AI Never Gets Tired and Always Knows What It Wants

Hello, and happy Sunday! This week we launched Every Studio, our AI product incubator. If you’d like to be among the first to try our earliest prototypes, sign up to be an Early Adopter. Read on for more details, along with everything else we published this week and our take on the latest tech news.—Kate Lee

Was this newsletter forwarded to you? Sign up to get it in your inbox.

---

Knowledge base

"How to Figure Out What People Want" by Dan Shipper/Chain of Thought: What is the key to building products people actually use? Forget about discovering hidden needs. Instead, learn to be the kind of person who can generate wants in others. Dan Shipper argues that needs arise in context, not as pre-existing facts. Read this if you want to understand why "make something people want" is terrible advice—and what to do instead.

🔏 "How to Use AI to Find Profitable Niches on the Internet" by Rhea Purohit/Chain of Thought: Ideas are abundant, but profitable ideas are rare. Rhea Purohit distills wisdom from internet entrepreneurs Steph Smith and Ben Tossell on using AI to find, validate, and execute business ideas online. From nurturing unconventional thoughts to creating your own AI toolkit, this guide offers practical tips for aspiring digital entrepreneurs. Read this if you're ready to turn your internet browsing into a profitable venture.

🔏 "AI Never Gets Tired of You Asking" by Michael Taylor/Also True for Humans: Want better results from AI? Just ask again...and again. Prompt engineer Michael Taylor shares a powerful trick: self-consistency sampling. By generating multiple AI responses to the same prompt, you can overcome inconsistencies and errors. Read this if you want to learn how to make AI work harder for you without complaining (unlike your human employees).

"Introducing Every Studio" by Dan Shipper and Brandon Gell: Every is launching a product incubation arm to develop new software incubation. We’re bringing on five new team members who will build prototypes in days, not months. Read this if you want to see how a media company can leverage its audience to create the next big tech product—and write about it along the way.

---

Fine tuning

Amazon hires Homer Simpson. The company became the latest to tap nuclear power for its data centers with a $500 million investment. Similar deals have been struck by Microsoft and Google, but don’t celebrate too soon. Details of the deal—who pays for development, how the power is purchased, and when cash is distributed—are what matters, and none are public. View these announcements as hopeful signals, not definitive futures. There’s a long way to go before these plants go live.

$525 to build crypto-utopia. Today, a company dares to ask, “What if we made a city that was not sucky?” Praxis raised a large debt vehicle from crypto venture capitalists to answer that question. Maybe that’ll work? A city is a bundle of low-margin assets, like real estate and electrical infrastructure, paired with high-margin subscription fees (aka taxes). Maybe the company will try to do all that? Or more likely, just handle the web3 equivalent of taxes? I’m confused. The list of problems you have to overcome to build this city—namely, who is going to give a bunch of crypto folks land to frolic on—is so overwhelming that I can’t help but admire the level of ambition. Good luck, fellas. 

No, crypto gaming has not “arrived.” Two weeks ago, battle royale game Off the Grid went viral: About 150,000 people watched it on Twitch (a large number for the platform), and the game reached number one in the Epic Store marketplace. Predictably, web3 folks took a victory lap about how crypto gaming was the next bull run. The one problem? No one is playing the game anymore. Just two weeks after debuting, it is the 32nd most streamed game on Twitch and the 11th most downloaded free game in the U.S. By comparison, The Sims 4, which was released in 2014, is sitting three spots above Off the Grid. What happened was quintessential crypto: Game creator Gunzilla Games paid two of the world’s biggest streamers—Ninja and TimTheTatman—to stream the game. As soon as they became less engaged—Tim switched to playing Overwatch 2 for a time—most of the player interest evaporated.—Evan Armstrong 

---

Alignment 

Silicon Valley, meet Pythagoras. Asking “What if?” has sparked more billion-dollar companies and Nobel Prize-winning breakthroughs than you'd find in 1,000 Silicon Valley pitch decks. It's also instinctive for mathematicians, who constantly challenge assumptions and probe the boundaries of what's possible. Michael Atiyah, a Fields Medal-winning mathematician, explained how to turn this question into a discovery framework. Surprisingly, Atiyah's mathematical playbook reads like a startup manifesto. His approach to problem-solving —embracing uncertainty and pivoting at dead ends—applies as much to founding a startup as to proving a theorem. So channel your inner mathematician. Your next breakthrough might be just one bold “What if?” away.—Ashwin Sharma

Collaborative filtering

AI consulting with Every. Are your employees automating 80 percent of their repetitive tasks? If the answer to that question is “no,” you’re behind. Every can help. 

We work with organizations of all sizes—startups, private equity firms, nonprofits, and corporations—to uncover where ChatGPT and custom AI solutions can be incorporated into your work. Our process: 

We conduct a month-long research project and interview members of your team across departments, from the C-suite to individual contributors. 

We share our findings in our proprietary platform, where you can discover all the GPTs your team needs, use cases they can solve with ChatGPT and AI, and a custom-made curriculum broken down by department. You can interrogate our report and interviews using AI to ask your employees your own questions. 

We train your organization on using ChatGPT and so they can build their own AI products, focusing on departmental-level use cases. Our trainings solve the real problems your employees face so that by the end of our engagement, they have both tangible prompts and GPTs they can use in their daily work as well as the knowledge to make more and improve those already made. 

Incorporating AI into everything your business does is critical not just to keep up, but to get ahead. By the end of this process with Every, your team members will be ChatGPT and AI power users—guaranteed. If you’re interested, email Dan at [email protected] with the name of your organization and a few details about the project.

---

Hallucination

A lot can happen in 1,000 years.

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

We also build AI tools for readers like you. Automate repeat writing with Spiral. Organize files automatically with Sparkle. Write something great with Lex.