Jul 15, 2026 · 50 min
How technical missteps reshaped silicon valley and AI birthed vibe coding
Former Intel CEO on What Went Wrong, What's Next + Lovable CEO on the Real Promise of Vibe Coding
Understanding the structural shifts in chip manufacturing and the rise of AI-driven software development is crucial to navigating the next wave of global technological dominance.
- 1Intel lost its technical edge by shifting leadership from engineers to financial managers, leading to costly missed product cycles.
- 2The rapid build-out of artificial intelligence infrastructure is hitting a hard ceiling dictated by global energy grid capacity.
- 3AI-assisted vibe coding platforms are lowering software development barriers, enabling non-technical creators to build full-stack web applications.
Don't miss
Pat Gelsinger details the internal decision-making that led Intel to pass on manufacturing the original iPhone chip, a pivotal moment that altered the course of mobile computing.
The brief
Intel's decline from a dominant chipmaker to a struggling giant reveals the cost of prioritizing financial metrics over technical leadership, as former CEO Pat Gelsinger reflects on missed opportunities like passing on the early iPhone chip.
While Nvidia captured the AI revolution by pairing high-performance GPUs with its CUDA software stack, TSMC's dedicated foundry model transformed global manufacturing, leaving Intel's integrated design-and-build model behind.
The massive build-out of AI infrastructure faces a hard physical limit in global energy capacity, making a 10,000x increase in chip efficiency essential to lower token costs and keep the technology scalable.
The software landscape is shifting rapidly with the rise of vibe coding, where platforms like Lovable allow users to build and deploy complex, secure applications entirely through natural language chat.
What was said on this episode
36 statements · 27 positive · 5 negative · 1 mixed · 3 neutral
Intel declined when business leaders replaced technically oriented leadership.
“one of the things that went off the rail was when it started to be run by business people”
Listen at 2:28
Spreadsheet-driven decisions are inadequate for major technical investments.
“when you're making these hardcore technical, you know, decisions that affect billions of dollars and you don't do that through a spreadsheet, that's a lousy investment”
Listen at 3:09
Technology companies should have technologists leading and hiring technical teams.
“you need technologists running technology that then hires technologists”
Listen at 4:28
Apple prepared core technologies before deciding to develop its own semiconductor designs.
“he had been preparing the core technologies inside of Apple for something that might happen in the future”
Listen at 7:06
Nvidia GPUs can support general-purpose workloads beyond graphics.
“These are general purpose computing devices that can start applying to these other workloads”
Listen at 9:38
TSMC’s open foundry model made semiconductor manufacturing available across chip designers.
“TSMC basically cut that in half and says, I don't care whose chip it is, I don't care what you're designing, I'll be your manufacturing partner”
Listen at 12:44
TSMC produced five times as many wafers as Intel.
“TSMC was producing 5x the wafers of Intel”
Listen at 13:14
The foundry model became the dominant semiconductor industry model.
“that model of Foundry became the model of the semiconductor industry”
Listen at 13:24
The CHIPS Act is increasing U.S. leading-edge semiconductor manufacturing.
“the CHIPS act is having benefit”
Listen at 15:16
The U.S. share of leading-edge semiconductor production rose from about 12% to 18%.
“the US was building about 12% of leading edge. Today that number is more like 18%”
Listen at 15:22
A shut-down semiconductor fabrication plant requires about 90 days to restart.
“When you turn off a fab, it doesn't come back on for 90 days”
Listen at 16:15
A Taiwanese brownout would cause economic damage exceeding the Great Depression.
“The economic impact of a brownout of Taiwan is greater than the Great Depression in the world”
Listen at 16:20
Global semiconductor supply chains should become more resilient.
“We need more resilient supply chains for it”
Listen at 17:17
Limited energy capacity constrains the growth of the AI infrastructure bubble.
“you have an upper bound on how aggressive and how hyped and bubbled that we get”
Listen at 19:03
AI tokens could unlock extremely large economic value.
“the potential value that we unleash in a token economic world is somewhat infinite”
Listen at 19:32
AI infrastructure and adoption will expand over multiple decades.
“I am an optimist that we are in a couple of decade build out”
Listen at 19:45
AI systems need to improve by 10,000 times in efficiency or capability.
“I have to make AI 10,000x better”
Listen at 19:56
The cost per AI token should fall by five orders of magnitude.
“we want to drop, you know, by five orders of magnitude the cost per token”
Listen at 20:03
Excessively high AI-company valuation multiples will periodically correct.
“anytime the multiples get too high, okay, some corrections”
Listen at 21:51
Quantum computing will become meaningful during this decade.
“This decade”
Listen at 22:48
Quantum computers will solve problems currently beyond conventional computation.
“you're going to be able to start doing things that cannot be computed today”
Listen at 22:57
Quantum computing will demonstrate supremacy results across multiple industries before 2030.
“this decade we will see quantum supremacy results across multiple industries”
Listen at 23:30
Lovable users build approximately one million new projects weekly.
“We're seeing a million new projects built every single week on the platform”
Listen at 25:45
Lovable-hosted applications receive over 700 million monthly visits.
“more than 700 million visits to the applications every month”
Listen at 26:10
Four-fifths of Lovable users are nontechnical.
“4 out of 5 are non technical”
Listen at 28:36
Lovable structures software architecture and helps prevent serious implementation failures.
“it creates a structure for the architecture of the software that you build and it makes sure that you don't go off a cliff”
Listen at 30:26
An employee built the Foundry University intranet in four hours using Lovable.
“It was built in four hours by an employee”
Listen at 33:48
Lovable is developing an AI cofounder that recommends strategic business directions.
“a co founder that works for you even when you're sleeping and comes back to you in the morning and says here are some strategic directions”
Listen at 36:04
Organizations will increasingly use bespoke software for specialized requirements.
“you will have more bespoke solutions”
Listen at 38:59
Lovable can provide bespoke interfaces while retaining existing enterprise software underneath.
“you can continue to use Salesforce, HubSpot and all the tools that you like to use under the hood, but with a bespoke interface on top of it”
Listen at 39:32
Lovable uses multiple AI models to serve customer requests.
“we're using multiple models”
Listen at 40:32
Lovable uses customer-relevant error data and reinforcement learning to improve model performance.
“we make the models, we create data sets. We did something called reinforcement learning specifically for the problems where the frontier models are making mistakes for us right now”
Listen at 42:15
AI generally accelerates work enough to justify its expense.
“AI usually lets you move much faster, so the spend is usually worth it”
Listen at 44:50
Rapid experimentation is valuable for software development.
“I'm actually a huge fan of very rapid experimentation”
Listen at 45:50
AI is improving implementation faster than product strategy and problem selection.
“more of the bottleneck. Whereas more intelligence is on some tasks. It's great. It creates really beautiful things, 3D games, for example. But on figuring what to build, figuring out what are the right strategic directions or experiments you should run to improve the outcomes for your business, that's not changing as fast”
Listen at 47:55
Businesses should try Lovable.
“Lovable is absolutely worth your time”
Listen at 48:37
Statements are attributed to the speaker as said on the episode and reflect their view at the time, not PodLume's. They are not advice.
Books & mentions
All-In with Chamath, Jason, Sacks & Friedberg
To hear more deep dives into the macroeconomic and geopolitical forces shaping the technology sector