In 2016, I received this e-mail from my oldest son.
Date: Saturday, October 22, 2016 at 7:09 PM Subject: FYI: Security Stuff
FYI – I just got an alert that my email address and my Gmail password were available to be purchased online.
I only use that password for my email, and I have 2-factor enabled, so I’m fine. Though this is further proof that just about everything is hacked and available online.
If you don’t have two-factor enabled on your accounts, you really need to do it.
Since then, security has only become a bigger issue. I wrote about the Equifax event, but there are countless examples of similar events (and yes, I mean countless).
When people think of hacking, they often think of a Distributed Denial of Service (DDOS) attack or the media representation of people breaking into your system in a heist.
In reality, the greatest weakness is people; it’s you … the user. It’s the user who turns off automatic patch updating. It’s the user who uses thumb drives. It’s the user who reuses the same passwords. It’s the user who falls for social engineering. Each of those choices may seem like a mistake, but they also represent some hacker’s favorite pattern to exploit.
Whether it’s malicious or unintentional, humans are often the biggest security weakness.
It’s impossible to protect yourself completely, but there are many simple things you can likely do better.
Use better passwords … Even better, don’t know them. You can’t disclose what you don’t know. Instead, use a password manager like LastPass or 1Password, which can also suggest complex passwords for you.
Check if any of your information has been stolen via a website like HaveIBeenPwned or F-Secure.
Keep all of your software up to date (to avoid extra vulnerabilities).
Don’t use public Wi-Fi if you can help it (and use a trustworthy VPN if you can’t).
Don’t put information into GPTs that you want to keep private.
Have a firewall on your computer and a backup of all your important data.
Never share your personal information on an e-mail or a call that you did not initiate – if they legitimately need your information, you can call them back.
Don’t trust strangers on the internet (no, a Nigerian Prince does not want to send you money).
How many cybersecurity measures you take comes down to two simple questions … First, how much pain and hassle are you willing to deal with to protect your data? And second, how much pain is a hacker willing to go through to get to your data?
It doesn’t make sense to put all your data in a lockbox computer that never connects to a network … Nevertheless, it might be worth going to that extreme for some of your data.
Think about what the data is worth to you, or someone else, and protect it accordingly.
My son reminds, “You’ve already been hacked … the important question is whether you’ve been targeted?” Something to think about!
For most of human history, scams were relatively simple. A stranger sold fake miracle cures from the back of a wagon. A con artist ran a shell game on a busy street corner. Someone forged signatures, counterfeited checks, or promised riches through a too-good-to-be-true investment scheme. The tools changed with each era, but the mechanics stayed familiar: gain trust, create urgency, exploit emotion.
The internet accelerated everything.
In the early days online, scams were often obvious and almost amateurish. Chain emails promised bad luck if you didn’t forward them to ten friends. Pop-ups claimed you had “won” a valuable prize. People in chatrooms pretended to be tech support employees asking for passwords, while others sold fake concert tickets or nonexistent items through forums and classifieds. On platforms like AOL, MSN Messenger, IRC, and early online marketplaces, anonymity created an entirely new playground for deception. Most scams were still relatively small-scale, relying on volume and the assumption that someone would eventually fall for them.
Today, fraud operates at an entirely different level.
Modern scams can involve organized criminal networks, rogue nation-states, stolen datasets, spoofed phone numbers, AI-generated voices, cloned websites, and highly targeted psychological profiling. A scammer might know where you work, who your bank is, what school your child attends, and which recent purchase is waiting on your doorstep. What once looked like a poorly written e-mail from a foreign prince can now look indistinguishable from a legitimate message from your employer, your bank, or even a family member.
Yet despite all the technological sophistication, the core principle has barely changed. Social engineering still depends on manipulating human behavior: fear, urgency, trust, greed, loneliness, authority, or curiosity. The software has evolved. Human psychology hasn’t.
The common thread across all of these attacks is that they exploit people more than technology. Fraudsters are increasingly bypassing traditional cybersecurity defenses not by breaking systems, but by manipulating trust, authority, urgency, and routine behavior. As AI lowers the cost of impersonation and allows scams to scale globally, social engineering is becoming one of the defining fraud risks of the digital era.
Sumsub developed the index based on four main factors: fraud activity, resource accessibility, government intervention, and economic health. Lower scores suggest better resilience against fraud, whereas higher scores indicate increased risk exposure.
Many lower-ranked countries face challenges such as weaker enforcement systems, limited digital protections, and economic instability. These conditions increase the likelihood of fraudulent activities proliferating.
But fraud isn’t only an issue in emerging countries.
America ranked 91st, putting us in the bottom 20% globally.
We’ve been preaching about cybersecurity for years, and as always, the greatest risk is the human-in-the-loop.
The difference now is that AI is making it cheaper and easier than ever to target that human at scale.
You may not have fallen for a scam yet … but I bet you’ve fallen for an AI video. Regardless, the odds that your team, your vendors, or your portfolio companies will be tested are rising fast.
Make sure you have protections and fail-safes at both the company and personal levels.
In the age of AI, we’re obsessed with better answers. But the real leverage may come from better questions.
It’s easier to solve someone else’s problem than your own. Why? Because your biases, emotions, and problem-solving frameworks become part of the problem. Likewise, your blind spots likely go unexamined when you’re both the observer and the subject.
As an entrepreneur, I strive to be objective about the decisions I make. Towards that goal, using key performance indicators, getting different perspectives from trusted advisors, and relying on tried-and-true decision frameworks all help.
Mindfulness as a Decision Framework
Combining all three creates a form of “mindfulness” that comes from dispassionately observing from a perspective of all perspectives.
That almost-indifferent, objective approach is also where exponential technologies like AI excel. They amplify intelligence by helping make better decisions, take smarter actions, and continually improve performance.
When I shot this video, AI was still relatively limited.
In just a few years, the technology has come so far. When I originally published the video, I suggested that:
The future of AI will likely be based on swarm intelligence, where many specialist components communicate, coordinate, and collaborate to view a situation more objectively, better evaluate the possibilities, and determine the best outcome in a dynamic and adaptable way that adds a layer of objectivity and nuance to decision-making.
Five years later, that prediction has largely materialized. Multi-agent frameworks, retrieval-augmented generation, and tool-using LLMs now orchestrate specialized components to tackle complex problems. The architecture isn’t identical to biological swarm intelligence, but the principle holds: better decisions emerge from coordinated, specialized perspectives, and from understanding the actual purpose of your tools.
What Hasn’t Changed
AI is a powerful solution for a seemingly infinite number of problems. But, much like the internet, it’s easy to get distracted by shiny objects, flashy intrusions, or compelling answers.
It is important to stay mindful and diligent as you apply AI and AI agents to your business.
Many of my friends are getting excited about these tools, and they’re using them for countless capabilities, but they’re not necessarily doing a good job of evaluating whether they should be.
Sometimes, you shouldn’t even be looking for the right answer, you should be looking for the right question.
The Importance of Better Questions
One of the lessons I teach to our younger employees is that an answer is not THE answer. It’s intellectually lazy to think you’re done simply because you come up with a solution. There are often many ways to solve a problem, and the goal is to determine which yields the best results.
Even if you find THE answer, it is likely only THE answer temporarily. It is a step in the right direction that buys you time to learn, improve, and re-evaluate.
Mindfulness comes from slowing down, stepping back, and looking at something from multiple perspectives, and AI can be a powerful tool for that when used intentionally. It can help us explore different viewpoints, challenge assumptions, and think more broadly.
But the greatest benefit of AI may not be in generating better answers. More often, it comes from helping us ask better questions.
Used mindfully, AI becomes less of a shortcut to conclusions and more of a tool for deeper thinking.
Recently, I’ve started using AI to sharpen my questions, and it’s changing the way I approach problems. At first, that sounds abstract, but in practice it forces a very different kind of thinking. Instead of immediately searching for conclusions, you start asking what actually makes a question “better” in the first place. How do you move from a vague sense of uncertainty to a question precise enough to reveal something useful?
When I’m evaluating a project now, I rarely ask AI something broad like, “Is this a good opportunity?” Questions like that usually produce predictable answers. Instead, I use AI to pressure-test my own thinking. I’ll ask it to identify the assumptions underneath the idea, explore what would have to be true for the project to fail, or point out the questions I haven’t considered yet. The process feels less like outsourcing thought and more like refining it.
That shift — from answer-seeking to question-sharpening — has changed how I handle ambiguity and make decisions. It has also changed what I consider trustworthy. I’ve started building what I think of as a “question pattern library”: prompts and frameworks that consistently help add structure to messy situations. Some questions help clarify the framing by forcing you to define the real decision being made rather than reacting to surface-level symptoms. Others establish criteria, helping determine how success should actually be measured before debating solutions. And some are designed to expose bottlenecks by identifying which assumption, if proven false, would completely change the next step.
Over time, I’ve realized these questions work best when they build on each other. At important checkpoints, I’ll often run through a simple sequence: What became clearer? What does this change? Why does it matter? What’s the next best move? The answers themselves matter less than the way the questions force clearer thinking.
The more I use AI this way, the more I think its greatest value may not be generating better answers at all. Used mindfully, its real strength is helping us examine our own thinking more carefully. Better questions create better distinctions, and better distinctions usually lead to better judgment. So before asking AI for an answer this week, it may be worth asking it to help you frame a better question first. You might discover that the most valuable part of the interaction isn’t the response, but the thinking process that led to it.
So whether you are a glass-half-full or a glass-half-empty person, you have plenty of ammunition.
The news cycle is designed to monetize fear, so it reliably amplifies what is fragile, broken, or uncertain. But if you shift focus from the headlines to the data, the global economy in 2026 looks far more resilient (and more opportunity-rich) than most people realize.
In this week’s commentary, I’ll walk through a few key charts that cut through the noise and highlight where growth, risk, and leverage are actually shifting.
The world economy is slated to reach $126 trillion this year, with four countries accounting for over half of that. Who tops the list?
The United States. As we have for over 100 years.
The graphic below visualizes the global economy as a whole using IMF projections from the April 2026 World Economic Outlook, breaking down nearly 200 countries by their share of nominal GDP.
Infographic showing just four countries generate roughly half of all economic activity worldwide.
Just four countries generate roughly half of all economic activity worldwide (U.S. ~$32T, China ~$21T, Germany ~$5T, and Japan ~$4T ). That concentration of economic power is striking, but as we’ll see, size alone doesn’t tell you who’s winning the next decade.
Size Doesn’t Equal Speed
Among the four largest economies, China is expected to lead with a projected 4.4% real growth in 2026, while the U.S. is anticipated to grow a solid 2.3%. In contrast, Germany and Japan (which have experienced years of stagnation) are forecast to grow only around 0.7–0.8%.
China’s strong performance continues a trend observed over the past several decades, despite facing challenges such as a demographic slowdown and an ongoing property sector crisis.
Once you look past the largest economies, there are real opportunities in large, fast‑growing markets across Asia. For example, India, at roughly $4.2 trillion in GDP, and Indonesia, at $1.5 trillion, are on track to play a much bigger role in the global order.
With a forecasted 6.6% growth rate in 2026, India could surpass the United Kingdom and potentially Japan by 2028 — driven by a demographic dividend, expanding services exports, and rapidly maturing digital infrastructure. For entrepreneurs and investors, that shift isn’t just trivia; it should inform where you place bets, partner, and build.
Tariffs, Trade, and the Debt Behind It All
Since early 2025, high-tariff policies implemented by the U.S. have caused downward revisions in growth forecasts for several economies, especially in North America.
Canada and Mexico are especially exposed. With U.S.-Canada relations strained and negotiations over a trilateral trade agreement progressing slowly, the North American economic bloc faces increasing uncertainty.
After World War II, it took over 60 years for U.S. debt to reach $10 trillion. The next $10 trillion took 9 years to reach following the 2008 financial crisis. In the 2020s, pandemic spending compressed the interval to just five years.
By the 2050s, each additional $10 trillion could take just one to two years.
That is under modest assumptions, with no new wars, no recessions, and manageable interest rates. Even so, debt projections still reach $182 trillion by 2056. For context, we’re at about $39 Trillion now.
The real story of the global economy isn’t just told with GDP rankings. While America and China dominate those numbers, it’s clear the landscape is changing.
Traditional economic metrics might become less relevant in a world where regional conflicts, supply chain dynamics, and technological innovation can reshape global power dynamics overnight.
In the longer term, birth rates and the growth of middle-class infrastructure are strong predictors of what lies ahead. That’s part of why we see so much growth in India and Indonesia.
GDP alone doesn’t measure what truly matters in the modern global economy.
The Variable That Changes Everything
Looking beyond traditional economic metrics, I believe artificial intelligence will emerge as one of the most critical factors driving power, progress, and wealth creation in the coming years. It’s likely to become both the most coveted resource and the capability we’ll most actively seek to deny our adversaries.
Economies that combine large markets, strong digital infrastructure, and responsive regulatory environments will be positioned to capture outsized gains. Those that lag on talent, compute, or data governance may see their nominal GDP grow while their strategic leverage erodes.
Obviously, AI is something I think about and write about in many other articles, so even though I won’t add a detailed section here, it’s worth noting that AI is going to change the relative weight and importance of many other things in increasingly exponential ways.
In conclusion, the scoreboard is changing on three fronts at once: where growth lives, how policy shapes risk, and how AI alters productivity and power. If you’re allocating capital or building companies in this environment, the advantage goes to leaders who can see beyond the fear‑driven headlines to where the real leverage is emerging.
I’ve spent enough time in the air to see the system from the inside … and things are definitely changing.
I grew up in a time when business deals were done face-to-face (and that didn’t mean Zoom). I’ve flown over 6 million miles butt-in-seat miles on American Airlines. To put that in perspective, it amounts to hundreds of flights a year at the peak. The kind of travel volume where small details (like upgrades, flight changes, and customer service) stop being luxuries and start being the difference between a manageable routine and a cascade of disasters.
That experience has changed.
I bet you’ve noticed it as well. Upgrades are harder to come by. Lounges are more crowded, and what used to be customer service has become a revenue center. The little efficiencies that made constant travel tolerable have been quietly stripped away.
That’s not just nostalgia. It’s a signal.
And the clearest version of that signal showed up somewhere else entirely.
On the surface, the reasons are straightforward: rising fuel costs, heavy debt, and an unsustainable balance sheet. But those explanations don’t fully answer the more important question … why does a company built around being the lowest-cost option no longer work?
For a long time, the airline industry operated on a relatively stable exchange.
At the bottom, you could sacrifice comfort for price. At the top, loyalty earned you a meaningfully better experience. And in the middle, there was enough balance that both ends could coexist. The average consumer would complain about travel, but not enough to stop them from booking that ticket.
That exchange is breaking down.
Spirit lived at one extreme. It stripped flying down to its bare minimum and charged for everything else. In doing so, it forced the rest of the industry to respond — introducing basic economy tiers and expanding access to cheaper travel.
But that model only works if there’s room to be the absolute lowest-cost option. As costs rise and pricing becomes more sophisticated, that edge disappears. “Cheap” doesn’t go away, but it gets redefined.
When there’s no longer enough margin to operate at that extreme, the model collapses.
Something structurally similar is happening at the other end of the spectrum as well.
Remember When Status Mattered
Elite status used to be scarce. It meant something because relatively few people had it. And, to get it, you had to be a real road warrior.
“I’ve flown over 6 million miles … and that used to mean something to the airline.”
I remember a time when I would see familiar faces on my routine flights. I also remember a time when the airline telephone agent actually knew who I was (and vice versa).
But over time, especially during and after COVID, airlines expanded access. And many of those road warriors have likely switched many of their flights to Zoom calls.
Is the Travel Business Still About Travel?
Credit cards became an alternative (and preferred) pathway to status. Short-term revenue became more important than long-term loyalty.
The reality is that more passengers are competing for fewer upgrades. The same lounge space. The same finite set of perks. The experience gets diluted and devalued.
That’s not an accident. It’s a reflection of where airlines are now making their money.
Breaking Down the Breakdown
Post-pandemic, carriers leaned heavily into premium travel. Higher fares, more segmented cabins, and more ways to extract value from passengers willing to pay for comfort or flexibility. At the same time, rising costs across labor, fuel, and financing have forced a more disciplined approach to pricing.
The system hasn’t gotten worse. It’s gotten more optimized. But optimization changes the experience.
Instead of a clear trade-off between price and comfort, we now have a layered system of constraints and upsells. Economy is fragmented into finer tiers. Premium is more expensive and more protected. And the space between them—where loyalty once created meaningful differentiation — has narrowed.
Which, while a bummer for the price-conscious seasoned traveler, theoretically creates a more distinct experience at the two ends of the spectrum.
That’s why both extremes are under pressure at the same time.
At the bottom, a pure low-cost carrier like Spirit has no room to absorb shocks. At the top, loyalty programs have expanded beyond the capacity of their own benefits. In both cases, the underlying exchange no longer holds the way it used to.
And when that happens, the outcomes start to look familiar.
The middle compresses. The edges strain. The players that survive are either large enough to absorb volatility or differentiated enough to command higher prices.
Airlines aren’t unique in this. You see the same pattern in retail, media, and parts of tech. More efficiency. More segmentation. More options on paper.
But a narrower lived experience.
So yes, flights feel more crowded. Perks feel less reliable. Even with millions of miles behind me, I recently found myself in a middle seat.
But that’s not really the story.
The system is still working. It’s just working differently.
The underlying exchange has shifted. Loyalty no longer buys what it used to. Price no longer guarantees access the way it once did.
More rational. More optimized.
Just not as rewarding for the people who built their routines around the old version.
Last week, while I was in Portland, I noticed that gas prices were over $5, compared to my normal $3 in Texas. And in Texas, even $3 feels high.
Gas prices rank among the most emotionally resonant economic indicators — visible on every corner, cited in earnings calls, and embedded in consumer sentiment surveys for decades.
It made me wonder: where are gas prices the highest … and why? But that doesn’t tell the whole story. Instead, it might be better to look at the average annual gas spend per driver. Below is a chart showing that analysis for each state.
Infographic looking at the average annual gas spent per driver
U.S. drivers spend between $1.6K and $3.3K per year on gas, depending on the state. The spread is significant. What’s counterintuitive is that gas prices aren’t the primary driving factor.
A more useful lens looks beyond the price at the pump and measures the total annual fuel spend against actual miles driven, because often, behavior matters more than cost.
High prices grab attention, but distance quietly does the damage.
As a result, rural states like Wyoming rank highest in annual fuel spend, while Northeast states rank near the bottom — not because gas is cheap there, but because residents drive significantly less.
When prices spike, the narrative focuses on the pump. Consumers feel it immediately. For some, it reshapes household budgets, travel plans, and business decisions.
But this kind of breakdown shows that travel patterns often matter more than price alone.
California is the clearest illustration. The state has the highest per-gallon prices in the country, yet ranks sixth in annual fuel spend at $2,705. Shorter average driving distances (11,780 miles per year versus a national average of 13,916) meaningfully offset the price premium.
The same pattern shows up in reverse in the Northeast.
In New York, drivers spend just $1,582 annually on fuel (about $700 less than the national average) largely because they drive fewer miles (9,185 per year). States like Rhode Island, Delaware, and New Jersey follow a similar pattern, where shorter commutes offset higher gas prices.
Driving The Point Home
Gas prices tell only part of the story. What really drives cost is how much we drive (and, to some extent, that’s demand-elastic). So, while high prices grab attention, decisions to minimize costs by doing less (or doing differently) have wide-ranging impacts.
The other article this week considered what happened to Spirit Airlines. Obviously, fuel costs mattered; too bad they couldn’t have just operated shorter flights … Expect to see more examples of tough choices because of demand-elasticity and rising costs.
For decades, I’ve been an Early Adopter of technologies. I love exploring tools to get an idea of where things are going and what’s possible.
In part, that means I don’t wait for things to settle down and a clear winner to arrive. Instead, I tend to try several tools that claim to do something that excites me.
On one hand, my wife questions whether this is a waste of time, energy, and money. But the practical realities of technology businesses make it a workable strategy for me in my role.
Companies have different levels of access to talent, opportunities, and resources. Consequently, the first tool that does something cool isn’t necessarily the one that takes off or gets big (or the one that continues to play the game, even if it does so slowly, committed to getting better till it wins). This is especially true in highly contested areas like large language models.
For a while, ChatGPT has been my default. Projects tend to start there and end there. It’s been my source of comfort.
Even though I start in ChatGPT, I might then show it to Perplexity and say, “Hey, here’s something I built in ChatGPT. What do you think and what would you change?” This process often results in a new idea or a different perspective. I tend to bring those ideas or perspectives back to ChatGPT, saying, “Hey, Perplexity recommended this … What do you think?”
As you might guess, I’ve tried various iterations of that game. For example, I might start something in Perplexity or Google Gemini … but over time (at least for the type of work that I do), ChatGPT earned its place as my default.
Now, in part, I’m writing this post because Claude has started taking more and more of my cycles: the answers it gives, the user interface, the integrations. It’s really interesting to see how fast it’s improving. Obviously, ChatGPT just released a new interim version to counter the momentum shift Claude is gaining from so much favorable press.
There’s another reason that I know Claude is getting better. It’s still critical of things I produce in other models, but other models are increasingly impressed with what I produce in Claude.
That’s notable because AI systems typically prefer their own outputs. The fact that one model regularly elevates another suggests something else is happening.
Meanwhile, the gap at the top is narrowing. And it’s changing quickly in part because people share outputs from one model with another. This process is a form of cross-pollination that allows LLMs to see (and learn from) a wider range of perspectives and techniques.
So, objectively, which models are really the best? That’s where things get murky. Benchmarks try to answer the question, but they only capture slices of capability.
Lists like this are less a “stamp of approval” and more like a snapshot in time. Models aren’t just getting better every day; new models based on radically different exponential capabilities are being created and released in shorter time cycles, too.
In the list above, Grok-4.20 Expert Mode and OpenAI GPT 5.4 Pro (Vision) tie for the top spot (based on TrackingAI’s April 2026 Mensa Norway benchmark), each scoring 145. The top tier is becoming more crowded, with leading models separated by just a few points. Scores have increased significantly since 2025, demonstrating the rapid progress in frontier AI reasoning on visual pattern-recognition tests. But even that doesn’t account for the fact that ChatGPT’s version 5.5 was released this week.
While this is only one test of AI capabilities, it’s very interesting to see how close the best models have gotten. It’s also worth noting that in 2025, the highest score was 135.
Meanwhile, use of these tools is skyrocketing.
Using cutting-edge AI isn’t a differentiator anymore — it’s the price of admission. The real question isn’t who has the best AI; it’s who can afford to keep up with the pace of change.
Which raises a more important question than “Who’s winning?”:
Can AI Firms Afford to Keep Up?
Last week, I talked about my eldest son lightly teasing me for still trying to overly direct Claude in performing tasks. It’s not just indicative of me getting older … it’s a broader, faster shift.
Early AI development was talent-driven. The limiting factor was human capital — researchers, engineers, and domain experts pushing systems forward.
That constraint is shifting. Today, leading AI firms are increasingly defined by access to compute. Training, fine-tuning, and running these models at scale require massive infrastructure investments, often dwarfing even the highest salaries in tech.
Anthropic spent almost $7 billion on compute in 2025.
Talent still matters, but it’s no longer the primary bottleneck.
Can You Afford To Keep Up?
As companies start leaning more heavily on tools like ChatGPT and Claude, the economics get a little less straightforward.
At first, AI feels like a no-brainer. You’re getting more done, faster— emails, summaries, code, all of it. And the cost? It barely registers. A few cents here and there, easy to ignore. But then usage creeps in. And with automation and agents, what was occasional becomes constant. It gets baked into workflows, products, and day-to-day habits.
And since everything runs on tokens, the meter is always running in the background.
Suddenly, AI stops feeling like “free leverage” and starts acting more like a quiet, always-on teammate. A fantastic and efficient teammate … but one that happens to bill you for every task, and more when you ask it to show its work. At that point, it’s not surprising that the costs can stack up to something meaningful. In reality, AI can cost more than human workers now.
That’s not a knock on AI — it’s just the reality of using industrial-grade AI at scale.
It’s easy to think of AI as a pure efficiency gain, something that just improves margins. But in practice, it’s both sides of the equation. It drives output, and it adds cost. The companies building these tools have always known that. Now the companies using them are starting to see it too.
I’m fully committed to AI, and yet I somehow continue to explore even further. But the deeper you delve, the more important it becomes to pause and catch your breath.
Activity isn’t progress if it doesn’t move you in the right direction.
