Apple’s Vision and AI’s Biggest Questions With Reid Hoffman

Happy Sunday! 

What does disruptive technology look like?

Marques Brownlee, aka MKBHD, doesn’t like the Humane AI pin. He did a fair, well-researched review, but his main idea was that it’s faster and easier to use his phone for common tasks. This is true, but it’s the wrong way to evaluate the potential of new technology. 

Disruptive technology is always worse along the dimensions of performance most valued by mainstream customers. That’s why large incumbents ignore (or ridicule) it. Instead, it’s always better along some new dimension of performance that only a small, low-margin niche of customers care about—but they care about it so much that they don’t mind its flaws in other areas.

The right way to evaluate the Humane AI pin is to ask: What are the situations in which customers might be willing to ignore its obvious flaws for the advantages that the pin form factor provides?

I don’t know the answer, but it’s the most important question to ask if you want to understand the future. For more, read my piece from 2020 on how new ideas happen, which explains how new ideas in both business and science often look terrible—right before they change the world.

Elsewhere this week: Evan Armstrong explains why distribution reigns supreme for building valuable companies and the entertainment potential of the Apple Vision Pro; Rhea Purohit shares how VC investor Jesse Beyroutey and I try to outsmart Wall Street with AI; and I try to predict the inevitable emotional resonance of AI-generated content. On How Do You Use ChatGPT? Reid Hoffman joins me to wax philosophical about AI's existential implications. Plus: our take on the latest tech and business news. —Dan Shipper

Our stories

🎧 "Reid Hoffman on How AI Might Answer Our Biggest Questions" by Dan Shipper/Chain of Thought: Reid Hoffman believes that studying philosophy will make you a better founder than an MBA. In this episode, he sits down with Dan Shipper to discuss how AI is changing what it means to be human and how we can use it to tackle life's biggest questions. Listen to or watch this if you want to learn how to use ChatGPT to clarify your thinking, customize answers, and act as your personal research assistant. 🔏 Paid subscribers have access to the episode transcript.

"Distribution Is King" by Evan Armstrong/Napkin Math: There is a popular saying in Silicon Valley: “First-time founders worry about product, repeat founders worry about distribution.” Evan uses examples from the last 40 years of video games to show just how true that is. Read this to learn how to do distribution that can help your company win. 

🔏 "Inside the Pod: How to Use AI to Be a Smarter Investor" by Rhea Purohit/Chain of Thought: Want to use Google’s Gemini Pro 1.5 to beat the market? Want to figure out the prompts to use to test your investment decisions? Read this to see how Dan and his friend, venture capitalist Jesse Beyroutey, are using AI to deploy capital.

"Don't Dismiss the Apple Vision Pro" by Evan Armstrong/Napkin Math: The Apple Vision Pro made Evan cry—and that's a good thing. In this piece, he argues that despite some flaws, the Vision Pro is a genuine breakthrough in entertainment technology. It may not have much virtual reality content yet, but it's an unparalleled screen for watching movies and being productive. Read this if you want to understand how the Vision Pro could revolutionize the way we interact with media.

“AI Writing Will Feel Real—Eventually” by Dan Shipper/Chain of Thought: Right now, AI writing feels like strumming fake plastic guitars in Guitar Hero—entertaining but artificial. Dan argues that our relationship with AI-generated content will transform as we learn to emotionally connect with it, just as we have with every other new communication technology throughout history. Read this if you want to understand how AI writing will become an integral part of our lives, imbued with real meaning and depth.

---

Chain of links

GPT-4 has a clear eye for medical problems. New research has come out that GPT-4 can achieve the same scores as most specialist medics for ocular problems. LLMs for medical care are useful—we know several people who got answers from ChatGPT that their doctors couldn’t provide—but building a real business for this use case will be challenging. Regulatory capture is real. 

54.8 percent of developers used GitHub Copilot last year. LLM applications that have scaled are few and far between. LLMs for coding are the most successful. They serve an audience that understands tech and a labor market that is wildly understaffed. Win/win. 

Microsoft continues to diversify away from OpenAI. This week Microsoft kicked $1.5 billion into Abu Dhabi-based AI startup G42. This deal is most fascinating from a geopolitical perspective—the company clearly believes that AI will be one of our most important technologies ever. Startup deals have real-world implications, and Microsoft is choosing to bet on the Middle East. 

China has a real ChatGPT competitor. Baidu’s chatbot “Ernie” (a delightful name far superior to any American competitor’s chatbot’s moniker) has 200 million users and 85,000 enterprise clients. Beyond fast fashion like Shein and TikTok, relatively few Chinese apps have made the jump to Western dominance, so it’ll be fascinating to see if LLM products can buck that trend. 

OpenAI begins to release agentic products. The company released an Assistants API that allows the AI to quickly and easily sift through 10,000 files for data. There is much more in the works for release later this year, but to get up to speed, check out my recent essay on the topic. —Evan Armstrong

---

The napkin math

77 boats keep the world running. The internet travels across the planet through 800,000 miles of garden hose-sized fiber optic cables crisscrossing the ocean. Repairing them is challenging, and of the companies that do it, more and more are being acquired by PE vultures, raising fears about the quality of service. The Verge did an incredible investigation—it’s our favorite piece of the week. 

180 million Prime members. Amazon has the most successful consumer membership on the planet. The prime bundle grew 8 percent year over year, with 75 percent of American shoppers now having a Prime membership. These gains are especially impressive when you consider market saturation—it looks like Amazon’s investment in NFL content is paying off. 

A16z raises $7.2 billion. The first fund was just two investing partners and $300 million. Now, the firm’s evolution into a behemoth mirrors the tech sector's rise in power and scope. The fund’s focus includes $600 million for games and $600 million for "American Dynamism"—the latter of which, based on purposefully limited understanding, are tech companies that are legally required to incorporate American flags into their decor. 

SaaS deals are ripping. While most startups are struggling to raise capital, those companies deemed as the “chosen ones” are raising money with ease. Both HR startup Rippling and finance startup Ramp have raised massive growth rounds from top investors. Rippling’s deal is expected to be $200 million in capital for the company, with $670 million in shares sold by existing shareholders. Ramp’s valuation rose 36 percent to $7.65 billion, including the $150 million in capital just raised. —EA

---

For the love of charts

Public SaaS companies on the skids? In the public markets, SaaS companies appear to be struggling. Key growth KPIs are in steady retreat, and have been for the last two years. Customer count growth is slowing:

Retained revenue is shrinking:

Payback periods are getting longer:

So what gives? Is SaaS doing a public face plant? Nope, not exactly. What SaaS is doing is trading growth for profitability:

As Charlie Munger used to say, “Show me the incentive, I’ll show you the outcome.” In this case, when the market rewarded growth, SaaS companies went all-in for it. Now that profit is de rigueur, they’ve reoriented around efficiency (and if they haven’t, investors want very little to do with them). Adapt to survive. —Moses Sternstein

---

Train your mind with tactical mindfulness

Meditation is a powerful tool for business leaders, but it can be hard to do alone. If you’ve ever tried meditation but found it didn’t stick, check out The Hidden Discipline, an eight-week cohort-based mindfulness course designed to instill a meditation practice. Taught by founder-coach and Every columnist Casey Rosengren, the cost is normally $1,000, but Every subscribers can register now for just $699—a 30 percent savings.

---

Eye candy

If Humane could do stuff like this, then we’re sold. 

That’s all for this week! Thanks for reading, and be sure to follow Every on X at @every and on LinkedIn.