Two Guys, Two Weeks, Two Hours: Episode 5
Phil (@pbeisel) and Cern talk Tesla & SpaceX, and are joined by @KenyonWells5 who discusses why AI needs to deliver results (and soon).
The Human Reckoning: Why AI’s Biggest Risk Isn’t Politics or Money — It’s Us (Trilogy Closer)
If you’ve been following the first two pieces in this series, you know the drill. The Final Countdown laid out the political headwinds. The Dutch disease sequel showed how the investment flood risks crowding out everything else economically. This final piece is about the people part: the human reckoning that could decide whether AI becomes something we embrace or something we eventually push back against.
The “AI jobspocalypse” narrative gets a lot of oxygen because it’s simple and scary. AI can do many tasks quicker and cheaper, and on pure cost-and-speed metrics, it usually wins. But most real-world value isn’t just about cost and speed. It’s about trust, judgment, empathy, creativity, accountability, and nuance. In those cases, AI + Human hybrids almost always deliver better overall value than AI alone.
We also need to retire a couple of other lazy narratives. The “nobody will need to work” utopia (or dystopia) misses basic human nature. People may not need to work just to survive if AI drives down costs for basics, but they will absolutely work to thrive — to get ahead, to have more than the neighbor or the brother-in-law, to build status or something meaningful. That competitive streak isn’t going away.
At the same time, the “everybody will be an entrepreneur” story is equally unrealistic. Not everyone wants to (or can) run their own thing. Larger companies aren’t disappearing — they’ll still need skilled, AI-augmented people who can add real value at scale.
The real risk isn’t mass unemployment or universal entrepreneurship. It’s that we botch the transition — leaving too many people without clear paths to the new hybrid roles, widening inequality, and eroding trust when AI screw-ups get blamed on “the system” with no human face to own it.
The fix has to be deliberate: education that teaches AI collaboration and judgment, industry reporting on hybrid outcomes, and trust-building design with human oversight where it matters. The countdown from the first piece still applies. By 2028, voters will judge AI on whether their own lives feel better — more productive days, better services that still feel human, and real pathways ahead.
We’ve got the politics warning, the economic risk map, and now the human playbook. AI alone is powerful. AI + thoughtful humans who still want to strive and build is where the real advantage lives. Let’s build for that.
Wooden Shoes Alert: AI’s Investment Boom Is Giving America “Dutch Disease” (And It’s Time to Retire the Gold Rush Picks-and-Shovels Story)
If you read the first piece in this series (The Final Countdown), you’ll see the political risks clearly. This one looks at the economic side: the very real risk that our massive AI investment boom is turning into its own flavor of Dutch disease — sucking up capital, talent, and focus while leaving the broader economy thinner, more concentrated, and vulnerable.
First, a quick explainer for the handful of non-econ geeks reading this. Classic Dutch disease comes from the Netherlands in the 1960s after they struck it rich with natural gas. The boom in that one sector strengthened the currency, made everything else less competitive, and pulled workers and investment away from manufacturing and other industries.
Now swap natural gas for server farms and GPUs. In 2026, the big hyperscalers — Amazon, Microsoft, Google/Alphabet, and Meta — are on track to drop a staggering $725 billion in capital expenditures, the vast majority tied to AI infrastructure. Amazon alone is guiding toward $200 billion. Microsoft around $190 billion. Google in the $180–$190 billion range. Meta bumped its number up to $125–$145 billion.
The crowding-out is already showing up. Capital and top talent flow heavily into a handful of AI-heavy firms. Manufacturing adoption is real but slower. Regional companies struggle. Energy gets locked up by data centers.
It’s past time to retire the tired old “gold rush” analogy. AI isn’t some speculative rush anymore; it’s mature enough that all companies should be actively implementing it deep into operations. The ones doing it well should be rewarded in valuations.
That means investors and analysts need to raise their game. Ask specifics on earnings calls: token cost per employee? Content optimized for LLMs? % of employees with enterprise AI licenses? % of customer interactions that are 100% AI or AI-assisted? How has AI reduced COGS or CAC? What % of processes are AI-augmented? Measured ROI on scaled pilots?
Affordability remains the key consumer issue — and this investment concentration makes solving it harder. People feel it in stagnant wages outside tech hubs and higher energy costs. They haven’t felt the broad productivity lift yet.
Here’s the smarter play: Spread the investment so AI helps factories, small businesses, and hospitals, not just hyperscalers. Reward companies that deliver hybrid value. The countdown from the first piece still applies — 2028 will judge whether this boom strengthens the whole economy or not.
The Final Countdown: Why 2028 Is Make-or-Break for AI (and Why Groceries Matter More Than Grand Promises)
Look, we’re in the final countdown here. AI and data center construction have become highly partisan — think AOC and Bernie pushing moratorium-style ideas, or actual legislative attempts like the one in Maine. The 2026 midterms matter for near-term regulation, but 2028 is the real turning point. If anti-AI sentiment becomes a political winner, both sides will race to strangle the industry.
Industry messaging isn’t helping. One side hypes the risks of doom as a marketing approach. The other plays defense by debunking while trying to paint a bright future. Voters are living hard times now, and the #1 issue remains affordability — especially groceries. If the AI industry can “pull forward” some of that deflationary effect so the average person actually sees and feels lower costs, neither party will make killing AI a plank in 2028.
Here’s a quick household budget breakdown (BLS data) to show where it matters most:
• Groceries/Food (~13% of spending): AI in supply chains, predictive inventory, and precision agriculture (Amazon, Walmart, and startups) can drive real deflation here.
• Housing/Transport/Energy (~50% combined): Tesla and others in autonomous + energy optimization; DeepMind-style efficiency in grids and data centers.
• Healthcare/Admin (~8-10%): Automation of billing and records, but broad longevity breakthroughs by 2028 remain unlikely.
• Retail/Consumer Goods: Amazon and others using AI for logistics and personalization.
The smarter path is clear: diffuse the benefits fast so voters feel them. Companies that deliver visible cost reductions in groceries, energy, and daily bills will make it much harder for politicians to run against AI.
The countdown is real. Get visible wins on affordability and hybrid value, or the politics will turn fast.