Chamath Palihapitiya

What I Read This Week…

In his first week back in office, President Trump issued a series of executive orders aimed at rapidly implementing his agenda.

His most significant actions focused on immigration and border security, declaring a national emergency at the U.S.-Mexico border to authorize military deployment and resume wall construction, while suspending the refugee program and designating certain international cartels as terrorist organizations.

On climate policy, Trump withdrew the U.S. from the Paris Agreement and expanded areas for oil extraction, reversing Biden-era environmental protections.

For federal workforce management, he mandated full-time in-person work, implemented hiring freezes, and streamlined the process for terminating employees.

The orders also included major shifts in domestic policy, with directives to address the cost-of-living crisis and restore freedom of speech, while dismantling diversity initiatives in federal agencies.

In foreign policy, Trump suspended foreign aid programs for 90 days and announced plans for new tariffs effective February 2025.

These executive actions, which included the revocation of approximately 80 Biden administration policies, signal Trump's determination to implement his "America First" agenda.

In December 2024, DeepSeek released its V3 model with 671 billion parameters, trained for just $5.58 million—a fraction of what other major AI labs typically spend.

This model, created by AI researchers in China, matched industry leaders like GPT-4 and Claude 3.5 Sonnet in benchmarks while being fully open source.

Following this, a few days ago, DeepSeek released DeepSeek-R1, a reasoning model that achieved strong results on mathematical and logical tasks, and the DeepSeek app is now the most-downloaded free app in the U.S. on Apple’s App Store.

DeepSeek demonstrates that high-performing AI models can be built with fewer resources than previously thought necessary and that China has developed advanced AI capabilities despite U.S. export controls on GPUs.

As more high-quality language models become available for free, market forces will drive down prices across the industry.

Moreover, a Chinese AI company releasing such a capable open-source model should be a wake-up call for America.

Leadership in AI will not only determine economic primacy in this era but will also be increasingly important for military effectiveness via enhanced decision-making and autonomous systems.

(The team and I are currently putting together a follow-up AI Deep Dive that will explain the story from the transformer model in 2017 to this current moment, which will be published in a few days.)

Chinese social media apps have engineered a uniquely addictive formula by seamlessly blending social networking, entertainment, and e-commerce.

While Western platforms like Instagram and YouTube remain focused on advertising revenue, Chinese apps like TikTok, Xiaohongshu, and Lemon8 have integrated shopping directly into their social experience, creating what industry experts call "shoppertainment."

This model has proven remarkably successful – Douyin (TikTok's Chinese counterpart) reported $2 billion in platform sales in 2023, while Xiaohongshu achieved its first net profit of $500 million.

The success stems from extremely powerful algorithms that not only surface viral trends but also compel impulse purchases, making China the global leader in social commerce with nearly half of users reporting recent in-app purchases in 2024.

When TikTok briefly shut down in January 2025, American users flocked to these Chinese alternatives, drawn to their hyper-targeted recommendations.

More below...
http://chamath.substack.com/p/what-i-read-this-week-db6

Deep Dive: The Media Landscape - Sense-making 2.0

Last week, Elon was on Joe Rogan's podcast talking about how the Associated Press had become a propaganda and misinformation platform.

Three months ago, Jon Stewart was on Ezra Klein's podcast criticizing the "moralizing nature" of traditional media.

Two years ago, Marty Baron commented in an interview that, after the rise of the internet, "the marginal cost of gaining an additional subscriber was essentially zero" and the lack of viability for local news is the "biggest crisis in journalism today."

At the same time, everyone is experiencing polarization and radicalization not just on the internet, but in their daily lives as well. People blame "the algorithm" for mindless scrolling, yet we still spend hours on these platforms, watching YouTube videos, browsing TikTok, and listening to podcasts (including the All-In Podcast) while driving or going for a run.

We're witnessing a profound transformation in how information travels through society. In just the past decade, we've seen the collapse of local and regional newspapers, social media becoming a primary news source for many Americans, and the emergence of creators and influencers who command larger audiences than entire media organizations.

So how do these things all connect, and what's going on?

This deep dive examines the evolution of media from the invention of the printing press all the way to the rise of algorithmic systems like Facebook and TikTok. We dive into the mechanics of how these algorithmic systems work, as well as how these systems cluster different subsets of the population into echo chambers, which explains why people are experiencing different realities from one another.

From this lens, we then analyze the decline of traditional media and the simultaneous rise of a new media ecosystem that's recreating the core functions of journalism, but leveraging technology to create more compelling and information-rich experiences.

One layer we'll focus on in this new media ecosystem are the "sense-makers" – individuals who leverage expertise and authenticity rather than institutional authority to help people make sense of an increasingly chaotic information landscape.

For creators and influencers, this deep dive offers a broad lens to understand the power of new media and a creator playbook that may permeate more aspects of American society, including running for U.S. President.

When I reflect on my entrepreneurial and investing career, I see it as the end product of thousands of decisions I made daily. Those decisions were based on the information I had consumed over time and how I shifted my information diet in key moments.

In many ways, whoever has the highest-quality information and the deepest understanding of topics forms the best judgement, and those people inevitably become more successful over the arc of time.

I hope you enjoy this one. Let me know what you think.

You can read our deep dive here:
http://chamath.substack.com/p/deep-dive-the-media-landscape-sense

We've talked a lot about this on the Pod, but the Great SaaS Meltdown has started and there's no going back.

What exactly is happening?

In short, hi growth, low/no profitability SaaS is no longer a winning strategy because the big question mark is the durability of that growth