

🍕 The Two Slice Team (Every)
In the ‘00s, you organized around 10-person, two pizza-eating teams. In the 2020s you run one person, two slice-eating teams with AI.

🧑‍⚖️ AI Will Not Make Your Lawyer Cheaper ()
Why AI will not radically reduce legal bills in the immediate term. The regulations will not bear it.

👏 How to Move Any Room ()
Straightforward frameworks for getting your message across.

💼 Favorite AI Tool Use Cases ()
How one leader fits Perplexity, Gemini, NotebookLM, ChatGPT and Claude together for a successful workflow.

🦔 A Brief History of SEGA ()
There was a lot more to SEGA than the Genesis and Sonic the Hedgehog.

The Look

The EPL won via .

Subscribe now

Did you enjoy reading this post? Hit the ♥ button above or below because it helps more people discover great Substacks like this one and it helps train your algorithm to get you more posts you like. Please share here or in your networks to help us grow!

In fact, there is a bit too much to work with on the subject. I was wondering if you could help me focus. See the poll below, if you have more input let me know in the comments.

And now for our regularly scheduled programming.

♻️ The Disposable Software Era ()
We are moving into an era where some software is disposable and tinkerers are enabled.

🔥 Software: Armageddon or Opportunity? ()
Software has been eating the world for the last decade and a half. Is it it’s turn to be eaten?

🤖 The Dawn of the Posthuman Age ()
What to make of a high technology, low fertility future.

🏎️ How F1 Conquered American Culture ()
How did F1 go from peripheral to important enough for Cadillac to start a team and take out a Superbowl ad to tell the world about it?

😎 Immersive Sports Revolution ()
Volumetric video, immersive entertainment venues and next generation headsets are conspiring to revolutionize the way you can consume sports.

👮 Fraud Investigations Follow Predictable Patterns (Bits About Money)
Fraud has become unfortunately politicized. In fact, it behaves like many other professions with identifiable supply chains and apprenticeship systems.

📈 Claude Coding a Portfolio Tracker ()
You could pay a lot for software you can just vibe code now.

🥈 Silver Supply Crisis ()
Part of the problem with Silver is we are not mining enough to meet demand, creating conditions for wild price swings.

🏭 Building the Wrong Factories for War ()
War needs mass production of shells, not highly differentiated production of consumer goods. America specializes in the latter.

🪦 Beware Q Day ()
Q day is the day quantum computing renders current encryption ineffective. Here is how nations are preparing for that eventuality.

🏫 School is Worse than Social Media ()
The studies about social media being bad for kids failed to adjust for schooling.

🏅 Amazing Olympics Coverage (Sports! by Rodger Sherman)
Rodger Sherman does brilliant coverage of esoteric sports. His Winter Olympic coverage shines.

The Look

China installs more robots than the rest of the world put together.

Subscribe now

Did you enjoy reading this post? Hit the ♥ button above or below because it helps more people discover great Substacks like this one and it helps train your algorithm to get you more posts you like. Please share here or in your networks to help us grow!

📉 Painfully Long on Silver ()
The story of one trader’s crash and why he is staying long in Silver.

💾 Opportunity in Software? ()
Lots of great graphs to study and one interesting question — software stocks are beat up, are they going to get onto the front foot again?

🤩 Why Open Source AI is Raising Serious Dollars ()
Open source is a way to win serious mind share in the AI world.

🤷‍♀️ Does Anyone Believe Anything? ()
Who would think tribal politics would lead to hypocritical beliefs?

🧊 Operators, Cosplay and ICE ()
Living the cult of the operator.

🪙 Inside OpenAI’s Unit Economics ()
Are AI models profitable? Can AI models become profitable?

💔 I Broke Up with ChatGPT ()
I reach for ChatGPT less and less these days. Apparently I am not alone.

🤖 When AI Agents Get Real ()
ICYMI, this publication’s take on the new wave of truly agentic tools that will run on your documents folder. Stay tuned for more on this subject.

👺 Backseat Software (Mike Swanson)
The pressure is on product managers to measure everything. This leads to push notifications, not software that gets out of the way.

😍 Ideas Have Sex ()
Ideas need to mate with other ideas to prosper and grow.

🏈 Fox’s NFL Deal Changed Everything ()
Rupert put Fox on the map with his NFL deal in 1993. It changed football broadcasting and funded Fox News among other feats.

⚽ Set Piece Revolution ()
What do the numbers say about set pieces today?

The Look

Measles are back!

Subscribe now

Did you enjoy reading this post? Hit the ♥ button above or below because it helps more people discover great Substacks like this one and it helps train your algorithm to get you more posts you like. Please share here or in your networks to help us grow!

Most of the world lives out of their desktop and their documents folder. It is where you find the photos and the notes. If AI agents are going to revolutionize your work, they need to land in your home directory. Patching agents into web applications via OAuth workflows is tricky and confusing for users, and requires developer time. Accessing the same data through the browser already logged into the website is much cleaner. Your ~/home is where worlds collide, and if AI Agents are to be successful they need to be able to work there, not just confined to a chat window with a few MCP tools.

This briefing focuses on a few tools you should be aware of that are changing the way we work. Claude Code is the granddaddy, a venerable 11 months old now. It has matured and made it out of the command line. Its cousin Claude Cowork is like Claude Code for the normies. Finally, Anthropic put Claude in Chrome and Excel, bringing the power of Opus 4.5 to just about everything important in the information world.

Counterbalancing this is the tool that has taken the computer geek world by storm over the last few weeks — a local‑first, persistent orchestration layer designed to run agents on your own machine rather than inside a hosted chat product. It was first called Clawd but that was a little too close for Anthropic’s comfort, so the tool has molted and become OpenClaw. It is not the first agentic AI tool you can run on your home computer, but it is so well done and adoption has taken off so fast that you can’t find a Mac Mini in Silicon Valley Best Buys because they are all sitting on shelves running agents.

It is an amazing time to be a technologist.

Claude Code Comes of Age

Claude Code has been an amazing product since its inception just last February. From the get-go, the power of having an agent running on your own PC with direct access to resources was a clear winner. Originally it was a command line tool and, as the name implies, aimed at coders. But soon many people figured out that it was a tool for much more than that. If you could throw the tokens, or a Max account, at the problem, Claude Code would work on it.

One key feature of Claude Code is skills. They are effectively small programs, written as a collection of text files, that Claude can use to execute. That, combined with MCP tools and the aforementioned file system access, made it a powerful tool. It was effectively the agent most people had. Buzz was growing.

Over the holidays it started to break out of its shell — Anthropic opened it up to more users. People figured out the command line was not too scary. Then they pounced and both enabled the feature globally for paid accounts and added a Claude Code client to the desktop app.

If you have a paid Claude account you should try the feature today. There is a ton you can do with it even if you never, ever will use it to write code.

Claude Cowork

All of this power matters less if it can’t meet people where they already work. For most users, that place is not a terminal window but a graphical desktop full of folders, documents, and familiar apps. GUI‑native agents are what finally bring the home‑directory thesis to life.

Claude Code is an amazing tool, but most people don’t work in the command line or with command line tools. They write documents in GUI word processors. This is where Claude Cowork steps in. It is a beta feature for paid Claude accounts that will take over your macOS desktop to do things like reorganize folders and write documents and presentations.

To Anthropic’s credit, they built it on a lightweight VM to keep the AI constricted to the data you share with it, rather than running in your context with all of your rights. It is scary — but not as scary as it could be — from a security perspective.

Claude in Your Apps

It would not be an overstatement to say the world is run in Excel. Anthropic has brought their platform to the place where the world counts by building Claude in Excel. It can now peer directly into your most important financial data and give you sound advice. It can clean up formulas, data and build graphs. You can use it to build complex models according to some users.

The parts of the world not run in Excel are run in browsers. Claude in Chrome brings it to the browser tabs you want. I’ll admit I use the tool to edit posts — it can interact directly with the Substack editing tool, beating the copy and paste or hunt and peck workflows I’ve used in the past. Running within your browser means you can give it access to anything you want. Your whole inbox if you are brave, or maybe a specific email if you are not so brave through a simple tab group sharing mechanism.

Your Own, Private AI has a Social Network

Ever since ChatGPT appeared, many technologists have dreamed of having a smart, persistent, context-aware private AI agent ready to do their bidding. Many tools have checked parts of that box, but none effectively packaged everything needed to run a truly local, orchestrated agent.

Peter Steinberger effectively solved this problem when he created what was once called clawdbot. The tool creates an interface over a local AI manager. You supply an API key and the tool supplies the connectivity and interface for a self-hosted persistent agent. The tool connects to any command line application — it just uses AI to understand them. It has infinite plugins out of the box thanks to the entire ecosystem of python and nodejs. It can be used as a personal agent, but it is really designed as an orchestration tool. I can see a world where you run multiple instances for discrete tasks and compartmentalization.

Superpowers aside, it excels as a personal agent. It remembers you, your likes. You can text it like your bestie. Like Claude it uses skills to perform tasks like managing your inbox or even booking travel. It has access to whatever you give it.

If this sounds cool to you, you are not alone. It went viral last weekend, with the Github project getting 9,000 stars on January 26 as Mac Minis sold out in the Bay Area. The newfound popularity caused a bit of a trademark issue and the name molted to OpenClaw as of January 30.

There is a lot of talk about security issues. It has the same tradeoffs any agentic AI system has — to make it useful you need to give it access to data and often the environment the data lives in. Like any other automated tool, limiting access is worth thinking about. If you use Clawdbot, think about giving it a dedicated machine, Gmail, iCloud and other accounts and only tapping into personal accounts as absolutely necessary. The design of the system is impossible to secure, but you can practice defense in depth to minimize damage.

What This Means

OpenClaw is not for everyone — it certainly requires knowing your way around the command line, being comfortable with API keys and doing a bit of troubleshooting. It is definitely not the kind of thing you should regularly run on your own PC in your own account. But give it a machine and some constraints and you can have a private setup on par with Claude Code at similar unit costs. Moreover you can deploy a fleet of them dedicated to specific tasks and separate security containers.

I did not set out to write this much of a Claude love story. But it is hard to hide — Anthropic is rapidly building a differentiated platform. Their tools are doing much more meaningful work for me than OpenAI’s ever have, even if you still cannot draw an image directly. The mileage you can get today out of their tools for most work is worth the investment or even switching providers. OpenAI probably is not in a great place here compared to their peers.

The tools are starting to think deep enough to do real, meaningful work. They are getting directly connected to the knowledge systems. It is a good time to start thinking about how you are going to work with them rather than for them. With agents embedded where work lives and orchestration running locally, 2026 is shaping up to be the year AI stops being a dream and starts doing real work on your machine.

Resources

Here are some things I found inspirational in writing this piece.

Claude

The Magic of Claude Code (, Alephic)
Why claude code is magical. Still very true four months after publication, which is roughly a century in AI years.

Claude Code and What Comes Next ()
How AI is capable of repeated, sustained work.

36 Claude Skills Examples ()
36 skills from 36 creators to show you want you can do with skills.

How to Use Claude Code for Everyday Tasks (Every)
Claude code can do a lot more than just code.

Will Claude Eat Excel ()
What you can do with Claude in Excel.

Claude Cowork: What It Is, Why It Matters and How To Use It ()
Excellent explanation piece about Claude Cowork

OpenClaw / Molt / Clawd

The Lobster Takeover (Starry Hope)
Good guide to setting up your OpenClaude including some hardware options beyond Mac Minis.

Interview with the Creator of Clawdbot ()
Peter Steinberger, and clawdbot, has a really interesting story.

OpenClaw is Wildly Insecure ()
OpenClaw is far from hardened, here are some factors to be aware of when you setup your instance.

The Best of Moltbook ()
AI agent social interactions are amazing to behold.

Subscribe now

Did you enjoy reading this post? Hit the ♥ button above or below because it helps more people discover great Substacks like this one and it helps train your algorithm to get you more posts you like. Please share here or in your networks to help us grow!

🚞 AI Device Watch in Davos ()
AI bubbles and AI devices were top of mind in Davos this week.

👍Modern Life is Good ()
Chris literally walks the word and has the perspective to see just how good modern life actually is.

🔥 Learning from the History of Fire ()
Deadly fires were once a frequent urban phenomenon. It required a variety of advances to create the relative fire safety we enjoy today.

🚫 Iran’s Internet Shutdown (Rest of World)
Starlink was thought to be un-jammable, then Iran broke GPS. What are the implications?

❌ The Social Media Ban Wagon ()
Banning social media for minors is the popular thing to get behind. Is it the right idea?

😡 Understanding Ragebait (Ragecheck)
Social media posts are designed to anger you. Use this tool to learn how they operate.

✍️ Content Operational Skills ()
Writers are experts in moving content through a system, a critical skill for successful AI integration even if the LLM takes the pen.

🦞 Clawdbot Moment (More or Less Podcast)
Clawdbot is the hottest thing among the AI geek set. Watch this to understand what it is about. It is both sweet and a security nightmare.

🦸 Unlock NotebookLM Superpowers ()
NotebookLM is an amazingly powerful tool. Here are six techniques to get the most out of it.

The Look

The guardrails have changed.

Subscribe now

Did you enjoy reading this post? Hit the ♥ button above or below because it helps more people discover great Substacks like this one and it helps train your algorithm to get you more posts you like. Please share here or in your networks to help us grow!

Big Picture

🤔 Who Captures the Value in AI? ()
AI unit economics are rapidly improving. Are they improving rapidly enough for Anthropic and OpenAI’s survival?

🏪 LLMs vs Marketplaces ()
Which businesses and marketplaces are more vulnerable to disruption by LLM? What should you do if your business is in the crosshairs?

🏘️ The American Dream Needs a Factory Reset ()
Building more housing might require building homes in factories. Here are the disruptors trying to make that work.

💩 Enshittification Incentives ()
How the incentives and obsession with measurement in product organizations lead to products that are hostile to users.

Technology Today

🏁 Talladega Bytes ()
Differentiating LLMs and Big Model Companies through car manufacturer analogies.

👎 AI Ruined My Sunday ()
Nick surrendered his Sunday to AI. It did not go well.

📈 The Vibe Graph ()
If you feed AI enough context, it can understand the vibe of decisions. It turns out emoji velocity can tell you things.

📝 The Magic of Markdown (Anil Dash)
Markdown is one of those Web 1.0 creations that stuck.

Work and Life

🎀 What is in for 2026 ()
What is in for 2026 according to the TikTok Girlie class. Are you ready for whimsy mixed metals?

🎬 Fan Edit Obsession (Inside Hook)
Gen Z is obsessed with creating mini trailers set to music. They are as amazing as that sounds.

🚀 How to Spot a Breakout Startup ()
Joining a startup is a major commitment. Here are some tips on spotting a winner.

🧑‍🤝‍🧑 Work Friends make Culture ()
Workplace friendships are a 20% salary premium and otherwise critical to retention.

🎙️1996 Was a Hell of a Year ()
I graduated high school in 1996, we had a hell of a soundtrack. Street dreams are made of this.

The Look

Bubble or not, TSMC is making money hand over chip.

Subscribe now

Did you enjoy reading this post? Hit the ♥ button above or below because it helps more people discover great Substacks like this one and it helps train your algorithm to get you more posts you like. Please share here or in your networks to help us grow!

Compute, Compute, Compute

💻 The Compute Theory of Everything ()
It seems that AI improvements closely track computing improvements.

💵 The Price of Intelligence ()
What if you could buy and sell GPU-hours like megawatt-hours?

📱 AI Has Left the Chat ()
Powerful compute driving local AI was the name of the game in Las Vegas.

🤖 CES 2026: Running With Robots ()
There were a lot of robots at CES.

Economics

🛑 The No Hire Economy ()
The economy seems fine by most numbers, but job creation has not followed. Why?

🛢️Venezuela’s Excrement ()
Oil has been Venezuela’s curse. Thomas Pueyo explains how it has come to dominate the economy, the government and life in the country.

📺 How did TVs Get So Cheap ()
The manufacturing tricks that let a 50” 4k TV land in best buy for $200.

Work

👀 Work Will Feel a Lot Different Next Year ()
Claude code, feral kids and AI property rights.

🗣️ Leadership by Extreme Ownership ()
How to succeed as a new leader by owning everything, blaming no one and building relationships for the win.

Culture

🎙️ How Video Podcasts Took Over Streaming ()
Video Killed the radio star, again.

💄 How To Build a Girl in Modern America (The Face)
Spend a weekend with University of Alabama influencer-student megastars Kylan and Izzy Darnell.

🏠The McModern (Architectural Digest)
Introducing the glassy, boxy current iteration of the McMansion.

The Look

Box office returns are declining. Via :

Subscribe now

Did you enjoy reading this post? Hit the ♥ button above or below because it helps more people discover great Substacks like this one and it helps train your algorithm to get you more posts you like. Please share here or in your networks to help us grow!

A Navigable Electronics Show

This year’s Consumer Electronics Show felt much more manageable than in years past. One could get from the Venetian to the Las Vegas Convention Center (LVCC) in a half hour rather than the hour of years past. Dining and casino floors seemed down commensurately. Hotel rooms on the strip at non-sketchy hotels were available on show nights for less than four digits. Inside the halls, one was not in the same crush of humanity. You could walk around most booths.

Some of this might not be CES being down. The show was a bit more distributed, with the NVIDIA-focused Foundry at the Fontainebleau. Perhaps the show’s early January timing, without a week buffer between the holidays, caused a little drop.

To be clear – the show was still busy, it was an exciting place to be and the conversations were great. One thing that struck me about the audience this time – it felt like there was a higher proportion of international attendees than in recent years.

Central Hall Changed

LVCC’s Central Hall has long seemed like CES’ center of gravity. It held the massive booths with the collections of televisions that drove the tradeshow. As long as I have been going, it was defined by three anchor points. You would walk in through LG’s massive booth, route back to Samsung’s giant enclosure before wandering to Sony’s massive presence in the back corner.

In 2026 LG was still in place. Samsung decamped to the Wynn – effectively getting a larger booth but also setting up to control access, hours and environment in a way you cannot in the middle of a tradeshow floor. Sony shrunk into a booth co-branded with Sony automotive. What was once the largest booth on the floor, an anchor tenant was no more.

In Samsung’s place in the center of the floor lay China’s TCL and Hisense. TCL was the largest by a smidge, and they both looked great. They had many large screens, all manner of other appliances and major sports tie-ins. The new Chinese majors are stepping up.

AI: Robotic Reality

The marquee AI news came from the chip giants. NVIDIA unveiled Rubin, while AMD announced the M14s—both representing significant architectural advances. LG and Samsung draped their entire presences in AI messaging, positioning artificial intelligence as the defining capability across their product lines.

The AI partnership landscape revealed interesting dynamics. “Works with Gemini” integrations appeared frequently, with scattered “Works with Copilot” presence as well. Notably absent: any visible “Works with ChatGPT” or “Works with Claude” certification programs. From conversation it seemed like Gemini’s advantage centered on its better vision capabilities, a function a bit closer to consumer devices which need vision to operate in many cases.

The big new AI product feature in TVs was that the AI has gone beyond helping you find something in the guide. It was now capable of interacting with content on screen, doing things like looking up characters, explaining scenes and context. Search gets a lot better too – you can do things like ask the TV to find scenes by description. It is convenient, but you are probably doing this with your phone already.

Robots were where one saw the most visible application of AI. They were everywhere this year, doing all manner of tasks, up to and including boxing with each other. They were nearly everywhere you looked, a little too close to customers to be that far off from production. I suspect you will see one sooner rather than later.

Overall AI is starting to get refined into the interface layer and into devices more than being a sticker on everything like it was in 2025.

New Product Categories and Trends

These are a few overall product trends which stuck out to me beyond the AI layers.

Retro Revival

If I had to pick a single visual throughline in products it would be retro. It crept into many different products from across the floor. Bluetooth keyboards shaped like typewriters for example. Some companies are building offline experiences – the Await camera behaves like a disposable camera of old, limited to 24 exposures and no previews. You have to send it to a device for processing. Clicks cases turn your smartphone into a BlackBerry of sorts with a keyboard. Nostalgia sells and it seems offline has a certain charm in 2026.

AI Enabled Smart Rings

There is a new wave of smart rings that are interactive tools rather than health trackers. In particular, the ring makes a great place to launch an AI assistant. They are big enough to have enough microphones. The placement on the hand makes it easy for you to tap a record button as required. Vocci was a good example – working even in tough Eureka Park noise and network envelopes. I am excited to see where these devices go.

E-Ink Art Displays

Brighter, color e-ink screens are powering a new wave of wall art displays. They make more sense than using a real screen – static screens consume next to no power, allowing for battery power with year-long charging cycles. The matte, paper-like screens match modern aesthetics. Lack of backlight means they don’t glow at night and behave like a painting not a screen. Some implementations, such as Framic, let you create AI art for your frame while others include public domain galleries.

Exoskeletons

Mobility augmentation can change lives. It feels like we are getting much closer to consumer exoskeletons – external frames that can support people doing repetitive labor or with mobility problems. There were multiple implementations on the CES tradeshow floor in consumer-ready implementations. Given our aging population this could be an exciting growth area in future years.

The Bottom Line

CES is a slog. I walked 25 miles over two days. The exhibits I saw during the march illustrated two things to me. First, the big Chinese companies are investing heavily in splash, not unlike their Korean and Japanese predecessors who have moved out to quieter venues for more refined conversations. Second, advances in batteries and sensors are combining with manufacturers learning how to apply AI.

Christmas 2026 should be an interesting shopping season.

References

Here are some handy references about this show in case you want to read further.

• Last year’s coverage from this newsletter.

• CES History from 1440.

• Wired’s CES 2026 Live Blog — great curation of the best devices.

• Tom’s Guide to CES 2026 — lots of close ups with devices

As I write this, I am off to fabulous Las Vegas to explore this year’s Consumer Electronics Show. The first of these longer-form posts will be the parts of my trip that do not have to stay in Vegas. It should land a bit later this week, hangovers and travel willing.

A Different-Feeling CES

For the better part of the last few decades, my year has effectively started at the Consumer Electronics Show. I have lived through several waves of technology and finance at that venue — from the Great Financial Crisis through COVID and now the AI boom. Booths and vendors have ebbed and flowed throughout these cycles, but I have never lived in a world where Central Hall did not feature giant booths from Sony and Samsung until 2026. Change is in the air.

The writing was on the wall last year. Sony was in the same place, and the same historical size. It was a giant, threadbare booth full of actors cosplaying game characters with nary a Bravia TV to be seen. Like they had wanted to pull out, but were too far along to completely pull out. This year Sony have punted entirely, in the same corner but a third the size and in partnership with Honda.

Samsung has gone the other direction. In hindsight, they have always been looking to break out of their boxy booth for a while. I recall a VR roller coaster in the concourse that was almost worth the wait in line sometime around 2012. Last year, the Samsung booth was packed to the gills and most of the point was lost in the noise and the crowd. This year they decamped and are aiming to better control their narrative by hosting a standalone Samsung Experience in the Wynn.

It is an exciting time in technology — perhaps more exciting than most. Between the AI boom-or-bust cycle and tariff uncertainties this will be an interesting CES. Here is what is top of my mind as I pack for Vegas:

• Is this an evolutionary or transformational CES? Are we getting good and orderly progress or a different world?

• How on-device AI is progressing. Last year, almost everything was little more than a physical wrapper around a cloud-based AI solution. Everything was slow, clunky and ultimately commercially unsustainable.

• Smart glasses have been looking promising. Is this the year smart glasses finally come together and become a viable mass consumer interface? It seems like the elements are in place to make it happen.

I will report back on what I see and learn later this week in our first CES special.

Subscribe now

Predictions Distilled

It is the season for glittering predictions about 2026. I read a lot of them, here are some that stuck out in three buckets for easy consumption.

AI-Focused Predictions

AI pundits love to predict. Here are some of the more interesting thesis, ranging from most negative to most positive. A few common threads include predictions about the status and outcome of the AI bubble, legal issues getting settled and AI backlash beginning to take various forms. World models are also something to keep an eye on as we evolve past the LLM.

Nine AI Predictions for 2026 ()
The bubble could burst, causing a financial crash. Sam Altman has a very public crash out as AI anxiety spikes.

Six (or Seven) AI Predictions ()
We are not getting closer to AGI, we are at peak bubble and expect the AI backlash to increase.

Six Themes for 2026 ()
Six fairly broad and workable predictions, the Apple one is especially interesting.

2026 Predictions ()
Another intriguing set of predictions with confidence ratings. He sees hollywood collapsing and the job loss narrative dissapting.

17 Predictions for AI in 2026 ()
Timothy B. Lee gathered a panel of experts to share a slate of interesting AI predictions including degrees of confidence. The experts see good and orderly progress by and large. Scorecarding this in late 2026 will be fun.

Graphs for Thought

The right graph can capture a trend. Here are a few to think about for 2026.

10+ Graphs that Picture 2026+ ()
A fascinating series of images that illustrate where parts of the economy are going. And voice first AI is certainly going to be a thing in 2026.

Six Major Trends to Watch in 2026 (Chartr)
Six eye popping charts to wrap your head around the next year.

The Wider World

Here are some other 2026 predictions I liked.

26 Predictions for 2026 (Part II) ()
Thirteen thesis worth considering for your 2026 investment strategy.

2026 Predictions Across Every Category ()
The title is ambitious but the article gets there. Great food for generalists.

15 World-Stunning Scenarios (Politico Magazine)
AI-driven stock market flash crashes and Patriotic Innovation Zones, to name a few ideas about what 2026 could bring from leading thinkers and futurists.

Sports Crystal Ball (Front Office Sports)
Sports sit at the intersection of culture, media and finance these days.

Synthetic Interests

The Story Behind the Roland TR-808 ()
The Roland TR-808 was a device putting sound generation, microprocessor control and onboard memory into one machine. To make it happen, they bought Mitsubishi chips that failed quality control, creating the signature kick, snap and crash.

Sweet Dreams are Made of This ()
How the Eurythmics’ iconic song came together.

Subscribe now

Did you enjoy reading this post? Hit the ♥ button above or below because it helps more people discover great Substacks like this one and it helps train your algorithm to get you more posts you like. Please share here or in your networks to help us grow!

Production Note: This is the last issue of 2025. We will be taking a short winter break, see you in January.

2025, Distilled

2025 has had certainly had its fair share of plot twists. We have covered many of them here. If you told a 25 year old me that, in a few decades time, I would be sitting at home being entertained by AI-generated videos in an America where Donald Trump just legalized marijuana I probably would have called you crazy. But here we are. I really appreciate everyone who read and followed. Thank you for your time and your trust.

When you are running a solo operation, retrospectives and lookbacks can be challenging. You really need a writers room. NotebookLM changes that game entirely. It is an amazing tool — one that we have touched on before. For the unfamiliar, it is a Google tool that lets you load a variety of sources into a notebook and then to interact with them using AI.

Like all of the other Google AI tools, NotebookLM is dramatically better of late. Gemini 3, Veo and Nano Banana Pro provide the tools to craft amazing presentations and videos about the content. Google has added a ton of new features and quality of life tweaks to existing ones. Not only can it generate video overviews, you have the ability to guide the style and focus of the output. It is a truly amazing way to explore information.

Rather than just talk about this you can see it in action. Google added the ability to share a notebook publicly. I invite everyone to come explore the Zeitgeist Distilled of 2025 today. Google requires authentication, and I don’t have a way around that so you will need a gmail account. Once you are inside of the notebook, you will be able to interact with the content using the chat features. In addition, see the studio panel on the right for videos and infographics. Or keep reading for some highlights from those.

Zeitgeist Distilled 2025 Notebook

The Video

There is a technological boom happening on one hand, and massive societal gloom on the other. The response is intellectual survivalism and we move into a probabilistic era.

The Deck

Another way to look at the Zeitgeist is as a tension between three paradoxes. The AI paradox of a creation boom and a jobs and economic bust. The reality paradox where technology is permeating reality while still remaining fragile. The human paradox where we are the most connected and the most isolated mankind has ever been. NotebookLM’s presentation explores those conundrums effectively.

The File

The Vibe Work Paradox

21.6MB ∙ PDF file

Download

Download

Key Slides

In case you are not in a good place to consume a 21mb PDF, here are the key slides.

Pillar 1

Pillar 2

Pillar 3

Conclusion

No, the Zeitgeist is not a Destination

Thanks for taking this journey with me, I really appreciate your time. I hope everyone has some time for some holiday relaxation and reflection. We will be back in mid-January to continue distilling the Zeitgeist in 2026!

Happy holidays!

Subscribe now

The Distilled Spirit

A Very AFCON Christmas

The football gods have given us an AFCON for Christmas. It is an amazing tournament — probably the best on the world. There is always drama and international excitement. There are a variety of play styles and real unknowns. Take some time to enjoy it however you can legally find beIN sports. Here are a few articles to get you in the mood.

🖼️ AFCON and the Art of Storytelling ()
AFCON is full of stories. Here are some good ones to get you started.

🌍 The Magic of AFCON ()
Modern AFCON is a celebration of African diaspora.

📕 AFCON Pocket Guide ()
An amazing guide to the whole tournament, including a team support selection flowchart.

Selected 2025 Wrap-Ups

The flood of 2025 wrap ups has started, this publication included. Here are some good ones.

📰 Best Reads of 2025 ()
Garrett curates amazing links, here are his picks for the year.

📈 8 Lessons for Investors ()
This has been a strange year in the market. Sam puts a bow on it.

🧠 2025 LLM Year in Review (Andrej Karpathy)
The man who coined the term “vibe coding” and is a leading mind in AI wraps up this year’s advances in five points.

🔗 20 Best Links of 2025 ()
Catlin wraps up the cultural story of 2025 in 20 links.

Big Ideas

Some though-provoking pieces to close out the year.

🧐 The Revolution of Rising Expectations ()
Rising expectations and requirements more than meet rising wages.

⛔ Make Everything Harder to Use ()
Can making things harder to use make them more valuable?

🎧 Why iPods are Making a Comeback ()
Your vintage iPod might be worth something. Here is why.

🗣️ How to Become Well Spoken ()
An exhaustive list of ways to become better spoken.

😨 Backtesting CNN’s Fear & Greed Index ()
Can you trade on the venerable CNN Fear & Greed index?

All About AI

🤖 AI Agents are Eating SaaS (Martin Alderson)
Software as a service ate the business world. Now your AI tools are replacing your SaaS tools.

🏗️ Configuration Operations is the Key to AI ()
AI infrastructure gets more expensive, the price per token drops. Downtime will be expensive. The winners will be able to deploy fast and clean.

⚡AI Power Accelerating Energy Transition ()
AI needs unimaginable amounts of power. Clean energy scales faster than fossil fuels and is better positioned to meet the needs.

🤝 What is Working For Companies (Every)
Success is driven by knowing what you want to achieve, having buy in from leadership and it really helps when you have documentation.

Holiday Fun

🎄 Six Christmas Questions You Never Thought to Ask ()
St. Nick is weird and people give some truly bad gifts.

🏰 67 is Old School ()
Your kid is probably saying 67. So did Chaucer.

The Look

shares some hope for the holidays, starting with a gentle reminder that the thing that has long term impact probably is not the leading headline.

Subscribe now

Happy Holidays!

If you read this far down — Thanks for reading, I really appreciate you taking it all in. Have a great holiday season and a happy new year!

Did you enjoy reading this post? Hit the ♥ button above or below because it helps more people discover great Substacks like this one and it helps train your algorithm to get you more posts you like. Please share here or in your networks to help us grow!

The chatbots we have today, tools like ChatGPT, Claude, Gemini and their competitors, are amazing. More often than not, they understand your imprecise instructions and come back with useful output. Most of them even help package your work through features like Custom GPTs, projects, and spaces. That convenience comes at a cost: abstraction hides how things actually work. Control can be imprecise.

Sometimes you need more finite controls or a deeper understanding. One way to get there is to write code, but most AI vendors also offer studio tools that let you work directly with the API. These tools pull back the curtain: you can see how prompts are constructed, compare variations, and more easily build applications, automations, or even lightweight publishing workflows.

What Platform Tools Can Do

Platform tools such as Google AI Studio, OpenAI’s Dashboard, Claude Console, and Perplexity Playground let you interact more directly with AI models than a standard chatbot. They provide structured tooling for developing, testing, visualizing, and deploying AI workflows.

These “Platform Tools” sit between pure end‑user applications like ChatGPT and working with the API directly in code, offering more control without requiring full‑scale development.

They are not free, but usage-based pricing is often cheaper and more flexible than paying for subscriptions that sit idle. They usually aren’t expensive, but costs can add up if you are pushing a lot of content or generating a lot of media via the API.

The magic of platform tools is that they remove a layer of indirection, giving you direct control over how the model behaves. They expose many of the knobs and switches behind the scenes of your favorite chatbot. Because you can be explicit about options, you can get much more consistent results. The tools also map to other behind-the-scenes AI tools.

Platform tools make you a better magician.

Google AI Studio is What Gemini Should Be

Gemini is great, but the consumer-facing tool is fairly limited, which is why AI Studio is such a step up. There isn’t a desktop app, and the web application has few bells and whistles. Nano Banana is hard to configure — getting aspect ratios right is hit or miss, you always have a little Gemini watermark to contend with. Google’s AI Studio is what Gemini should be. It is feature-rich, with lots of options and ways to manage your AI conversations and projects.

The name fits: there is a lot packed into the tool. Some features it includes are:

• The Playground is the central feature. It lets you prompt most of Google’s models, lets you leverage System Instructions for most of them. These can be saved and reused, dramatically speeding up development.

• Using the Playground, you can harness the astounding image generation powers of Nano Banana much more effectively. Beyond System Instructions, you can control output settings like aspect ratio and specify quality from 1k to 4k. Images do not have the Gemini watermark like they do with the normal tool.

• You can generate two-person voice conversations in a variety of voices using scripts you supply.

• There is a Veo 3.1 app for creating video from text, frames or with an example image.

• You can build small, interactive applications that run within the tool. These apps can be published to a gallery or to Google Cloud. You can also browse the gallery and run other creator’s apps.

The tool has a lot of nooks and crannies, but its depth is its strength. It is the way to get the most out of the Google toolchain and a markedly better prompting environment than Gemini.

Finally, if video is more of your thing, see Futurepedia’s walkthrough on YouTube. Also note that Google has a few competing AI studio products.

OpenAI’s Omniscient Dashboard

OpenAI’s dashboard, compared to Google’s Studio, is more developer-focused. A primary use case is a place where advanced end users can build prompts that can be versioned, saved, and called in other applications, bridging the gap from the front lines to development. Like everything else, OpenAI does a good job of making a functional yet deep user interface.

Here are some things you can do with the tool:

• Side-by-side prompt testing is very well implemented. You can tweak just about any conceivable knob in a repeatable manner. Your chats can include calls to storage, MCP or tools.

• The new and exciting Agent Builder tool lives here; you can build workflows in ChatGPT that can be embedded in other apps via the ChatKit SDK.

• OpenAI’s lead in audio shows — you can test conversational tools with real-time audio. You can have a conversation with any prompt you choose to share. A text-to-speech tool is also included for the occasions you need to generate audio.

• You can create media, though this looks to be little different than the Sora app from what I can see. Unlike Google, audio is not included so the text-to-speech might become handy for the right clip. Note that you must verify your account to create media, OpenAI is strict about API access here.

• OpenAI includes a vector store interface that makes it easy to store documents for your chats to work with. If you have a static set of knowledge, this tool is for you.

Claude’s Console & Perplexity Playground

Compared to Google and OpenAI, Claude and Perplexity do not offer full, end-to-end platform suites. They still offer important features for developers and advanced users.

Claude is most famous for Claude Code, but for advanced users they also make Claude Console. The killer feature of this tool is the prompt rewriter — it can step in and help make your prompt a lot better before it executes.

Perplexity features the Playground, where you can execute a Perplexity query with fine-grained options. It is primarily an API test tool, but it is a good one, and I have found it useful from time to time.

The World Behind the Curtain

Sometimes you need to see behind the curtain. Sometimes you need more consistent control. With these API-adjacent tools, you can do just that on all of these platforms. Many other vendors have similar features — do not be afraid to rummage around the settings screens and developer portals to see what options you do have on a given platform. You never know what superpower you will discover.

Subscribe now

The Distilled Spirit

Warning Signs

🧓 Boomer Caregiving Will Break Us ()
The oldest baby boomers are turning 80. Senior citizens will outnumber children in America by 2034. The caregiving tsunami is here.

🏭 The Manufactured Rise of Nick Fuentes ()
According to a recent study, right wing influencer Nick Fuentes’ message is amplified by an array of foreign-controlled fake speech bots, not actual fans.

💪 The Era of Brotein ()
What does all the protein do to a man?

2026 Predictions

🔮13 Predictions for 2026 ()
The Digital Native sees Anthropic having an IPO, you Insta feed will be 25% synthetic and more in part 1 of his 2026 predictions.

💡 26 Ideas for 2026 ()
Derek Thompson sees an America on drugs going through an AI backlash while toiling in a casino economy.

🛍️ Consumer Trends for 2026 ()
Gen Z and Millennials are entering their prime spending age and they are different consumers than the previous cohorts. A fascinating look into the mind of the American consumer in 122 chart-heavy slides.

AI Trends

🤖 Embodied AI Hits Stride in 2025 ()
Embodied AI, such as Robots, really took off on 2025. World Models, VLAMs and Chain of Action are making real world deployment more of a question of “when” and “how many” than “if?”

📉 That’s No Enron (Where’s Your Ed At?)
NVIDIA is not like Enron, but does that mean NVIDIA is not a scam? An exploration of the bear case for AI investment.

🎈 The AI bubble is not Popping Yet ()
The market thinks Wall Street’s NVIDIA estimates are way overvalued. As long as the hyperscalers keep needing to scale it will not fall. There might be better places to look for returns.

How To AI

✍️ Writing Faster With AI ()
Using AI like a team makes you a better writer. Here is how one author does it.

5️⃣ ChatGPT Leadership Therapy ()
Leadership is lonely. ChatGPT can be very cathartic.

🔁 Recursive Prompting ()
A five step process to iterate over questions and get the right advice for questions. Really quantifies a lot of things I have been doing in an explainable package.

🔍 Understanding Perplexity ()
Perplexity has lots of knobs. Here is how you can use them effectively.

🔨 Useful Patterns for Building HTML Tools (Simon Willison)
HTML tools are single file applications you can easily make with LLMs. Here are tips on building them from one of the best.

📊 Guide to Data Visualizations ()
One great use of said HTML tools: data visualization.

Fun Stuff

📅 52 Things I Learned in 2025 (Tom Whitell)
Tom’s annual list is back and not to be missed.

🧑‍🎄 ACL’s Holiday Gift Guide ()
This one is so good it does not matter it is so late.

🔪 The Life Changing Magic of Opting Out ()
Ho ho ho, just RSVP no . . . .

🧸 A History of the World in Japanese Toys ()
You can tell a lot from toys.

🏀 Maryland’s Basketball Era ()
Let’s go Maryland!

The Look

and do an amazing job covering How to AI for Non-Technical Late Bloomers for AI beginners.

Subscribe now

Did you enjoy reading this post? Hit the ♥ button above or below because it helps more people discover great Substacks like this one and it helps train your algorithm to get you more posts you like. Please share here or in your networks to help us grow!

The Lightweight Database in the Sky

Airtable looks like a relational database, feels like a spreadsheet and smells like an event store. It has aspects of lots of different data tools and does a very good job of what it sets out to do. It is not perfect for everything but it is good enough for many things. It can be a flexible layer in your stack if you can fight the integration battles successfully.

I have been using Airtable successfully for the last four years. Some of the more interesting projects include:

• A multi-Airtable base solution for managing event space requests, event planning and fulfillment with a conjoined production tracking and management system for a fairly busy event space and team. Includes a dedicated multi-tenant approval system, integrations with Box and Slack via Zapier to manage file storage and communications channels.

• A heavily automated AI news tracking system. Ingests articles from SLACK and other tools, articles are automatically loaded and summarized. Distributes newsletters including AI summaries via multiple channels weekly.

• An automated social media tracking system that watches a large number of users but uses AI to filter the posts down to a more manageable number for human curation.

• The brains of a very high profile exhibit.

Here are my notes after a few years of using and building on the platform.

Elegant Simplicity

Airtable fulfills its core promise: it provides a simple, approachable database that empowers advanced end users to build effective solutions quickly. The interface is clean, intuitive, and accessible without sacrificing capability.

For many users, Airtable never gets past the spreadsheet. Just like Microsoft Access many users just use the tables and never get to the forms. In Airtable’s case, the interface layer is known as “Interfaces”. The tool started pretty rough and fairly limited, but Airtable’s product team has put a lot of work in and it has dramatically improved. Mobile is there too now in some capacities, closing a major hole in the stack.

Airtable forms are another example of this. The updated versions are a huge step up over the legacy implementation. They look much cleaner and are much more flexible. They are still limited — no ability to create related records is a particular pain point in many workflows. At the end of the day, most users can build a table, then build a form and share it which is a net win.

Synchronization and Automation

One of Airtable’s best features is sync — the ability to synchronize data between bases. You can choose a set of records which are effectively included in the other database instance, sharing data automagically. You can even extend this data in the other base, it is effectively a first class citizen. Airtable has recently enabled 2-way and related record synchronization, opening more doors with this great feature.

Automation is another core feature of the product. It features a pretty limited but effective for what it is automation tool. Simplicity is the main selling point — it is not as complicated as a Zapier or a Make, making it a lot easier for many users to approach. They have some pretty elegant triggers — the ability to have different triggers on different fields is useful. It supports incoming webhooks as a trigger which opens the door to many possibilities.

Automation is another space where they have been investing. The recent addition of an email trigger is a nice feature. Especially combined with the new AI features, you can imagine workflows where users don’t even need to use forms, they can just email you the request and AI can extract the information to the system just like they wanted to do in their heart of hearts.

Airtable automation is great — but it has a lot of limitations. You can only have 50 automations for a given base. It does support branching logic, but it is limited to a single instance within your automation. Loops are also supported, but mutually exclusive from the branching logic. Finally error handling is very, very limited at best. Notifications only go to the last editor with no way to have someone on pager duty. If the ETL process matters, you probably need to manage it upstream from Airtable and hand Airtable very clean data for your own sanity.

Highs and Lows of Collaboration

Airtable collaboration within the enterprise is great. Sharing among users in your workspace or on your team is fun and easy. If anything, the IT administrator in me thinks it is a bit too easy to share and too hard to control.

Sharing beyond your workspace — especially in the enterprise — is difficult, and outside editing for smaller groups is particularly tricky. As a result, many teams rely on tools like Softr or Fillout to build workable 2‑way extranet solutions.

Lack of an exportable reporting layer is another challenge. Sometimes you just need a printout or a digital equivalent to ship. If you are not in Google Docs it is a challenge to get anything useful out of the tool on that front. They need Office 365 integration and Word on the list. Excel (and Google Sheets) would also be an important addition to the stack.

This Developer’s Wishlist

Airtable is great, but there are some things it really could do better.

As mentioned above, automation management needs some love on a few levels. Error handling would be a great place to start there.

External collaboration and better reporting options would be great. A native JSON export would be a really good add too.

Airtable is not a relational database, but it would be great to do some set based operations somehow. Having some form of SQL would be absolutely amazing.

The biggest flaw is the lack of validation — you cannot validate a record before executing a button click, at least not directly. You have to roll back the code, then you have no way to tell the end user their action did not take. Some way to do this would be life changing.

An Exciting Future

Airtable is a great platform and it is a really good choice for a generic, lightweight database in the sky. I would not hesitate to use it in most operations. It is flexible yet reliable and easy enough to pick up. They are doing a good job of continually improving the platform and I’m excited to see where it is going.

At its best, Airtable embodies a modern tooling philosophy: give people just enough structure to stay organized, and just enough freedom to build what they actually need.

Subscribe now

The Distilled Spirit

The Lead

🥰 I Love AI, Why Doesn’t Everyone? ()
It turns out being excited about AI is a minority position in today’s America.

📉 The Vibecession Era ()
Are we are living a vibecession? What is causing it? A deep dive into both questions.

🫧 Which Parts of AI are in a Bubble Today? ()
AI investment exists on several planes, some of it is a saner play than others. What you are in matters.

🎲 The Danger of a Gambling Society ()
Prediction markets make it easy to bet on anything, creating a financialization of reality. Gambling has moved from the fringes of society into center stage. What could be the harm?

Technology Then and Now

🤖 Agentic E-Commerce Is Coming ()
Major AI vendors are building commerce experiences directly into their chatbots. It is going to change retail, but how?

🌐 JavaScript: the Hack that Runs the Internet (Arstechnica)
Thirty years ago Brendan Eich hacked together a working prototype of the language that runs the internet and many other apps you know and love.

🪖 Syndication Format Wars (Buttondown)
RSS won the syndication wars against some big companies. It works today — every substack has a RSS feed, including this one.

AI Pro Tips

📓 NotebookLM Video Guide ()
Google’s NotebookLM is an amazing tool. This is a great text and video guide to taking advantage of all of its new features.

🧑‍🤝‍🧑 The Ultimate Guide to Gemini ( )
This is a great intro to Gemini especially if your world was ChatGPT before the recent updates converted you.

🖼️ Creating Complex Infographics ()
AI made cool images, but complex infographics were beyond the tool until Nano Banana Pro hit the scene.

🧑‍💻 Claude Code for Non-Coders ()
Claude code can also organize your files and do research.

Fun Stuff (Read: More Gift Guides)

🥚 Egg Color Guidance ()
It turns out it what is inside that counts.

🧳 Things We Never Travel Without ()
Travel accessories for the discerning traveler.

🎁 Good Thinking Gift Guide ()
Gift like the cool kids. Lots of interesting brands.

The Look

The pandemic recovery has been K-shaped.

Subscribe now

Did you enjoy reading this post? Hit the ♥ button above or below because it helps more people discover great Substacks like this one and it helps train your algorithm to get you more posts you like. Please share here or in your networks to help us grow!

Three years ago, ChatGPT burst onto the scene. Since then, the rise of Large Language Models has dominated nearly every conversation about technology. Here are five thoughts on the current state of AI, knowing what we know today.

1. Forget the benchmarks

There is a lot of digital ink spilled about which LLM model is the latest in what benchmark. It is largely immaterial. Switching tools to whoever has the best benchmark score this week is a false strategy. You will just confuse yourself. There is always going to be a marginally faster model coming, but you don’t need to pay attention. Focus on tools that connect to your data for the big wins.

2. Connectivity Becomes the Differentiator

As AI becomes commoditized, connectivity starts to become the big differentiator. Technologies like MCP are starting to unlock this. Today, most of your major tools let you connect with your cloud files, your email and other tools. On the other hand, many system of record sorts of tools are developing their own AI integrations. You can start to see the world where your data is easily accessible to AI take shape.

Tools like Notion and Airtable have pretty impressive AI baked into them, letting you use AI alongside your data in real time — as explored further in this piece. The future is now.

3. Google 🚀🚀

Google stumbled hard out of the AI gates. For readers who may not have tracked this closely, Google’s early releases—like Bard—were widely seen as rushed responses to ChatGPT that lacked polish and competitiveness. Their first versions were not good, and they took a long time to get the product together. The giant is now awake and ready to rumble. Gemini 3 is very, very good. Nano Banana’s image creation prowess is unsurpassed; 🍌 has supplanted ChatGPT for this publication’s image needs. NotebookLM has grown from side project to an amazing knowledge tool I could not live without.

Behind the scenes, Google has long built their own hardware, and their homegrown TPU computing units now power much of their AI infrastructure. Recently, they have started looking at selling the devices commercially. Beyond hardware there is a huge data moat. YouTube and Gmail are huge assets. The internet graph is another.

Last and certainly not least, Google can afford to do this out of the massive cashflow generated from their existing business. They do not need anyone’s capital to build it out. They are very well positioned to lead with AI. The $160 share prices we saw not too long ago might have been a generational buy.

4. OpenAI 🤷‍♂️

Google and the other hyperscale tech companies in AI are largely spending current cashflow to fund future dreams. OpenAI is burning capital at an impressive clip, underscoring their urgent need to monetize more effectively without repeating the same point about scale. They have an impressive number of active users, but not on Google’s or Meta’s scale. Sam Altman is forced to keep pushing the envelope and increasing the leverage because he has no other options. The ChatGPT dream is getting bigger because it has to get bigger or die trying. Currently it seems they are at some sort of American super-app a la WeChat. You can see this starting — group chat and shopping are here, ads are coming. They need to figure out some way to monetize a lot of those users fast.

When I hear people talking about the AI bubble, I wonder if they are really talking about how far over their skis OpenAI might be and using that as a proxy for all of the technology being that far over its skis. Don’t mistake one early adopter’s stumbles for failing technology. My Space did not last either.

5. Limits of LLMs?

It is not entirely impossible that we will soon hit a hard wall in the performance of Large Language Models. Compared with the dramatic capability leaps of 2022 and 2023, today’s advancements feel far more incremental, reinforcing the sense that we may be approaching the limits of this architecture. Recent updates are certainly more incremental. Gary Marcus, who has long been skeptical of the longevity of this technology, is bordering on doing victory laps. If the LLMs do run into a wall, what happens with OpenAI and the other AI-only companies that are largely bets on continually increasing LLM performance?

Year four of the AI boom will likely hinge on how quickly connectivity matures, whether Google’s resurgence continues, and whether LLMs can push past their potential ceiling. Year four of the AI boom will be exciting.

Subscribe now

The Distilled Spirit

AI Depreciation Bubbles

AI, depreciation and bubbles were all over the news again.

🌍 AI is Eating the World (Benedict Evans)
Benedict Evans’ annual presentation on technology reminds you that AI is advancing like every wave of tech before it in many ways. The waves are also getting bigger every cycle.

👀 Google > OpenAI + NVidia (Stratechery)
Google strikes back — do they have the wherewithal to penetrate NVIDIA’s moats and OpenAI’s lead?

🌐 Financial and Other Bubbles ( )
Why technology bubbles like what might be happening with AI are different than financial bubbles like we had in 2008.

📝 The Depreciation Battleground ()
Michael Burry was a main character in the 2008 crash. He is betting on this bubble, and his thesis is centered on the deprecation of GPUs and datacenters. Does he have a point?

Things to Think About

The most interesting reads of the week.

📰 Read the Damn Article ()
A very eloquent reminder on the value of reading the article, not just the re-written headline.

😀 What Captures Our Attention Now? ()
Media and entertainment are deeply fragmented. Figuring out what captures attention now required a exploration and aggregation across a number of data sets to find a thread that might not be there.

😠 The World’s Polarization Side Hustle (404 Media)
A new feature on X showed that many of the divisive and hateful accounts fanning the flames of social discord hailed from places like Bangladesh, Vietnam, Cambodia and Russia. Most of it is tied to monetization programs where American eyeballs pay the most.

AI Applied

How you can apply AI to your life and work.

🧑‍🎨 Can AI Replace Powerpoint? ()
Gemini 3.0 and NotebookLM can actually make decent presentations you can really present from. Very close — you can only re-prompt, not edit. But the writing is on the wall.

🤖 Reverse Engineer Yourself to Train Your AI ()
A really cool prompt to examine your own writing, spot your patterns and train a tool to help you compensate for them.

🥇 How to AI Like the Best ()
Using AI effectively requires teaching AI exactly how to do it so you don’t get randomly generated slop.

Gift Guides

I heard people here like gear. We had a great gift guide last week, but we were not alone. Here are a few I liked.

• has great curation and a great and wide-ranging holiday list.

• Steven Raichlen writes great BBQ and a wonderful gift guide for the outdoor cook in your life.

• writes about product and AI has a solid list of modern gifts.

• The writes about fashion and has a cool sub-$150 list.

• is running a series of gift guides for the soccer fan in your life. They started with stocking stuffers and intend to continue all week.

• PI.FYI is an amazing community. They have a crowdsourced gift guide worth exploring.

The Look

does it again in the latest Six Chart Sunday where he shares how to navigate this age of disruption we share.

I’ll add subscribing to this publication speaks to the first row explicitly.

Subscribe now

Did you enjoy reading this post? Hit the ♥ button above or below because it helps more people discover great Substacks like this one and it helps train your algorithm to get you more posts you like. Please share here or in your networks to help us grow!